european weed research society - ewrs · zaragoza, spain 9-11 march 2009. european weed research...

EUROPEANWEED

RESEARCHSOCIETY

Proceedings8th EWRS Workshop on

Physical and Cultural Weed ControlZaragoza, Spain

9-11 March 2009

EUROPEAN WEED RESEARCH SOCIETY

Proceedings

8th EWRS Workshop on Physical and Cultural Weed Control

Zaragoza, Spain

9-11 March 2009

ii

Abstracts were compiled and edited by:

Daniel C. Cloutier Weed Science Institute 102 Brentwood Rd. Beaconsfield (Québec) H9W 4M3 Canada Tel.: +1 514 695-2365 E-mail: [email protected]

Scientific organisers

Bo Melander University of Aarhus Faculty of Agricultural Sciences Department of Integrated Pest Management Research Centre Flakkebjerg Forsøgsvej 1 DK-4200 Slagelse Tel: +45 8999 3593 E-mail: [email protected]

Alicia Cirujeda Centro de Investigación y Tecnología Agroalimentaria Av. de Montañana, 930 50059 Zaragoza Spain Tel: + 34 976.71.40.00 ext. 2032 Fax.: + 34 976.71.63.35 E-mail: [email protected]

Carlos Zaragoza Larios Centro de Investigación y Tecnología Agroalimentaria Av. de Montañana, 930 50059 Zaragoza Spain Tel.: +34 976.71.63.22 Fax.: + 34 976.71.63.35 E-mail: [email protected]

Daniel C. Cloutier Weed Science Institute 102 Brentwood Rd. Beaconsfield (Québec) H9W 4M3 Canada Tel.: +1 514 695-2365 E-mail: [email protected]

Local organisers

Alicia Cirujeda Centro de Investigación y Tecnología Agroalimentaria Av. de Montañana, 930 50059 Zaragoza Spain Tel: + 34 976.71.40.00 ext. 2032 Fax.: + 34 976.71.63.35 E-mail: [email protected]

Carlos Zaragoza Larios Centro de Investigación y Tecnología Agroalimentaria Av. de Montañana, 930 50059 Zaragoza Spain Tel.: +34 976.71.63.22 Fax.: + 34 976.71.63.35 E-mail: [email protected]

Joaquín Aibar Escuela Politécnica Superior de Huesca University of Zaragoza 22071 Huesca, Spain Tel: +34 974 239 417 Fax: +34 974 239 302 E-mail: [email protected]

Sonsoles Fernández-Cavada Centro de Protección Vegetal Dept. de Agricultura y Alimentación Gobierno de Aragón Av. de Montañana, 930 50059 Zaragoza Spain Tel.: +34 976.71.63.79 Fax.: + 34 976.71.63.88 E-mail: [email protected]

Produced April 23, 2009

iii

Group photo (courtesy of Jesper Rasmussen)

iv

Table of contents Preventive and cultural methods Weed suppression by canola and mustard cultivars H.J. Beckie, E. N. Johnson, R. E. Blackshaw and Y. Gan (CAN) ......................................................2 Effect of innovative crop and weed management systems on organic cauliflower in Central Italy S. Carlesi, F. Bigongiali, D. Antichi, M. Fontanelli, C. Frasconi, L. Lulli, F. Sorelli and P. Bàrberi (ITA) ............................................................................................................................3 Mechanical barriers for Cyperus esculentus (yellow nutsedge) control in strawberry O. Daugovish and M. J. Mochizuki (USA) .........................................................................................4 Weed patchiness: implications for physical and cultural control C. Fernández-Quintanilla, J. Dorado, D. Andújar and D. Ruiz (ESP) ................................................5 An integrated physical approach to control purple nutsedge (Cyperus rotundus) Hershenhorn, J., Weissblum, A., Dor, E., Lande, T., Achdary G. and Smirnov, E. (ISR)..................6 Combining enhanced competition and cultivation for improved weed control in organic cereals L.N. Kolb, T. Molloy, and E.R. Gallandt (USA) ................................................................................7 Using ecological processes to manage cropping systems for weed suppression and other services in the U.S. Corn Belt Matt Liebman, Paula R. Westerman, Andrew H. Heggenstaller, Carol L. Williams, Brent J. Danielson, Mark D. Tomer, and Michelle M. Wander (USA) ............................................................8 Integrated winter/summer cover crops into weed management program in organically grown vegetables H. Mennan, M. Ngouajio, D. Isık and E. Kaya (TUR)........................................................................9 Trials of a crimper-roller for killing cover crops for organic and non-herbicide, no-till cropping C.N. Merfield (IRL)...........................................................................................................................11 Effects of cereal rye mulch and soybean density on weed suppression M.R. Ryan, D.A. Mortensen, S.B. Mirsky, J.R. Teasdale and W.S. Curran (USA) .........................16 Comparison of several recommended cultural practices for weed management and their effects on yield of coconut in tropical coconut plantations S.H.S. Senarathne and M.J.I. Costa (LKA) .......................................................................................17 The relative importance of cultural weed control methods: A survey of results from western Canada Steven J. Shirtliffe, Eric N. Johnson, Dilshan Benaragama, Yvonne E. Lawley, and Julia M. Baird. (CAN) ................................................................................................................23

v

Physical and preventive weed control in uncultivated areas The invasion of weeds in the archaeological sites and innovated methods for their control Economou G, Papafotiou M. and I, Kanellou (GRC) ..................................................................... 25 Results and experiences of physical weed control on hard surfaces D. Hansson and H. Schroeder (SWE) ............................................................................................... 26 Experiences with physical weed control on hard surfaces in central Italy L. Lulli, M. Fontanelli, C. Frasconi, M. Ginanni, M. Raffaelli, F. Sorelli and A. Peruzzi (ITA)..... 27 Weed problems on pavements B. Melander, N. Holst, A.C. Grundy, C. Kempenaar, M.M. Riemens, A. Verschwele and D. Hansson (DNK) ..................................................................................................................... 28 Mechanical weed control Morphological differences between carrot and weeds: its usefulness in selective mowing as a weed control technique Diane Lyse Benoit (CAN) ................................................................................................................ 30 New innovations for intra-row weed control Pieter Bleeker and Rommie van der Weide (NLD) ......................................................................... 31 The use of flex-tine harrow, torsion weeder and finger weeder in Mediterranean crops A. Cirujeda, J. Aibar, S. Fernández-Cavada, P. Zuriaga, A. Anzalone and C. Zaragoza (ESP) ...... 33 In pursuit of effective mechanical/physical weed management in organic lo-till W.S. Curran, R.T. Bates, S.B. Mirsky, R.S. Gallagher, D.A. Mortensen and M.R. Ryan (USA).... 34 Innovative operative machines for physical weed control on organic cauliflower in Central Italy M. Fontanelli, C. Frasconi, L. Lulli, F. Sorelli, S. Carlesi, F. Bigongiali, D. Antichi and A. Peruzzi (ITA) ......................................................................................................................... 35 Innovative operative machines for physical weed control on processing tomato in the Serchio Valley (Central Italy) Fontanelli M., Raffaelli M., Ginanni M., Lulli L., Frasconi C., Sorelli F. and Peruzzi A. (ITA)..... 41 Non-chemical weed control on open-field fresh market tomato in the Serchio Valley (Central Italy) Fontanelli M., Raffaelli M., Ginanni M., Lulli L., Frasconi C., Sorelli F. and Peruzzi A. (ITA)..... 49 Physical weed control on cabbage in the Serchio Valley (Central Italy) M. Fontanelli , M. Raffaelli , M. Ginanni , L. Lulli , C. Frasconi , F. Sorelli and A. Peruzzi (ITA) ........................................................................................................................ 56 Managing weed seed rain to enhance physical weed control efforts E.R. Gallandt (USA) ......................................................................................................................... 66

vi

Performance of a min-till rotary hoe in field pea (Pisum sativum L.) E. N. Johnson and S.J. Shirtliffe (CAN) ...........................................................................................67 Current weed management and the problem of the highly adaptive, cosmopolitan weeds L. Radics and M. Glemnitz (HUN) ...................................................................................................68 Physical weed control in protected leaf-beet in Central Italy M. Raffaelli, M. Fontanelli, C. Frasconi, L. Lulli, M. Ginanni, F. Sorelli and A. Peruzzi (ITA) ....69 Sensor based selective weed harrowing in cereals in Germany Victor Rueda-Ayala and Roland Gerhards (GER) ............................................................................76 Comparison of three tillage intensities on grass weed occurrence in cereal rotation J. Salonen (FIN).................................................................................................................................77 Evaluation of finger and torsion weeders for cultivating cool season vegetables in Salinas, CA, USA R.F. Smith and M. Silva Ruiz (USA) ................................................................................................78 Using a spring-tine harrow to weed a navy bean crop (Phaseolus vulgaris var. Tabella Brisa) destined for human consumption. Taberner A., Llenes JM., Roque A., Orri J. and Miralles M. (ESP) .................................................79 Use of the flex-tine harrow for grain production in an Amaranth crop. Taberner A., Zamora N., Llenes J.M. and Roque A. (ESP) ..............................................................80 Thermal weed control Thermal weed control – a review of current techniques J. Ascard (SWE) ................................................................................................................................82 Weed control with steam and solarization for field-grown flowers and strawberry (Fragaria ananassa L.) Celeste A. Gilbert, Steven A. Fennimore, Krishna Subbarao, Rachael Goodhue, J. Ben Weber and Jayesh B. Samtani (USA).....................................................................................83 Growth stage impacts tolerance to broadcast flaming in agronomic crops S.Z. Knezevic, A. Datta and S.M. Ulloa (USA) ...............................................................................86 Response of corn (Zea mays L.) types to broadcast flaming S.Z. Knezevic, C.M. da Costa, S.M. Ulloa and A. Datta (USA).......................................................92 Tolerance of selected weed species to broadcast flaming at different growth stages S.Z. Knezevic, A. Datta and S.M. Ulloa (USA)................................................................................98 Winter wheat (Triticum aestivum L.) tolerance to broadcast flaming S.Z. Knezevic, J.F. Neto, S.M. Ulloa and A. Datta (USA) .............................................................104 Dose-response of weeds to flaming M.L. Leblanc, D.C. Cloutier and E. Sivesind, E (CAN) .................................................................111

vii

Solarization as a tool for non-chemical weed management Baruch Rubin (ISR) ........................................................................................................................ 112 Effects of soil steaming on weed seed viability F. Vidotto, M. Letey and D. Ricauda-Aimonino (ITA) .................................................................. 113 Various weed control systems Autonomous navigation in weed-infested maize fields Abadia D., Ballano S., Uson F., Paniagua J., Seco T., Cirujeda A. and Zaragoza C. (ESP) .......... 115 Vision based crop plant identification for weeding operations Abadia D., Gonzalez S., del-Hoyo R., Paniagua J., Seco T., Cirujeda A. and Zaragoza C. (ESP) 116 Control of docks (Rumex spp.) in organic fodder production – experiments for optimizing the soil tillage effect on docks when renewing highly infested areas L. O Brandsæter, K. Mangerud (NOR) ........................................................................................... 117 Evaluation of different mulches for weed control in processing tomato A. Cirujeda, J. Aibar, A. Anzalone, M. Gutierrez, S. Fernández-Cavada, A. Pardo, Mª L. Suso, A. Royo, L. Martín-Closas, J. Costa, A. M. Pelacho, M.M. Moreno, A. Moreno, R. Meco, I. Lahoz, J.I. Macua and C. Zaragoza (ESP) ........................................................................................................ 118 Is yardwaste mulch a weed-free substrate? O. Daugovish, B. Faber and J. Downer (USA) ............................................................................... 119 Management issues related to the use of the herbicide glyphosate and the transformations operated on urban vegetation in Genoa (Northern Italy). First note. A. Di Turi and G. Paola (ITA) ........................................................................................................ 120 Possibility of using mustard meal of Sinapis alba and Brassica juncea for weed control Fredrik Fogelberg (SWE)................................................................................................................ 121 Impact of weed-control mulches and disks on fertilizer placement and water use in nursery container production M. Lanthier, S. Peters, J. Atland, S. Harel (CAN) .......................................................................... 125 Weed community response to six IWM systems in a four-year crop rotation A. Légère, A.G. Thomas, J.Y. Leeson, F.C. Stevenson, F.A. Holm, B. Gradin and D Kratchmer (CAN) ................................................................................................................. 126 Influence of different soil cultivation systems on weed population in soybeans G. Malidza, S. Vrbnicanin, I. Kurjacki and D. Pavlovic (SRB)...................................................... 127 Weed population dynamics in fields with different management D. Piliksere (LVA) .......................................................................................................................... 128

viii

Effect of Elymus repens on yield of winter wheat, spring barley and faba bean in an organic crop rotation experiment. I.A. Rasmussen, M. Sønderskov, C. Damgaard and K. Kristensen (DNK) ....................................129 Identifying weed distribution using soil properties H.Salehian and S.Soltani (IRN).......................................................................................................130 Control of weeds in flooded rice by non-chemical means Taberner A., Cónsola S., Llenes JM., Roque A. (ESP) ...................................................................138 False seedbeds in organic grown winter wheat A.Verschwele (DEU).......................................................................................................................139 Round table reports Report of the Round Table “Relevant and Non-relevant Parameters When Studying Cover Crop and Mulching Effects.” Eric Gallandt ....................................................................................................................................141 Report of the Round Table “Unifying parameters in mechanical weed control research” Jesper Rasmussen ............................................................................................................................142 Report of the Round Table “Research methodology in thermal weed control” Johan Ascard and Daniel Cloutier ...................................................................................................146

8th EWRS Workshop on Physical and Cultural Weed Control 1 Zaragoza, Spain, 9-11 March 2009

Preventive and cultural methods


Weed suppression by canola and mustard cultivars

H.J. Beckie1, E. N. Johnson*2, R. E. Blackshaw3 and Y. Gan4

*2presenter Agriculture and Agri-Food Canada (AAFC), Box 10, Scott, SK, Canada S0K 4A0 email: [email protected]; 1AAFC, Saskatoon, SK, Canada; 3AAFC, Lethbridge, AB, Canada; 4AAFC,

Swift Current, SK, Canada.

Competitive crops or cultivars can be an important component of integrated weed management systems. A study was conducted from 2003 to 2006 at four sites across semiarid prairie ecoregions in western Canada to investigate the weed-suppression ability of canola and mustard cultivars. Four open-pollinated canola cultivars, four hybrid canola (Brassica napus L.) cultivars, two canola-quality mustard (Brassica juncea L. Czern. & Coss) cultivars, two oriental mustard (Brassica juncea L. Czern. & Coss) cultivars, and two yellow mustard (Sinapis alba L.) cultivars were grown in competition with indigenous weed communities. Yellow mustard was best able to suppress weed growth, followed in decreasing order by oriental mustard and hybrid canola, open-pollinated canola, and canola-quality mustard. Competitive response of cultivars, assessed by weed biomass suppression, was negatively correlated with time to crop emergence and positively correlated with early-season crop biomass accumulation (prior to bolting) and plant height.


Effect of innovative crop and weed management systems on organic cauliflower in Central Italy

S. Carlesi, F. Bigongiali, D. Antichi, M. Fontanelli, C. Frasconi, L. Lulli, F. Sorelli and P. Bàrberi

Land Lab, Scuola Superiore Sant'Anna, Piazza Martiri della Libertà 33, 56127 Pisa, Italy, [email protected]

The aim of this study was to compare three crop and weed management systems in an on-farm research project on organic vegetable production. The three systems tested, corresponding to increasing levels of innovation, were: the Standard Crop Management System (SCMS), an Intermediate Crop Management System (ICMS) and an Advanced Crop Management System (ACMS) applied to the same crop sequence (spinach-potato-cauliflower-tomato). The SCMS is the ordinary crop management system practised on farm, the ICMS consists in the use of innovative machines for physical weed control such as the rolling harrow, patented by MAMA-DAGA, University of Pisa, and the ACMS has the same features of ICMS plus the inclusion of hairy vetch (Vicia villosa Roth) as living mulch crop interseeded in cauliflower.

The field experiment was carried out at the Colombini vegetable organic farm in Crespina, Pisa (lat. 43°35′ N, long. 10 34′ E) in 2007/2008 within the FertOrtMedBio research project (Mediterranean Organic Horticulture Fertilisation) aiming to develop improved and sustainable crop management systems for organic vegetable crops.

The experiment was carried out as a completely randomised block design with three replicates. Yield and number of cauliflowers were progressively recorded in 10 sampling areas (1 x 1.35

m) per plot until the end of harvest (19 February 2008), while weeds, hairy vetch biomass, and cauliflower leaves were collected also on 19 March 2008 when the cover crop and crop residues were ploughed under. .

The yield was split in marketable (diameter > 10 cm) and unmarketable heads. Both innovative systems (ICMS and ACMS) gave a number of marketable heads three-fold that of SCMS (P<0.01). Total (marketable + unmarketable) head fresh matter in ICMS and ACMS was 2.7-fold that of SCMS (P<0.01). Significant differences were also observed in percent marketable biomass 81.3% as average of ICMS, and ACMS vs. 16.4% in SCMS (p<0.01). Total weed biomass did not differ between the three systems on 19 February 2008, while on 19 March 2008 the ACMS system showed 40% lower weed biomass (P<0.05) compared to ICMS. The weed biomass in SCMS was not tested on 19 March 2008 because at this date the soil seedbed had already been prepared for the next crop.

Results showed clear differences between the SCMS and the two innovative systems for each parameter studied. No evidence was found to indicate that hairy vetch influenced the yield of cauliflower in ACMS. However, this system showed a better weed control than that relying only on physical weed management system (ICMS) when soil was left uncultivated after cauliflower harvest.

This study was conducted within the MiPAF FertOrtMedBio Project.


Mechanical barriers for Cyperus esculentus (yellow nutsedge) control in strawberry

O. Daugovish1 and M. J. Mochizuki1 1University of California Cooperative Extension Farm Advisor and Staff Research Associate, 669

County Square Drive, Suite 100, Ventura , CA 93303, USA, Email: [email protected]

Cyperus esculentus is a most difficult weed to control in an annual plasticulture strawberry production systems in California, USA. While most opaque plastic tarps prevent germination of annual weeds, shoots of C. esculentus penetrate standard polyethylene tarps and rapidly establish in strawberry beds.

A series of RCB designed experiments with four or five replications were conducted at Oxnard, California (2006-2009) to compare emergence of C. esculentus in naturally infested planting beds covered with black polyethylene tarp alone or in combination with mechanical barriers. Mechanical barriers included: Novovita paper (recycled newspapers, gypsum) placed under plastic, paper placed between the two layers of plastic, weed barrier matt (landscaping fabric), water resistant Tyvek (DuPont) home wrap installed under plastic and high density embossed Dura Skrim 8 BBR plastic applied alone. All plots were 1.5 by 8 m.

In fall and winter the combination of paper alone under plastic completely eliminated C. esculentus germination that otherwise germinated through standard plastic at a density of 5 plants/m² per week. In spring, when the paper disintegrated due to contact with wet soil and when soil temperature increased above 16 C the weed resumed germination at a rate of 3 to16 plants/m²

per week in all treatments. However, weed barrier matt, paper layered between two plastic mulch layers and water resistant Tyvek (DuPont) all provided near 100% control of C. esculentus shoots throughout the 9 months growing season. Evaluation of embossed Dura Skrim plastic is currently in progress and shows 100% C. esculentus control.

In 2007-2008 seasons, in addition to C. esculentus control, plots with mechanical barriers had 45-67% less wind dispersed weeds in planting holes, thus potentially reducing costs for hand-weeding. These studies suggest that persistent mechanical barriers prevent C. esculentus germination and are especially valuable in non-fumigated and organic production in the absence of other control tools.


Weed patchiness: implications for physical and cultural control

C. Fernández-Quintanilla, J. Dorado, D. Andújar and D. Ruiz

Instituto de Ciencias Agrarias, CSIC, Madrid, Spain

Weed distribution within fields is generally heterogeneous. This heterogeneity may have important implications on weed management. However, up to date, the concept of site-specific weed management has been mainly applied to chemical control. In this paper we will explore some of the implications of weed patchiness for physical and cultural control.

Surveys conducted in 31 barley fields have shown that Avena sterilis (winter wild oats) infestations tended to be concentrated in certain topographical positions (flat lowland, concave zones, northern slopes) (Ruiz et al., 2006b). These zones should receive individual attention by the farmers when planning crop production practices, avoiding the use of cereal monocultures, delaying the sowing date or increasing sowing densities.

In the case of Sorghum halepense (johnsongrass), surveys conducted in 47 commercial maize fields have shown that infestation is concentrated in the borders, decreasing progressively as we go inwards the fields. This fact suggests that this weed species infests the fields starting from the field margins and irrigation channels. Farmers should conduct intensive tillage operations of field borders during the winter in order to destroy the rhizomes with frost. Furthermore, they should try to minimize movement of rhizomes to the centre of the field via tillage operations.

Patch size is an important factor to take into account when selecting and using weed control tools. Our A. sterilis survey has shown that 85% of area covered by this weed was formed by rather large patches (> 2000 m2) (Ruiz et al., 2006a). This implies that most of the infested area could be managed with control tools with a relatively coarse spatial resolution (i.e., cultivation implements). On the contrary, controlling small patches (< 200 m2) would require the use of implements that work at a micro scale. Various non-chemical treatments (flaming, steaming, UV-radiation, electrocution, high speed hoe) may be applicable at this small scale (Rask & Kristoffersen, 2007).

Patch shape is also important. In the case of A. sterilis, we have shown that patches are relatively regular (~ circular). This fact introduces complexity when using coarse resolution control tools. In contrast, patches of S. halepense tend to be elongated, following crop rows (i.e., the direction of tillage). This shape may simplify site-specific mechanical weed control.

References Rask AM and Kristoffersen P (2007) A review of non-chemical weed control on hard surfaces. Weed Res 47, 370-380. Ruiz D, Escribano C and Fernandez-Quintanilla C (2006a) Assessing the opportunity for site-specific management of

Avena sterilis in winter barley fields in Spain. Weed Res 46, 379-387. Ruiz D, Escribano C and Fernandez-Quintanilla C (2006b) Identifying associations among sterile oat (Avena sterilis)

infestation level, landscape characteristics and crop yields. Weed Sci 54, 1113-1121


An integrated physical approach to control purple nutsedge (Cyperus rotundus)

Hershenhorn, J1., Weissblum, A2., Dor, E1., Lande, T1., Achdary G1. and Smirnov, E1. 1Dept. of Phytopathology and Weed Research, Newe Ya’ar Research Center, Agricultural Research

Organization; 2Dept, of Growing, Production and Environmental Engineering, Agricultural Research Organization

Purple nutsedge (Cyperus rotundus L.) is considered to be one of the world's worst weeds, especially in the tropical and subtropical regions. In Israel, where purple nutsedge is also regarded as an important weed, seeds only rarely germinate and reproduction is vegetative by underground tubers. The rapid growth, prolific propagation through a complex underground system of rhizomes and tubers and its narrow leaves with thick cuticles makes it extremely difficult to control by mechanical or chemical means. The weed affects mostly summer irrigated vegetables causing high yield losses.

In order to control the weed we integrated mechanical removal of the tubers and plastic mulching. In pot experiment purple nutsedge tubers were buried in depths of 0, 10, 20 and 30 cm and the pots were covered with 200 µm thick black nylon. The tuber sprouting was monitored for several months.

In a field experiment, a mechanical removal of the tubers was preformed followed by 200 µm thick black nylon mulching. Results will be discussed.


Combining enhanced competition and cultivation for improved weed control in organic cereals

L.N. Kolb, T. Molloy and E.R. Gallandt Sustainable Agriculture Program, University of Maine, Orono, ME, USA

Email:[email protected]

Despite advances in individual methods of weed control in organic small grain cropping systems, weeds continue to reduce yield and quality. Our research objective is to combine physical and cultural weed management practices to provide reliable and economical weed control. To do this, field trials were established in 2007 and 2008 comparing standard cereal production practices to strategies that increased crop interference through narrow row spacing or increased spatial uniformity, in combination with elevated seeding rates and harrowing. We measured the efficacy of harrowing on both final weed density and yield in these more competitive systems. In addition, a wide row planting, subject to both harrowing and inter-row hoeing was added to compare the effects of enhanced physical weed control to systems that rely extensively on enhanced crop: weed competition.

Multiple control tactics, for example the narrow row planting with increased seeding rate or wide row sowing with elevated seeding rate, showed significant improvement in barley yield and reduced weed biomass over standard practice and systems that only included one control method. Ongoing efforts to quantify weed seed return in each of the tested systems with allow better assessments of the long-term benefits of adopting increased integration as a weed management method in organic cereal production.

We speculate that choice of both cereal species and cultivar could interact with these systems depending on the severity of weed pressure and have initiated field trials comparing alternative physical and cultural management practices over several barley cultivars.


Using ecological processes to manage cropping systems for weed suppression and other services in the U.S. corn belt

Matt Liebman1, Paula R. Westerman2, Andrew H. Heggenstaller1, Carol L. Williams1, Brent J. Danielson1, Mark D. Tomer3, and Michelle M. Wander4

1 Iowa State University, Ames, Iowa 50011, USA; e-mail: [email protected] 2 Universitat de Lleida, 25198 Lleida, Spain

3 USDA-ARS-National Soil Tilth Laboratory, Ames, IA 50011, USA 4 University of Illinois, Urbana, IL 61801, USA

Conventional cropping systems in the central U.S. have low levels of biological diversity, rely heavily on synthetic fertilizers and herbicides, and commonly emit environmental contaminants. Ecological theory suggests that diversified cropping systems integrated with livestock should foster reduced reliance on agrichemical and fossil fuel inputs and should lower production costs and pollution. Since 2002, we have conducted a 9-hectare experiment in Boone, Iowa, within the core area of the U.S. Corn Belt, to test the hypothesis that a four-crop system (maize/soybean/oat/alfalfa) receiving only small amounts of agrichemicals can match or exceed a conventionally managed two-crop system (maize/soybean) in weed suppression, yield, and profitability (Liebman et al., 2008).

Data collected during 2003-2007 indicate that weed biomass in maize and soybean was low (<4.2 g m-2) in both the conventional and LEI systems, despite the use of 84% less herbicide in the LEI system. Experimentally supplemented seedbanks of the annual weeds Abutilon theophrasti and Setaria faberi declined in both the conventional and LEI systems. Substantial losses of A. theophrasti and S. faberi seeds to predators were detected during both warm (April-November) and cold (November-April) times of year. Trap data indicated that mice, field crickets, and carabid beetles were the most important weed seed consumers present. Temporal patterns of seed predation differed among crops in a manner that suggested that planting strips of crops with asynchronous patterns of canopy development should enhance the impacts of weed seed predators. Modeling analyses showed that predation of A. theophrasti seeds could contribute strongly to regulation of weed density in the LEI system, but was unnecessary in the conventional system.

Fossil fuel energy use was 53% lower and nitrate leaching was 26% lower in the LEI system than the conventional system, whereas crop yields were higher in the LEI system than the conventional system (maize: 12.9 vs. 12.3 Mg ha-1; soybean: 3.8 vs. 3.4 Mg ha-1). Concentrations of particulate organic matter carbon in the surface 30 cm of soil were 27% greater in the LEI system than the conventional system. Labor requirements were greater in the LEI system than the conventional system (3.41 vs. 1.82 hr ha-1 yr-1), but average returns to land and management were also greater in the LEI system ($540 vs. $504 ha-1 yr-1). With subsidies, differences between the systems in net returns were smaller, but rank order of the systems was maintained. Data from this experiment indicate that cropping systems can be redesigned for productivity, profitability, and improved environmental quality, and that ecological processes such as weed seed predation can reduce requirements for agrichemicals.

References Liebman, M., L.R. Gibson, D.N. Sundberg, A.H. Heggenstaller, P.R. Westerman, C.A. Chase, R.G. Hartzler, F.D.

Menalled, A.S. Davis, and P.M. Dixon. 2008. Agronomic and economic performance characteristics of conventional and low-external-input cropping systems in the central Corn Belt. Agronomy Journal 100: 600-610.


Integrated winter/summer cover crops into weed management program in organically grown vegetables

H. Mennan1, M. Ngouajio2, D. Isık3 and E. Kaya1

1Ondokuz Mayıs University, Agriculture Faculty, Department of Plant Protection, 55139 Samsun, Turkey Email:[email protected]

2Department of Horticulture, Plant and Soil Science Building, Michigan State University East Lansing, MI 48824, USA

3Black Sea Agricultural Research Institute, Samsun, Turkey

Weed control is a major concern for organic farmers around the world and non-chemical weed control methods are now the subject of many studies. Organic vegetable production has high potential for niche market in Turkey. Despite recent research effects, weed management is still a major constraint in organic production. Many farmers are willing to transition to organic vegetable production because of premium prices and the fact that it can be processed into various products. But, they also face the lack of effective weed management strategies. The use of cover crops could be alternative method for organic vegetable growers due to their ability control weeds, improve soil fertility and structure, conserve soil moisture and reduce erosion.

Field studies were conducted in tomato (Solanum lycopersicum L.), pepper (Capsicum annuum L.), kale (Brassica oleracea var. acephala) and lettuce (Lactuca sativa L.) from 2004 to 2006 at the Black Sea Agricultural Research Institute experimental field to determine the suppressive effects of winter/summer cover crops on weeds. The treatments consisted of winter and summer cover crops. Ryegrass (Lolium multiflorum L.), oat (Avena sativa L.), rye (Secale cereale L.), wheat (Triticum aestivum L.), gelemen clover (Trifolium meneghinianum Clem.), egyptian clover (Trifolium alexsandrinum L.), common vetch (Vicia sativa L.) and hairy vetch (Vicia villosa Roth.) were used as winter cover crops. Grain sorghum [Sorghum bicolor (L.) Moench.], sudangrass [Sorghum vulgare Pers. var. sudanense (Piper) Hitchc.], hairy vetch (Vicia villosa Roth.), grain amaranth (Amaranthus cruentus L.) and pea (Pisum sativum L.) were used for summer cover crops. Treatments were arranged in a randomized complete block design with four replications. To determine the weed suppressive effects of the cover crops weed density, and total weed dry biomass were assessed at 14, 28, and 56 days after desiccation (DAD) from all plots using a 50 x 50 cm quadrat placed randomly in each plot.

At the time of winter cover crop termination and incorporation, about 50% of the cover crops were at flowering stage. Cover crop biomass was established well and ryegrass produced the highest (3567 kg ha-1) biomass fallowed by common vetch, hairy vetch, oat, rye, wheat, egyptian clover and gelemen clover. Weed dry biomass production just before termination of the cover crops was lowest in rye, ryegrass, common vetch, and hairy vetch plots. After termination the cover crops, hairy vetch, ryegrass, oat and common vetch residue showed the greatest potential to reduce total weed dry biomass in both years at 14 DAI. This reduction was very distinctive at 28 and 56 DAI than 14 DAI. Ryegrass was one the most effective cover crops at suppressing weeds throughout the season. It caused more than 70% reduction in emergence of P. oleracea, S. arvensis, V. sativa, L. aphaca, S. nigrum, F. officinalis, and T.pratense at 28 DAI. But it failed to inhibit seedling emergence of C. arvensis, R. crispus, A. theophrasti, P. convolvulus, C. nutans, C. bursa pastoris, M. chamomilla, E. crus-galli, A. myosuroides, L. perenne, S. glauca, B. sterilis and S. verticillata. At the exception of egyptian clover, all cover crops increased pepper and tomato yield.

After incorporation of summer cover crops, hairy vetch and sorghum treatments showed fewer weed species, and lower weed density and total weed dry biomass than the other cover crops. Cover crop residue suppressed emergence of Portulaca oleracea, Chenopodium album, Amaranthus retroflexus, Heliotropium europaeum, Thlaspi arvense, Sonchus oleraceus, Solanum nigrum, Capsella bursa-pastoris, Sinapis arvensis, Euphorbia helioscopia, Veronica persica, Mercurialis


annua, and Datura stramonium up to 56 DAI. Total kale and lettuce yield in hairy vetch treatments was more than double that of the control, and was significantly higher than yield from the other cover crop treatments.

These results indicate that ryegrass and hairy vetch as a winter cover crops, grain sorghum and sudangrass as an summer cover crops have ability to suppress weeds in early season of organic tomato, pepper, kale and lettuce production.

References Fisk JW, Hesterman OB, Shrestha A, Kells JJ, Harwood RR, Squire JM and Sheaffer CC (2001). Weed suppression by

annual legume cover crops in no-tillage corn. Agron. J. (93) 319-325. Mennan H, Ngouajio M, Isık D and Kaya E (2006). Effects of alternative management systems on weed populations in

hazelnut (Corylus avellana L.). Crop Prot. (25) 835- 841.

Moore MJ, Gillespie TJ and Swanton CV (1994). Effect of cover crop mulches on weed emergence, weed biomass, and soybean (Glycine max) development. Weed Technol. (8) 512-518.

Morales-Payan JP, Santos BM, Stall WM and Bewick TA (1997). Effects of purple nutsedge (Cyperus rotundus) on tomato (Lycopersicon esculentum) and bell pepper (Capsicum annuum) vegetative growth and fruit Yield. Weed Technol. (11) 672-676.

Ngouajio M and Mennan H (2005). Weed populations and pickling cucumber (Cucumis sativus) yield under summer and winter cover crop systems. Crop Prot. (24) 521-526.

Reddy KN (2001). Effects of cereal and legume cover crop residues on weeds, yield, and net return in soybean (Glycine max). Weed Technol. (15) 660-668.

Teasdale JR (1996). Contribution of cover crops to weed management in sustainable agricultural systems. J. Prod. Agric. (9) 475-479.


Trials of a crimper-roller for killing cover crops for organic and non-herbicide, no-till cropping

C.N. Merfield

Teagasc (Irish Agriculture and Food Development Authority), Johnstown Castle, Co. Wexford, Republic of Ireland. Email: [email protected]

Abstract Dependence on tillage for weed management in organic systems potentially reduces soil quality

/ health, which is contrary to the founding principle of organic agriculture: healthy soils. One weed management technique that can eliminate the need for tillage for cash crop establishment is ‘crimper-rolling’. This is where a cover crop is killed and flattened at anthesis using a ‘crimper-roller’ and the cash crop is planted into the resulting mulch. The technique was originally developed in South America and subsequently independently discovered in North America where there is currently considerable research and farmer interest. However, the agricultural and climatic conditions of the Americas differ considerably from other areas that operate generally comparable agronomic systems, such as Europe and New Zealand where the technique is less well known. Therefore, a trial was conducted in the Canterbury region of New Zealand on four cover crops: A mix of oat (Avena sativa) and pea (Pisum sativum), rye (Secale cereale), common vetch (Vicia sativa) and field / tick bean (Vicia faba) to study the potential of the using the crimper-rolling technique under local conditions. Analysis was made of how easy the cover crops were to kill by the amount they regrew after treatment, their weed suppressing ability and the ratios of volunteer grass spp. to white clover (Trifolium repens), which was considered to be an indicator of available soil nitrogen levels. The oat pea mixture showed poor weed suppression, while weed suppression and N availability were excellent under common vetch but the crop was unaffected by crimping. Rye showed good weed suppression and was completely killed by crimping but tied up soil N. Beans showed moderate weed suppression and N availability and died very quickly after crimping. The results indicate that cereal-legume mixtures e.g., rye and beans, may be preferable over cover crop monocultures in order to reduce soil nitrogen ‘robbery’ while maximising weed suppression. Nonetheless, uptake of the technique in the Canterbury region may be hampered because anthesis occurred after the standard planting dates of many common crops. Despite this lack of initial success the technique appears to have considerable potential and significant further research is required.

Introduction Maximising soil health is a founding principle of organic agriculture. However, the decision to

eschew synthetic chemical herbicides has left organics dependent on tillage for many weed control operations. There is substantial evidence that tillage damages soil health, while no-till agricultural systems are in general able to promote and maintain high soil health (Baker & Saxton, 2007). However, no-till systems rely on synthetic herbicides for weed control which are not permitted in organic production. This has resulted in a conflict within organic agriculture between tillage based weed management and the need to reduce tillage to improve soil health. The introduction of no-till techniques that are consistent with organic principles could help address this dilemma. Such techniques may also be increasingly valuable in non-organic systems because of the decreasing number of permitted and/ or effective herbicides due to the combined effects of legislative review of pesticides and increasing numbers of weeds evolving herbicide resistance.

In the 1980s, a low-tech machine was developed in South America to kill green manure / cover crops when they have reached full anthesis, without soil disturbance (Ribeiro, 2001). It consists of a smooth roller with a series of thin metal bars attached to the surface across the full width of the


roller. Originally called ‘knife rollers’, they are also known as roller-crimpers in N. America, crusher-rollers and crimper-rollers. The effect of the machine is to flatten the cover crop and crimp its stems, i.e., the stem is not severed, but bent / crimped. This way, the cover crop is converted into an in-situ biological surface mulch, anchored in place by the plants’ roots, that can suppress weeds and protect the soil. Crops planted into the mulch benefit from multiple effects, including weed suppression, moisture retention, slow release nutrient supply and improved soil conditions. In addition, the cover crops compete with weeds while they grow, thus reducing weed prevalence through interactions such as allelopathy and agronomic activities associated with cover crop establishment and management, further aiding weed suppression. As the technique was unknown in New Zealand (NZ), an initial demonstration trial was conducted to study its potential under NZ climatic and seasonal conditions.

Materials and methods The design was a two factorial randomised complete block with four replicates. First factor was

cover crop type with four treatments: 1. oats 80 kg ha-1 and peas 180 kg ha-1; 2. rye 120 kg ha-1; 3. vetch 25 kg ha-1; 4. beans 300 kg ha-1.

The second factor was crimping date with two treatment dates: 11 December 2006 (first) and 9 January 2007 (second). The cover crops were drilled on 6 May 2006 (autumn), into 1.70 m × 75 m beds at the Biological Husbandry Unit, Lincoln University, Canterbury, NZ (43°39'3.04"S 172°27'24.73"E). The cover crop were at anthesis at the first date and had finished flowering at the second date. A herringbone crimper-roller was used with a roller diameter of 48 cm, width 2 m, with 10 blades, 5 mm thick and 13 cm deep with the gap between blades at their tips of 20 cm (The Rodale Institute, 2007). It weighed 550 kg, contained approx. 200 kg water ballast and had additional metal ballast of 380 kg for the second crimping, as it was thought that this may improve cover crop mortality. Visual assessments including photographs of the cover crops attributes such as height and morphology and the effect of the crimper-roller on the cover crop were made over a range of dates.

On 3 February 2007, 54 d after the first and 25 d after the second crimp all weeds, i.e., all vegetation growing through the cover crop but excluding the cover crop itself, were collected from four random 0.25 m2 quadrates per plot, hot air dried and weighed.

Clover takes up nitrogen (N) from the soilin preference to fixing it from the atmosphere due to the higher metabolic cost of N fixation. In soils with high plant available N, grasses tend to outcompete clover as they are better at N assimilation. When available soil N is low, clover has the competitive edge over grass due to fixation. Therefore, the ratio of grass to clover in a mixed pasture gives an indication of available soil nitrogen levels. Significant amounts of mostly ryegrass (Lolium perenne) and white clover (Trifolium repens) ‘volunteers’ germinated under the cover crops soon after planting but remaining small while the cover crops were alive. After crimping however, the grass and clover grew up through the mulch. The proportion of grass to clover growing through the cover crop was used as a surrogate measurement for available soil N under the cover crop. It was visually estimated in four random 0.25 m2 quadrates per plot with the data from the first and second crimpings combined. Grass and clover were also classed as weeds for the above weed dry weight measurements.

The dry weight of the weeds and the proportion of grass to clover were analysed by ANOVA on untransformed data.


Results The proportion of grass and clover and the dry weight of weeds in the respective cover crops

are shown in Table 1. Rye and beans were completely killed at both treatment dates and grew the tallest. Rye showed good and beans moderate weed suppression. Vetch had very dense foliage, which formed a spreading mat that very effectively suppressed weeds. However, the crop was impossible to kill by crimping, and continued to grow strongly after treatment. There was a small amount of oat re-growth in the oat and pea mixture, which had a short open growth habit that failed to suppress weeds.

Clover was dominant under rye while beans and the pea-oat mix had close to a 50:50 ratio of glass to clover, and vetch had 75% grass, which was almost the opposite of rye. Table 1 The proportion of grass and clover, and the dry weight of weeds growing through four crimper-rolled cover crops at two crimping dates.

Clover Grass Weed dry weightfirst crimping

Weed dry weight second crimping

Oats and peas 55% 45% 252 g m-2 105 g m-2 Rye 93% 7% 113 g m-2 42 g m-2 Vetch 25% 75% 58 g m-2 13 g m-2 Beans 59% 41% 236 g m-2 70 g m-2 P value < 0.001 < 0.001 LSD0.05 24.3 35.3

Visually, there was no noticeable benefit of using the extra weight on the crimper-roller at the second crimping, either on the crimping effect itself or crop re-growth.

Discussion The poor weed suppression of oats and peas and the re-growth of oats mean these spp. appear

unsuitable for crimping. However, pea cultivars with more biomass and particularly leafed types which are considered to be more competitive against weeds may be suitable. The strong re-growth of vetch also rules it out, however, other vetch species, e.g. hairy vetch (Vicia villosa), and black oats, (Avena strigosa), are killed by crimping, show good weed suppression and are successfully used by farmers (Anon., 2007; Baker et al., 2007). Rye is considered to be allelopathic (Dhima et al., 2006), which may be partly responsible for its better weed suppression than beans. As both rye and beans were effectively killed by crimping, they are considered suitable spp. to use for crimper-rolling. The lower weed suppression of peas, beans and oats may also be a function of their populations; some proponents of crimper-rolling recommend sowing rates two or even three times higher than for cash crops to maximise weed suppression and biomass production (Jeff Moyer, 2009, pers. comm.). These interacting factors indicate that it is not possible to make a final decision on any of the cover crops based on this single experiment and that considerable further research is required.

While the grass and clover that had established itself at the time of cover crop sowing grew little while the cover crop were standing, they rapidly grew through the mulch after crimping. This is considered a clear example of the need for effective weed management within the cover crops, with higher sowing rates probably making an important contribution. With the caveat that the effect of soil N on the grass: clover ratio is an assumption in the context of this experiment, i.e., no direct relationship was established between soil N and grass: clover ratio, the relative proportions of grass and clover are still illuminating. If a cash crop planted through the mulch is to grow vigorously there must be enough soil N available. A high clover content indicates that N is low and that therefore crop growth could be restricted. The dominance of clover in the rye indicates that available soil N was scarce, which is likely due to rye having taken up a large proportion of soil N


during growth, which had yet to be released by decomposition of the straw. The leguminous spp. on the other hand had more grass, which is expected for N fixing legumes. A common recommendation for cover crops is a 50:50 mix of legume and cereal, because the cereal has a high N demand forcing the legume to fix N thereby maximising fixation while minimising leaching. Assuming the grass: clover ratio is an accurate reflection of soil N availability, the results presented here support the idea of such mixtures. A combination of rye and beans would appear to be good contenders due to the large biomass, good weed suppression and rapid death on crimping, although the optimum ratio of rye to beans for crimping may not necessarily be 50:50. This needs to be studied in more detail. Other vetch spp. that are suitable for crimping should also be tested due to their good weed suppression and high levels of N fixation (see below).

It was initially planned to sow a cash crop following crimping, in order to assess the follow on effects, such as weed suppression and soil N availability. This was prevented, not only by the late date of anthesis, but also by the impenetrability of the dense mats of the crimped cover crop vegetation by any available drill high residue seed drills. North American researchers using this technique find that only purpose designed no-till drills have the capability of successfully drilling through such thick residues (John Luna and Jeff Moyer, 2009, pers. comm.). An alternative approach may be to drill into the standing cover crop, which high residue drills may be able to penetrate, and then crimp after drilling. However, this approach may be impossible with vetch spp. due to the dense mat of interwoven stems and branches that they form.

A major problem for crimper-rolling in Canterbury was the date of anthesis, which occurred in late November (mid summer). This is far later than the September to October planting date of most summer annual crops, which means the technique would be of no value for any but the latest planted crops. However, for late-planted crops such as squash (Cucurbita mixta) the technique could be especially advantageous, as the mulch could keep the pepo (fruit) off the soil surface thus improving skin quality. It may also be practical for crops that are manually transplanted, e.g., vegetables such as cabbages (Brassica spp.). The technique may be viable for more crops in climates that experience longer summers and shorter winters. Alternatively if cover crop species and/or cultivars can be identified that flower sufficiently early, especially if flowering date can be manipulated through cultural means, e.g., earlier planting dates, then it may be applicable to a wider range of crops within New Zealand.

While crimper-rolling is proving to be a valuable technique globally, to date, no research is known that has studied its effect on plant physiology (Miguel Altieri, 2008; John Luna and Jeff Moyer 2009, pers. comm.). It is considered important that to improve our understanding at a more fundamental level, i.e., understand causal mechanisms, and establish why crimping kills plants while rolling or mowing does not. Such research should be able to assist in designing better crimper-rollers with optimum blade spacings, weight and effectiveness, determine why some, often closely related, plant species react differently to crimping, all of which could lead to improvements in the technique.

Conclusions Crimper-rolling has the potential to allow short-term no-till organic cropping in NZ with the

potential benefits of reduced tillage for weed management and soil preparation. However, the lateness of anthesis in the Canterbury region for the cover crops tested means that other species and cultivars need to be investigated. Mixtures of cereals and legumes are likely to offer the best combination of weed suppression and soil N availability for the following crop and cover crop plant population may also be a critical factor. Clearly, a substantial amount of further research is required including fundamental work on the effects of crimping on plant physiology, determining the best mixtures of cover crops that provide optimum weed suppression, high residual soil nitrogen, and good cover crop mortality; as well as establishing specific farmer guidelines for individual cover crop × cash crop combinations. It is hoped that such research would make crimper-rolling a viable


new technique for both organic and non-organic farmers, in NZ and other locations, to improve environmental sustainability and profitability of farm systems.

Acknowledgments The Sustainable Farming Fund for funding the research. Mr Ivan Barnett for looking after and

crimping the crops, organising construction of the crimper-roller and help with data collection. The Biological Husbandry Unit for hosting the trials. Mr Holger Kahl for being project manager. Dr Christine Stark for her detailed review of this paper.

References The Rodale Institute (2007): The no-till + page. http://www.newfarm.org/depts/notill/, (accessed 2007-09-21) Baker C J, Ribeiro, F and Saxton K E (2007): Residue Handling. In Baker C. J., Saxton K. E. (eds.): No-tillage Seeding

in Conservation Agriculture, 2nd Edition. Food and Agriculture Organization of the United Nations. Wallingford, UK. p. 134-158.

Baker C J and Saxton K E (2007): No-tillage Seeding in Conservation Agriculture, 2nd Edition. Food and Agriculture Organization of the United Nations, Wallingford, UK, 352 p

Dhima K V, Vasilakoglou I B, Eleftherohorinos I G and Lithourgidis A S (2006): Allelopathic potential of winter cereal cover crop mulches on grass weed suppression and sugarbeet development. Crop Science 46:4:1682-1691.

Ribeiro M F (2001): No-tillage equipment for small farms in Brazil. In García-Torres L, Benites J and Martínez-Vilela A (eds.): Conservation agriculture, a worldwide challenge. European Conservation Agriculture Federation, Córdoba, Spain, p. 237-243


Effects of cereal rye mulch and soybean density on weed suppression

M.R. Ryan1, D.A. Mortensen1, S.B. Mirsky2, J.R. Teasdale2, & W.S. Curran1 1The Pennsylvania State University, University Park, PA, USA Email: [email protected]

2Beltsville Agricultural Research Center, Beltsville, MD, USA

No-tillage crop management can be successfully implemented in organic cropping systems if cultural practices are employed to manage weeds otherwise controlled by tillage and herbicides. Cover crops can serve as the primary agent of weed suppression in such situations through multiple mechanisms including reduced light resource availability and allelochemical interference. Increasing crop density is another cultural practice that can compliment the use of cover crops for weed suppression. Interspecific competitive ability of soybean increases with increasing crop density, and although this can be an effective weed suppression strategy, recent research has focused on reducing soybean density because of increasing seed prices (technology fees), the high efficacy of glyphosate, and the somewhat elastic yield response to soybean density.

In order to gain a better understanding of the interaction between cereal rye residue and soybean density, a study was conducted in 2008 at the Rodale Institute in Pennsylvania, USA, and was replicated at the Beltsville Agricultural Research Center in Maryland, USA. The experiment was designed as a full factorial split-block with five levels of rye residue representing 0, 0.5, 1, 1.5, and 2 times the ambient level (~10,000 kg ha-1) and five levels of soybean density ranging from 0 to 740,000 seeds ha-1. Cereal rye residue and soybean density were measured in each experimental unit. Biomass of each weed species was quantified in late-August (peak biomass) by clipping all weeds to the soil surface, drying to less than 5% moisture and then weighing. Soybean yields were quantified using a plot combine in November.

Weed biomass decreased with increasing cereal rye residue. In addition to the effect on weeds, soybean plants were also suppressed at the high level of residue. Weed communities differed across the different cereal rye residue levels. Setaria faberi L. (giant foxtail) was a dominant weed in the no-residue plots whereas it was completely absent in the plots that received cereal rye residue. Ambrosia artemisiifolia L. (common ragweed) was a dominant weed species in the mid-range of residue, but was completely suppressed at the higher levels. Calystegia sepium L. (hedge bindweed) was the only weed species present at high residue levels. Although not as dramatic as the effect from cereal rye residue, weed biomass also decreased with increasing soybean density. Unlike cereal rye residue, there was no shift in weed communities across the soybean density gradient. Results indicate that cereal rye residue can provide adequate levels of weed suppression in organic no-till planted soybean. Soybean seeding rate can also be used to compensate for lower cereal rye levels or in situations with large weed seedbanks. Together, these tactics can overcome many challenges of weed management in systems that do not rely on cultivation or herbicides.


Comparison of several recommended cultural practices for weed management and their effects on yield of coconut in tropical coconut plantations

S.H.S.Senarathne and M.J.I.Costa Email: [email protected]

Coconut Research Institute, Lunuwila, Sri Lanka

Abstract The influence of five different cultural practices on nut yield of coconut were evaluated to

determine an economical and effective method of controlling weeds in coconut plantations in the low country, intermediate zone in Sri Lanka. Treatments imposed were cover cropping with Pueraria phaseolodies (T1), planting Gliricidia sepium (T2), tractor slashing (T3), application of glyphosate (N-(phosphonomethyl)-glycine) at 1.44 kg ai ha-1(T4) and cattle grazing at two months interval (T5). All the treatments were applied twice a year except for the cover cropping with Pueraria and cattle grazing treatments. Cattle grazing were applied at two monthly intervals. Based on the reduction of weed biomass, treatments T1 and T4 were found to be significantly effective over other treatments. Coconut yield was increased significantly (P<0.05) in glyphosate applied and cover cropping with Pueraria plots. Control of weeds with glyphosate (1.44kg ai ha-1 and cover cropping with Pueraria resulted in 16% and 8% increase in nut yield over the uncontrolled weed plots respectively. These two methods were found to be the most effective and economical methods of controlling weeds in coconut plantations. Cover cropping with Pueraria phaseolodies was effective in controlling weeds in the long term, but was not economical compared with the glyphosate application. Establishment of cover crop was helpful to conserve soil moisture and to improve the soil fertility when compared with other treatments.

Introduction Coconut is a tropical perennial plantation crop and its canopy structure requires a wide spacing

between palms, which permits abundant sunlight to the understory. As a result, the unutilized space beneath the plantation becomes invaded by a wide range of perennial and annual weed species (Liyanage & Liyanage 1989). Such weeds invariably compete with coconut for soil moisture and nutrients, affecting its growth and yield and obstructing routine estate practices (Senarathne et al. 2003). Management of the understory weed growth is, therefore, considered an essential step in maintaining a coconut plantation. In fact, the cost of weeding (20% of the total cost of production of the estate) accounts for a substantial proportion of the total recurrent expenditure for maintenance. Therefore, weeds in coconut plantations in Sri Lanka not only reduce yields due to crop weed competition but also add to production costs. Thus the yield losses due to crop weed competition have a substantial impact on productivity and the high cost of weed control operations reduce the profitability of coconut estates. Therefore, there is an acute need to introduce effective and economically viable weed control strategies for coconut plantations in Sri Lanka.

Weeds in coconut plantations are managed in different ways, under the broad categories of mechanical, chemical and cultural methods (Liyanage and Liyanage, 1989). However, complete eradication of weeds is not expected and weeds have to be managed to some extent so that they do not compete with coconut. Slashing either by hand or tractor harrowing, use of herbicides, inter-row cultivation of fast growing leguminous creepers or bushy type cover crops and grazing by ruminants are currently popular weed control methods in coconut plantations locally. But selection of a suitable herbicide to be used in chemical weed control is very important and glyphosate (N-(phosphonomethyl)-glycine) is a commonly used herbicide in coconut plantations and which controls a wide range of monocotyledons and dicotyledonous annuls biennials and perennials (Boyall, 1998).


Four-wheel tractor mounted slashers and harrows are the main mechanical methods practiced in coconut plantations. Slashing removes the aerial parts of the weeds, resulting in a depletion of the food supply to the rhizomes. Repeated slashing removes the aerial parts can serve as mulch on the ground, and help to conserve soil moisture. Removal of aerial parts also reduces water loss through transpiration. Finally, slashing performed at sufficient frequencies is capable of keeping the growth of ordinary weeds down to an acceptable level (Pethiyagoda, 1980). Weeds that do not tolerate shade and competition can be controlled by other faster growing species. In such situation cover crops can be used as fast growing species and they can suppress weed growth effectively. Suitable cover crops such as Pueraria phaseolodies, Centrosema pubescens (creepers) and Gliricidia sepium (bushes) can be introduced. At the same time the leguminous cover crops grown for weed suppression can provide additional benefits by supplying nitrogen to the coconut palm (Senarathne, 2008). In order to develop a sustainable integrated weed management strategy, a detailed understanding of weed population changing pattern is required. Therefore, the objective of this research was to evaluate the effect of different cultural practices for weed management on the nut yield and weed population changes in coconut plantation.

Materials and methods This experiment was carried out at the Melsiripura Estate, in the Low country Intermediate

Zone of North Western province of Sri Lanka from January 2000 to December 2005. The area is characterized by bi-modal pattern of rain fall with an annual mean precipitation of 1500 mm. Approximately 65% of the annual rainfall is received from September to February (Maha). There is a smaller peak of rainfall from March to May (Yala), but the rainfall is erratic. Higher ambient air and soil temperatures (about 28 C0 - 32 C0) and bright sunshine hours (about 6 – 8 hours per day) are more common especially during the dry periods from May to September.

The soil at the site is a Reddish Brown Latasolic (RBL) (USDA soil taxonomy - Rhodudalfs fine loamy, non calcareous, isohyperthermic), (FAO/UNESCO soil taxonomy - Rhodic Cutanic Luvisols) soil. Soils are very deep and well drained. Surface soil is dark brown in colour with a sandy clay loam texture. The sub surface soil is dark reddish brown to dark red in colour and texture ranges from clay loam to clay. Structure of the sub surface soil is moderately developed coarse sub angular blocky with patchy cutans on ped faces. However, during the dry season ground water table remains at 10 – 12 meters below the ground surface. Reaction of the soil is slightly acidic (pH 6.0 – 6.5) throughout the soil profile (Mapa et al, 2005). The trial was established under 55 years old selected coconut plantation with palms were regularly fertilized with 3kg per palm a year of adults palm mixture (800g urea, 600g rock phosphate and 1600g muriate of potash) with 1kg of dolomite. Major weed species in this site were Panicum maximum, Pennisetum polystachion, Imperata cylindrica, Panicum repens Cynodon dactylon, Chloris barbata, Chromoleana odorata, Mimosa pudica, Urena lobata, Croton hirtus, Allmania nodiflora, Mitracarpus villosus, Tephrosia purpurea, Vernonia cinerea, Tridax procumbens, Sida acuta, Scoparia dulisis, Stachytarpheta jamaicensis and Hyptis suaveolens.

The following treatments were used in a randomized complete block design with three replicates. Each plot had six effective coconut palms. (Coconut spacing of the square planting system is 8.2m x 8.2m). The treatments of the experiment were: T1. Establishment of cover crop (Puereria phasioloides) T2. Establishment of Gliricidia (two rows T3. Tractor slashing (once in six month) T4. Chemical weeding (Application of Glyphosate 1.44 kg a.i. per hectare) T5. Cattle grazing (once in two months) T6. Un-weeded (Control)


To control weeds, different weeding methods were applied according to the schedule. In the chemically weeded plots glyphosate (1.44 ai kg /ha) was applied twice a year at 6 monthly intervals, at the latter part of the rainy season using a knapsack sprayer in the morning. There was no rain for five to six hours after applying glyphosate. The cover crop was established to control weeds and the over grown conditions of cover crop was managed to overcome competition by twice a year. Cattle’s were allowed to graze the plots to control weeds at two monthly intervals. Tractor slashing was done at the latter part of the rain season at six monthly intervals. Gliricidia was planted between coconut rows as two row systems (spacing 2m x 1m) and plants were lopped at 6 monthly intervals. Lopping,s were mulched around the coconut palm.

The data collected were as follows Weed biomass The weed biomass was collected within 1m x 1m quadrates from four random points per plot

every two months. Weed samples were dried at 80 C0 for five days and weighed. The dry weight of both monocotyledonous and dicotyledonous weeds was measured separately every month from January 2000 to December 2005.

Soil moisture content Soil samples (four per plot) were taken after having two weeks rain free condition in February

and August months of the year. Sampling was done from random points from 30cm depth at each location, and dried at 105 C0 to a constant weight in order to estimate soil moisture content gravimetrically as follows.

Initial weight of the soil = Yg Final dry weight of the soil = Xg Soil moisture content per 100g of soil = ((Y – X)/ Y) x 100 Coconut Yield Nut yield was counted every two months interval and calculated the total nut production per

palm per year. Data analysis: The experiment was conducted using a Nested Design with three replications.

Data were analysed by a General Linear Model and Adjusted Means were separated by Least Significant Means test at the 0.05 significance level. The statistical program was the Statistical Analysis System (SAS, 1999).

Results and discussion Effect of different cultural practices on weed biomass

Lowest weed biomass was recorded in glyphosate (N-phosphonomethyl-glycine) (1.44 kg ai/ha) applied plots as well as in plots with cover crop Pueraria (Figure 1). Initially Pueraria phaseoloides took several months to establish a good cover. The weed biomass was very high at the initial stages in cover cropped plots which declined gradually later with time. However with time, cover management was essential to avoid possible competition between coconut palms and cover crops which suppressed the growth of weeds.


0

50

100

150

200

250

300

350

2000 2001 2002 2003 2004 2005

Year

Wee

d b

iom

ass

(g/m

2)

T1- Cover cropping (Pueraria) T2- Planting Gliricidia T3- Slashing

T4-Chemical weeding T5-Cattle grazing T6- Unweeded

Figure 1. Weed biomass (g/m2) in five different cultural practices between 2000 and 2005.

(Vertical bars indicates ± SE of the means) Initially, the slashing treatment suppressed weed growth, but thereafter faster regrowth was

observed in monocot weeds than in dicotyledonous weeds. Generally slashing damaged the aerial parts of the weeds but with no damage to the root system or under ground plant parts such as stolen and rhizomes of the monocotyledonous weed species. Thus during the favorable weather conditions, underground plant parts produced new shoots or new flushes. For, example, the monocotyledonous weeds Imperata cylindrica, Panicum maximum and Cynodon dactylon and several dicotyledonous weeds Lantana camara and Chromolena odorata produced a new flush within a few weeks of slashing. Olaoye (1977) found that slashing of Chromolena odorata caused rapid regeneration. Therefore, repeated slashing and burning resulted in effective control of the weeds. Planting gliricidia between coconut rows and cattle grazing methods were less effective methods to control weeds in coconut plantations. However, use of cattle to control understory weeds in mature coconut plantations is more economical than manual weeding (Osborne, 1972) Effect of different cultural practices measures on coconut yield

Application of glyphosate (chemical weeding) and cover cropping (Pueraria) practices significantly increased (P<0.05) nut yield over other treatments. This increase in nut yield commenced from the third year after the first application of all treatments and the trend continued. Cumulative yield of five years showed 8% and 16% yield increase in cover crop and glyphosaate applied plots over unweeded plots. This could be explained in terms of reduced competition for soil nutrients and soil moisture due to lower weed density in glyphosate applied plots. Creeping legumes like Pueraria have the additional potential of bringing in substantial quantities of atmospheric nitrogen and would helpful to increase the soil fertility. Because they have a big potential to fix atmospheric N (Lehmann et al., 1999). Control of weeds with slashing, cattle grazing and planting gliricidia did not produce any yield increase in coconut. This might be a result of the presence of competition by weeds for nutrients and water. It is interesting to note that there was no significant difference between the slashing and cattle grazing (Table, 1). This indicates that competition was more for soil moisture rather than soil nutrients and is further supported by the fact that in areas with inadequate and poor distribution of rainfall, control of weeds with slashing and cattle grazing did not produce any yield in coconut.


Table 1. Effect of application of different weed control methods on nut yield (Nuts palm-1year-1) of coconut in Melsiripura Estate, Sri Lanka.

Nuts palm-1year-1 Treatments 2001 2002 2003 2004 2005

Cumulative average yield

T1.Puereria phasioloides cover T2. Planting Gliricidia T3. Tractor slashing T4. Chemical weeding T5. Cattle grazing T6. Unweeded -control Significance LSD (P=0.05)

45 59 54 44 58 46 ns

52 49 50 61 46 56 ns

61 56 51 78 53 59 ns

61 53 50 69 47 38 * 12

74 62 62 85 62 58 * 10

59 (08) 56 (05) 53 (02) 67 (16) 53 (02)

-

Values in parentheses are the percentage increase of nut yield over the control Significantly different at P=0.05 Ns: non significant

However, cover cropping with Pueraria phaseoloides was the most effective long term method for weed control, as indicated by the lowest weed biomass in the experiment. Although cover crops control weeds, they are also expected to compete for soil nutrients and water. In this situation, soil water might be the most critical factor. To overcome this situation cover crops should be managed. If don’t mange the cover, it produce the desired out come of weed control but failed to show benefits in coconut yield improvement. Cost/benefit analysis of different weed control methods under coconut

The cost of different weed control practices are outlined in Table 2. Benefits were calculated as the average incremental yield per year over the unweeded control. The highest nut yield was achieved with application of glyphosate and establishment of cover crop (Pueraria). The highest return to investment (B/C ratio of 5.22 and 2.69) were given by the application of glyphosate and cover crop establishment methods. Therefore, the most cost effective methods of controlling weeds in the present study were application of glyphosate at the rate of 1.44 ai kg ha-1 and cover cropping with Pueraria. Table 2. Cost and benefit analysis of different weed control practices

Treatment

Cost (SL Rs/ha per annum)

Average annual incremental yield ha-1

Incremental benefits (SL Rs), (SL Rs 20/nut)

Non discounted B/C ratio

T1. Puereria cover T2. Planting Gliricidia T3. Tractor slashing T4. Chemical weeding T5. Cattle grazing

9500 11300 10000 9800 12000

1280 800 320 2560 320

25600 16000 6400 51200 6400

2.69 1.41 0.64 5.22 0.54

Average price of commercial product of glyphosate SL Rs 850/L Average labour wage SL Rs 400 day-1 US 1$= Sri Lanken (SL) Rs. 110

The other methods like slashing, cattle grazing and planting gliricidia were not highly economically viable as indicated by the B/C ratios of 0.64, 0.54 and 1.41 respectively.


Effect of different cultural practices measures on soil moisture Soil moisture content was significantly high in cover cropping and glyphosate applied

treatments plots compared with other weeding treatments at every sampling time (Table, 3). Cover crops produce large amount of biomass and finally make a good litter on the soil surface. It helps to conserve soil moisture and to increase the soil fertility. The increased nutrient recycling by using leguminous cover crops may also improve fertilizer use efficiency as shown in comparison to natural grass cover for an oil palm plantation (Broughton, 1977). Table 3. Soil moisture content (%) as affected by different cultural practices in

controlling weeds at Melsiripura Estate

Soil moisture content (%) at 1.0 ft (depth) Treatment

2004 February

2004 August 2005 February

2005 August

T1 - Unweeded 5.71 4.25 4.21 3.51 T2 - Cover crop (Pueraria) 9.38 7.31 8.96 8.52 T3 - Gliricidia 6.20 4.81 4.32 4.78 T4 - Slashing and mulching 4.58 4.37 3.85 4.38 T5 - Chemical weeding 7.65 8.20 7.66 7.22 T6 - Cattle grazing 5.22 4.50 4.62 4.19 Significance * * * * LSD (P=0.05) 3.02 3.88 3.52 1.8

Conclusion The conclusion from the present study is that the application of glyphosate and cover cropping

with Pueraria produced the best cost-effective cultural practice for the control of weeds in coconut plantations. However, an integrated approach, application of glyphosare followed by establishment of leguminous creeping cover crops is very effective in controlling grass weeds biomass such as Imperata cylindrica, Panicum maximum and Pennisetum polystachion.

References Boyall LA (1998). The control of perennial weeds. In Recent Advance in Weed Research (ed, by Fletcher, W.W.) The

Gresham Press, Surry, 141-166. Liyanage LVK and de Liyanage MS (1989). Weed control under-story weed management in coconut lands. CORD, 1:

48-56. Mapa RB Dasanayake AR and Nayakekorale HB (2005). Soils of the intermediate zone of Sri Lanka, Special

Publication, Soil Science Society of Sri Lanka, 70-95. Olaoye SOA (1977). The effect of slashing on the performance of Eupatorium odoratum in Nigeria. Proceedings of 7th

Nigerian Weed Science Confernces (Abuja, Nigeria, 5-8 March 1977). S.M. Publishers, Abuja, 23-28. Osborne HG (1972). Cattle production and management under coconut (ed: E. Hugh) South Pacific Commission, New

Caledonia, 141-144. Pethiyagoda U (1980). Weed control, In Hand Book on Coconut Cultivation, 68-70. [SAS] Statistical Analysis Systems (1998). SAS 1, STAT Users Guide, Release, 7.00 Cary, NC: Statistical Analysis

Systems Institute, 1028. Senarathne SHS Samarajeewa AD and Pererea KCP (2003). Comparison of different weed management systems and

their effects on yield of coconut plantations in Sri Lanka. Weed Biology and Management. (3), 158-162 Senarathne SHS (2008). Weed populations and seed bank dynamics in coconut plantations of Sri Lanka. PhD, thesis,

Post Graduate Institute of Agriculture; University of Peradeniya, 68.


The relative importance of cultural weed control methods: A survey of results from western Canada

1Steven J. Shirtliffe, 2Eric N. Johnson, 1Dilshan Benaragama 1, Yvonne E. Lawley, and 1Julia M. Baird.

1Department of Plant Sciences, University of Saskatchewan, Saskatoon, SK, Canada S7N 5A8. 2Agriculture and Agri-Food Canada, Scott Research Farm, Box 10, Scott, SK, Canada, S0K 4A0.

There are several generic recommendations given to farmers to increase the competition of a given crop with weeds. However, little is known of the relative importance of these techniques or the interaction of these techniques when used in combination. The objective of this presentation is to examine the relative effect of cultural weed control methods on crop yield loss and weed biomass in western Canada. Crop genotypes differ in their ability to both tolerate and suppress weeds. However in many cases crop genotypes are not evaluated for competition and therefore this information is not know. Increasing the crop population density by increasing the seeding rate usually has a large effect of crop/weed competition. Furthermore, the competition benefits of other cultural weed control techniques are usually additive when used with higher crop densities. Increasing the size of individual seeds that are sown reduces weed biomass, however the mechanism through which this acts is not yet clear. At high crop densities, reduced row spacing may increase crop competition. Although there has been little work done in the synergy and relative importance of these techniques, in many cases using more than one cultural weed control technique will increase the crop competition in an additive fashion especially when the practices are repeated over time.


Preventive and cultural methods


The invasion of weeds in the archaeological sites and innovated methods for their control

Economou1 G, Papafotiou2 M. and I, Kanellou2 Agricultural University of Athens, Department of Plant Production, 1 Laboratory of Agronomy 2 Laboratory of Floriculture and Landscape Architecture, 75 Iera Odos, 118 55 Athens, Greece

e- mail:[email protected]

One of the most serious problems we have to face in the archaeological sites is the occurrence of shrubs and weeds which inhabit the monuments and the surrounding area, in a short or long period of time making mechanical and chemical destroy. In many cases, considerable part of the historic sites had fallen prey to the abundant wild species that grow uncontrolled in the semiarid climate particularly in the Mediterranean zone. Actually, a considerable effort needs to counter the problem as the delicate context of the historic sites would easily suffer irreparable damage if inappropriate mechanical and chemical means were used.

Taking into consideration all these reasons we applied a plan of vegetation management in the historic site of Eleusis, forty km apart of Athens. Weeds, bushes and shrubs cover a great part of the monuments constituting a severe problem. The control of the undesirable vegetation by herbicides is considered prohibitive by the authorities in order to avoid the deterioration of the monuments. We recorded the vegetation following a stratified procedure in selected sites, whereas we applied alternative methods which were proved effective to control the undesirable vegetation. The soil solarization was applied for weeds control, while an integrated method consisted by mechanical and chemical means was applied to bushes and shrubs. On this vegetation, after cutting the branches, we treated a dense pastry of glyphosate by specialized application. Additionally, the glyphosate was directly applied, in a dense suspension formula, by injection inside the cambium. A particular problem is the olive trees seedlings occurrence on the monuments. One spray with 300 or 400 mg l-1 naphtalinacetic acid, at th many end of the spring, resulted in complete fruit abortion, suggesting an effective control of the great distribution of olive trees. It is worth mentioning the particular method applied for weed control on a mosaic floor in a Roman Villa. The floor surface was covered by layers of quartz sand, matting, LECA and gravel, inhibiting the weeds development. Furthermore, a study was carried out for installation of particular plant species aiming to the restoration of the Villa.


Results and experiences of physical weed control on hard surfaces

D. Hansson1 and H. Schroeder2 1Swedish University of Agricultural Sciences, P.O. Box 104, SE–230 53 Alnarp, Sweden. Email:

[email protected] 2Swedish University of Agricultural Sciences, Alnarp, Sweden. Email: [email protected]

There is currently a focus on research and development of alternatives to chemical weed control methods on hard surface areas, such as flaming, steaming and brush weeding. One important reason is the growing awareness of the disadvantages of herbicides and public resistance to them. Residues from herbicides have been found in surface and well water. Many authorities in Europe have decided not to use herbicides on urban hard surface areas or recommend a restricted use of herbicides.

Weed control research has mainly focused on arable land. Some of these experiences can be applied to hard surface areas. However, weeds on hard surface areas cause problems that are different from those on arable land. Hard surface areas include areas with ground-cover such as asphalt, paving-stone and concrete or surfaces with a top layer of sand, gravel or crushed material. Weeds impair the function of the hard surface area and shorten its lifetime and thereby cause substantial increases in expenditure for the authorities responsible. Furthermore, weeds lead to an increased risk for fires because they produce a mass of dry debris, and they make the surface unattractive and more difficult to clean.

Weeds can be prevented on hard surface areas by changing the design of the surface, and by selecting suitable materials and construction techniques. However, there is a need for suitable non-chemical methods that are acceptable for weed control on hard surface areas. Such methods could be mechanical, thermal or based on low-toxic natural substances. Only a few non-chemical weed control methods are in practical use on hard surface areas today, namely mechanical methods i.e. brush weeding, weed harrowing, hoeing and the thermal method, flame weeding.

Mechanical weed control methods like rotating wire brushes may cause additional costs, mainly depending on mechanical damage to the surfaces which shorten their permanency. The weed brushes are capable to damage some of the roots of weeds growing in the joints between stones and tiles (Hein, 1990). Weed harrowing works only on larger continuous gravel areas with sufficiently thick layer of gravel. Hand weeding often results in overload injuries and high expenses and it may be difficult to find labour.

Thermal weed control methods based on hot water, steam or flames are interesting alternatives to herbicides and mechanical methods. They cause less wear on the treated surface compared to mechanical methods like rotating wire brushes. However, flame weeding involve risks close to houses and cars etc. Hot water and steam treatment eliminates the fire hazards associated with flame weeding. Studies showed that hot water killed most annual weeds, but perennial weeds required repeated treatments (Daar, 1994; Hansson & Ascard 2002).

There is a need for new methods to improve and rationalise the work of assessing the response to weed control on hard surface areas. In field studies the weed control effect has traditionally been assessed as the weight or number of surviving plants. On hard surface areas with a heterogeneous naturally developed weed flora, it is more convenient to assess the weed cover.

References Daar S (1994) New technology harnesses hot water to kill weeds. IPM Practitioner 16, 1-5. Hansson D & Ascard J (2002) Influence of developmental stage and time of assessment on hot water weed control.

Weed Research 42, 307-316. Hein R (1990) The Use of Rotating Brushes for Non-chemical Weed Control on Paved Surfaces and Tarmac. (in

Swedish with English summary). Swedish University of Agricultural Sciences. Department of Agricultural Engineering. Alnarp. Report 141.


Experiences with physical weed control on hard surfaces in central Italy

L. Lulli1, M. Fontanelli1, C. Frasconi1, M. Ginanni2, M. Raffaelli1, F. Sorelli1, A. Peruzzi1 1University of Pisa, Department of Agronomy and Agroecosystem Management, Division

Agricultural Machinery and Farm Mechanization, Italy 2University of Pisa, C.I.R.A.A. “E. Avanzi”, Italy

In Italy weed control in urban areas is mainly performed by means of mowing cutting and herbicide distribution. While trimmers are not effective in reducing weed density and they are also potentially injurious for hard surfaces and the safety of citizens and operators, chemical control induces resistance to active compounds in spontaneous plants and it is surely a source of environmental pollution and a risk factor for the health of human beings and animals. For this reason the use of herbicides in urban areas is strictly regulated by laws. As an alternative to ordinary weed control devices, thermal equipments can be used successfully for weed control on hard surfaces. Flaming machines are the most efficient among thermal devices and they are suitable for treatments in many urban contexts.

The aim of this research was to evaluate the effects of different weed managements (flaming, mowing, herbicide application, and flaming+herbicide application) on weed dynamics in two cities of Tuscany (Central Italy) and to compare the total working time and costs of operations in order to define a proper strategy for the control of weed flora growing on hard surfaces in a typical mediterranean environment.

A hand knapsack device and three motorized flaming prototypes were also projected, built and tested at the University of Pisa. One of the self-propelled versions of the flaming machines was also used for a trial of weed management in the city of Livorno between October 2006 and June 2007 in order to evaluate its technical performances and effectiveness.


Weed problems on pavements

B. Melander1, N. Holst1, A.C. Grundy2, C. Kempenaar3, M.M. Riemens3, A. Verschwele4 and D. Hansson5

1 University of Aarhus, Faculty of Agricultural Sciences, Department of Integrated Pest Management, Research Centre Flakkebjerg, Slagelse, Denmark

2University of Warwick, United Kingdom; 3Wageningen University, the Netherlands; 4Julius Kühn Institute, Germany; 5 Swedish University of Agricultural Sciences, Sweden

Weeds on pavements in urban areas are unwanted mainly because they cause an untidy appearance and at worst structural damage. Municipalities and other public authorities responsible for the maintenance of pavements invest considerable funds and time into keeping pavements clear of unwanted weed growth and in a good state of maintenance. In addition to regular sweeping, glyphosate spraying is the predominant method of weed control in most European towns, usually requiring two applications per year for satisfactory control. However, herbicide use on hard surfaces is under pressure due to the risk of leaching of herbicides into ground water and nearby surface waters. These concerns have influenced policy making in some North European countries leading to federal restrictions in the use of herbicides on hard surfaces. Many municipalities in the Netherlands, Denmark and certain regions of Germany and Sweden have now restricted use of herbicides and there has been a shift towards alternative non-chemical methods. However, these methods are less effective than glyphosate and an optimisation of non-chemical methods would require an improved knowledge about the pavement flora to be controlled, especially the species composition, growth pattern and coverage of the pavement. Consequently, we surveyed the flora on pavements in five North European towns (Braunschweig (DE), Malmö (SE), Næstved (DK), Royal Leamington Spa (UK) and Wageningen (NL)) by recording weed species and their coverage in 56 sampling points randomly placed in each town. The points were scattered over three zones: industrial, residential and town centres, and weeds were recorded at several dates in 2005 and 2006. No weed control was applied during the survey apart from sweeping. Weed coverage increased during the survey (averaging 1.4% in late 2006) and was highest in the towns having the strictest policies on herbicide use. Industrial zones were mostly more weedy than the other zones with 3.5% weed coverage on average and more weed coverage (averaging 2%) was found along the pavement edge away from the road. Poa annua was the most frequently recorded species followed by mosses, Sagina procumbens and perennial grasses. Grasses and other species frequently found, notably Taraxacum officinale, should receive particular attention when planning a non-chemical weed control campaign on pavements.

Selected references Kristoffersen P, Rask AM, Grundy AC, Franzen I, Kempenaar C, Raisio J, Schroeder H, Spijker J, Verschwele A and Zarina L (2008). A review of pesticide policies and regulations for urban amenity areas in seven

European countries. Weed Research 48, 201-214. Rask AM & Kristoffersen P (2007) A review of non-chemical weed control on hard surfaces. Weed Research 47, 370-

380.


Mechanical weed control


Morphological differences between carrot and weeds: its usefulness in selective mowing as a weed control technique

D.L. Benoit Agriculture and Agri-Food Canada, Saint-Jean-sur-Richelieu (Quebec) Canada J3B 3E6 Email:

[email protected]

Carrots have basal leaves with compact internodes forming a rosette compared to most weeds which have specific internode length. The aim of this research was to take advantage of this morphological difference between carrot and weeds to develop a selective mowing method to achieve maximum damage to weeds with minimal impact to carrots. An experimental protocol was set up under controlled conditions (greenhouse or growth chamber) and under field condition to determine the regrowth potential of carrots, common ragweed (Ambrosia artemisiifolia-AMBEL) and redroot pigweed (Amaranthus retroflexus-AMARE) following topping of seedlings at different early phenological stages. Two carrot cultivars were tested Appache and Sugar Snap. Under controlled conditions, pots filled with artificial soil mixture are seeded with 10 seeds of each species, placed at 25 C day and 15 C night with a photoperiod of 16 hr day and 8 hr night. After emergence, seedlings are thinned down to 6 seedlings per pot. Under field conditions, all species were seeded at the same time as 1 row/species and each row was 11 m long with 1 m between rows. Weed species were seeded at 10 cm intervals while carrots were seeded with a precision seeder. Four contiguous seedlings were monitored for each combination of cutting height and growth stage. Seedlings are cut at 5 different phenological stages (cotyledon, 2, 4, 6, and 8 leaves stage) and at various heights (below cotyledons (CS1), above cotyledons (CS2), 1st internode (CS3) , 2nd internode (CS4), 3rd internode (CS5) and uncut control (T)). The treatments were completely randomized with 2 replicates. Plant height before and after cutting, cutting height, number of leaves and branches at 7 days intervals and 14 days after CS5 and aerial biomass were recorded 14 days after CS5. Under both controlled conditions, carrot cultivars did not differ from each other in their growth development and AMARE had consistently a slower growth development rate than AMBEL. If cutting occured at the 1 or 2 leaves stage, carrots grew 1-2 new leaves within 14 days. However, if cutting was done at the 3-4 leaves stage, carrots grew 2-4 new leaves depending on the variety, Appache producing more leaves than SugarSnap. Cutting above the 1st internode stimulated branching to a greater extent in AMBEL than in AMARE. More branching was stimulated when ragweed was cut at the 4-6 leaves stage than when cut at a later stage. Based on the results under controlled conditions, the selective mowing should be done minimally at the 6 leaves stage of AMBEL and AMARE, approximately 27 days after seeding. This means that the carrots would be at that time between the 1 to 2 leaves stage and 6-7cm in height. Because of the mechanical limit of the cutting tool, the minimal cutting height is ~7cm. Thus selective mowing is theoretically feasible as early as the 2 leaves stage of carrots at which time the cutting blade will cut the weeds and barely touch the carrot leaves.


New innovations for intra-row weed control

Pieter Bleeker and Rommie van der Weide Applied Plant Research Wageningen UR (PPO-WUR), Lelystad, the Netherlands.

Email: [email protected]

In mechanical weed control recent developments aim at increasing capacity and accuracy in inter-row and intra-row weed control. With accurate guidance the width of the untreated strip in the crop row can be reduced considerably without yield loss. Cultivator-mounted RTK-GPS in the Netherlands resulted in untreated strips of 4 cm with a driving speed of 4 km h-1 and strips of 6 cm with higher speeds.

With appropriate machine adjustments, the machinery have made intra-row control of small weeds possible in many crops and in several growth stages. Important aspects that need to be improved are the higher driving speeds of the machines to make the techniques practical and economical and ways to get still closer to the crop. Each centimetre closer to the plant saves a lot of manual weeding. For larger crops there is an intelligent weeder with a simple crop detector system based on a light interceptor, with guides a hoe in and out of the crop row around the crop plants. The first intelligent intra-row weeder for lettuce has been improved by fitting the weeder with a cutting blade at either side of the crop row. This reduces the uncultivated area around the crop plant to half, leaving fewer weeds. And because the blade only needs to bridge half of the distance in the row it is now possible to drive almost twice as fast (4 km per h) without damaging the crop. The machine has been tested in sugar beet in 2008, and onion will be next in 2009. The working speed of 4 km per hour is still rather slow. The aim is a speed of 6 to 7 km. Speeds can still be increased by improvement of the sensors and the pneumatic and hydrologic system. In Dutch research together with different manufacturers the different opportunities are used to improve the machines. For recognizing of smaller crop plants cameras are used that calculates whether a plant is weed or crop. In 2009 an organic lettuce grower will start using a faster hoe with this camera. Other actuators (blowers, pneumatic, hydraulic and rotating hoes) are also in development.

In slow germinating, small seeded and low-competitive crops like carrots, direct-sown leeks and onions, intra-row weeding remains difficult. In this case Dutch research is focusing on applying clean compost in the crop rows. Such a compost layer preventing the germination of annual weeds at early crop development, substantially reduces the required amount of hand weeding. This reduces the need for application of finger and torsion weeders in very small crop stages. One of the research questions was how much compost should cover the onion seed. The thickness of this layer appeared to be very critical. A layer of about 2 cm resulted in best crop emergence combined with good weed suppression. Results of a 2-cm compost layer again were best in a follow-up experiment with carrots. Weed emergence was reduced by 75 to 85 per cent in both cultures. After the successful experiments in cooperation with an organic grower an machine was developed for simultaneously sowing and putting down a compost layer: the band sowing machine.

Currently, advanced technologies such as sensing crop and weed plants and robotics are regarded as an important way to improve mechanical weed control in existing cropping systems. However, the adjustment of cropping systems to new developments and technologies may be equally important. One of those developments is societal pressure to reduce the energy consumption of plant production, which may lead to reduced ploughing and no-till systems.

References Cloutier, D.C., R.Y. van der Weide, A. Peruzzi and M. LeBlanc (2007) Mechanical weed management. In M.K.

Upadhyaya, R.E. Blackshaw, eds. Non-chemical Weed Management: Principles, Concepts and Technology. CABI, Oxon, UK, p. 111-134.

Kropff, M.J., L. Bastiaans , C. Kempenaar and R.Y. Van der Weide (2008) The changing role of agriculture and tomorrow’s weed research agenda. Journal of Plant Diseases and Protection, Special Issue XXI, 2008, p.3-8.


Van der Schans, D.A., P.O. Bleeker, L. Molendijk, M. Plentinger, R.Y. van der Weide, L.A.P. Lotz, R. Bauermeister, R. Total and D.T. Baumann. (2006) Practical weed control in arable farming and outdoor vegetable cultivation without chemicals. Applied Plant Research, Wageningen University and Research Centre, Lelystad, The Netherlands.

Sukkel, W. (ed) (2008) Dutch Research in Organic Agriculture. Biokennis Publication pp ....... (in press) end this year on www.aboutdutchorganic.nl

Van der Weide, R.Y., P.O. Bleeker, V.T.J.M. Achten, L.A.P. Lotz, F. Fogelberg and B. Melander (2008) Innovation in mechanical weed control in crop rows. Weed Research, p. 215-224


The use of flex-tine harrow, torsion weeder and finger weeder in Mediterranean crops

A. Cirujeda1, J. Aibar2, S. Fernández-Cavada3, P. Zuriaga4, A. Anzalone5 and C. Zaragoza1

1Centro de Investigación y Tecnología Agroalimentaria (Gob. de Aragón), Avda. Montañana 930; 50059 Zaragoza. Spain ([email protected])

2Escuela Politécnica Superior de Huesca, Carretera de Cuarte s/n, 22071 Huesca. Spain. 3Centro de Protección Vegetal, DGA, Avda. Montañana 930; 50059 Zaragoza. Spain.

4Serv. Prov. de Agricultura y Alimentación, Gob. Aragón. C/ San Francisco 1; 44001 Teruel. Spain. 5Dept. Fitotecnia. Universidad Centroccidental “Lisandro Alvarado”, Apdo Postal 400 Venezuela.

Many parameters having an influence on weed control are different in Mediterranean conditions compared to the North-European situations. There parameters are the crops themselves, weed species, soil conditions (stoniness, soil crust, heavy soil textures), irrigation, which has a big influence on weed emergence and on the soil conditions, etc. The use of mechanical weed control devices designed in Northern Europe needs to be adapted to those conditions to achieve high efficacy. The problems found in two very different crops, saffron and irrigated processing tomato, are discussed in this communication.

Saffron is grown usually in cold drylands at 20 cm distances between rows and needs weed control after harvesting the flowers in late October until summer, where few weeds grow due to drought. The leaves develop after flowering and die in May. The corms with the new flower buds develop in March, in Spanish conditions. Weed control during winter needs to respect the leaves but after May, tillage can be conducted overall without respecting the crop, which remains buried at 20 cm depth. One problem found in this crop is the presence of tap-rooted weeds like Descurainia sophia or grass weeds needs to be controlled very early with mechanical methods. As the plantation of the corms is conducted in rows at 20cm distance, the devices need to be fitted at small distance and accurate steering is needed with torsion and finger weeding. In spring, weed harrowing is probably the best solution for weed control, provided that the weeds are small enough.

Processing tomato completes the cropping cycle in 3 to 4 months. Irrigation is provided in the conditions of North-Eastern Spain normally by drip irrigation in warm areas. Row distance is between 1 and 2 meters, allowing the use of different weed control tools. The continuous moistering and drying of the soil frequently causes soil crusting, so that torsion weeder and finger weeder may need a previous device loosening the soil, e.g. by harrowing previously. Another problem found is the frequent presence of perennial species with subterranean rhizomes as Cyperus rotundus, which is very difficult to be controlled by the tested devices. In processing tomato an early treatment is a key-factor to achieve good control, as the crop plants are able to compete with small new emerged weeds later on. It is important to work with crop protectors when using aggressive weed control tools because the tomato plants are susceptible to harms. Another additional problem can be the presence of drip pipes, which can hinder a correct mechanical control. In these cases, the burial of the pipes can be a good solution. In fields where different summer weed species are frequent, emerging gradually, two control moments may be necessary e.g. when Portulaca oleracea and C. rotundus emerge in the first flush short time after planting but again when Amaranthus retroflexus and Digitaria sanguinalis emerge in the second moment at flowering stage of the tomato.


In pursuit of effective mechanical/physical weed management in organic lo-till

W.S. Curran1, R.T. Bates1, S.B. Mirsky2, R.S. Gallagher1, D.A. Mortensen1, and M.R. Ryan1

1Penn State University, University Park, PA, USA Email: [email protected] 2USDA, ARS, ANRI, Beltsville, MD, USA

Organic farmers in the northeastern USA face many challenges including farming on erodible soils with vulnerable watersheds and the strong desire to use sustainable farming practices. Historically, organic crop farming has relied heavily on tilling the soil to prepare the seedbed and to help manage weeds. Most organic farmers make between 7 and 12 trips over their fields during the first half of the growing season with various implements that disturb the soil and control weeds. The negative impact of tillage (loss of soil quality, carbon, energy inputs, etc.) has stimulated a growing interest in identifying and adopting practices that are less tillage intensive, while at the same time providing sufficient weed control in organic crop production systems.

The primary tactics that we are pursuing include using cover crop mulches for physical weed suppression along with shallow tillage (lo-till) to control weeds and preserve crop residues. These practices are being tested in rotational tillage systems (i.e. not long term organic no-till). In these systems, winter cover crops are established in late summer or early fall to protect the soil over the winter and perhaps provide N (legumes) for the following cash crop. In late spring, cover crops are controlled with a roller/crimper and cash crops are no-till seeded into the rolled mulch. Depending on the cover crop and amount of biomass residue as well as weed species and severity, 4 weeks or more of weed suppression can be achieved. As weed severity increases, the efficacy of the cover crop surface mulch declines. Some of this work is examining how to optimize cover crop growth and biomass by altering planting and termination date to target specific weed species. Early emerging summer annual weeds such as Ambrosia artemisiifolia require large amounts of cover crop surface mulch earlier in spring to suppress emerging seedlings, while later emerging weeds like Amaranthus hybridus do not require the mulch until late spring through early summer for adequate weed suppression. In maize and soyabean planted in 76 cm rows, a high residue row cultivator can be used if necessary as an in-season rescue to control weeds between the crop rows. The wider row system is being compared to soyabean seeded in 19 cm rows without cultivation. In 2008, in-row cultivation improved control in maize no-tilled into a hairy vetch mulch, but was less consistent in the soyabean-cereal rye system. Additional implements being tested include a vertical coulter followed by a rotary harrow prior to planting the cash crop to control emerged weeds and a high residue rotary hoe after cash crop planting to disrupt newly germinated weed seedlings. In 2008, the complement of surface tillage implements that included in-season row cultivation significantly reduced weed populations, but also reduced maize surface residues below the 30% minimum. Like the surface mulch, the severity of the weed population influenced efficacy of the mechanical control tools. We believe using cover crops for surface mulch and incorporating some shallow tillage for weed control can be an effective weed management approach, while maintaining the benefits of reduced tillage. Research will continue to identify and refine ways for improving this approach in organic cropping systems.


Innovative operative machines for physical weed control on organic cauliflower in Central Italy

M. Fontanelli1, C. Frasconi1, L. Lulli1, F. Sorelli1, S. Carlesi2, F. Bigongiali2, D. Antichi2 and A. Peruzzi1

1 MAMA - DAGA, University of Pisa. Via del Borghetto 80, 56124 Pisa, Italy mail address: [email protected]

2 Land Lab, Scuola Superiore Sant’Anna. Piazza Martiri della Libertà, 33, 56127 Pisa, Italy.

Abstract Organic fruit and vegetable consumption in Italy has been considerably increasing over the last

three years (about 25%). Furthermore, this sector is actually involving nearly one fifth of the total Italian organic market.

In this context, an on-farm field trial was carried out in the municipality of Crespina (Tuscany, Central Italy) for one growing season (2007-2008), aiming to develop and set up innovative machines and strategies for weed control on organic cauliflower (Brassica oleracea L. convar. botrytis (L.) Alef. var. botrytis).

Three different weed management strategies were compared within an organic farming context: 1) standard crop management system (SCMS) that consisted in the use of biodegradable plastic film; 2) an intermediate crop management system (ICMS), that consisted in the use of innovative operative machines for weed control; 3) an advanced crop management system (ACMS), that includes the same operations as ICMS plus the use of a hairy vetch (Vicia villosa Roth) living mulch.

In this work performances of innovative operative machines and hand labour requirements for weed control and transplanting were assessed. Moreover economic parameters were estimated.

ICMS and ACMS showed a sensibly lower labour time requirement for weed control and transplanting (-60%) and considerably higher gross income with respect to SCMS (on average 7000 € ha-1 vs -490 € ha-1).

Moreover ACMS showed a slightly higher labour time requirement and a slightly lower gross marketable production with respect to ICMS.

Introduction Italian organic harvested area has been sensibly increasing during the last three years and now

its value is almost over one million hectares, so that Italy is at the moment the first European country and the fifth nation all over the world for organic cultivation (Biobank, 2008; Sinab, 2008).

Furthermore organic fruit and vegetable consumption in Italy has been considerably increasing over the last three years (about 25%), and this sector is actually involving nearly one fifth of the total Italian organic market (De Ruvo, 2008).

This trend is probably due to the strong concern about pesticides and herbicides pollution and food contamination by chemical residues, that is actually a very common feeling of the consumers all over the world (van der Weide et al., 2008).

In this context, there is the clear need for Italian organic vegetable growers to have effective machines for physical weed control at their disposal, in order to reach high yields and a good quality of their production (Peruzzi et al., 2007c).

One of the most serious problem for organic farmers is weed control in the crop-row, where it is more difficult to reduce weed density using machinery, so that very expensive manpower interventions are required (Peruzzi et al., 2007c; van der Weide et al., 2008).

For that reason, many tools for physical selective intra-row weed control are available now, as many kinds of harrows, finger and torsion weeders and the weed blower (Cloutier et al., 2007; van der Weide et al., 2008). Furthermore, non selective tools can be also used for intra-row weed


control when combined with electronic devices for crop-detection (Dedousis et al., 2007; van der Weide et al., 2008).

Aiming to develop and set up innovative machines and strategies for weed control on organic cauliflower (Brassica oleracea L. convar. botrytis (L.) Alef. var. botrytis) – one of the most important Italian open-air vegetable crops – an on-farm field trial was carried out in the municipality of Crespina (Tuscany, Central Italy, 43°34’26.66’’ N, 10°33’55.20” E) for one growing season (2007-2008).

This work takes into account just data regarding machines performances, operative times and cultivation costs, while all the agronomical aspects of the research are reported in another paper presented at this Workshop (Carlesi et al., 2009).

Materials and Methods The on-farm experiment

The field trial was included in the FertOrtMedBio research project (Mediterranean Organic Horticulture Fertilisation), in which a two year organic vegetable crop rotation was adopted (spinach-potato-cauliflower-tomato) (Bàrberi et al., 2008; Fontanelli et al., 2008).

Three different weed management strategies were compared within an organic farming context: 1) standard crop management system (SCMS); 2) an intermediate crop management system (ICMS); 3) an advanced crop management system (ACMS).

SCMS consisted in the use of biodegradable black plastic film as preventive weed control method. It is the standard practice ordinarily adopted by the farmer. In this case cauliflower was manually transplanted.

ICMS consisted in the use of innovative operative machines for physical weed control. False seedbed technique was carried out by means of two passes of rolling harrow. Post-transplanting interventions were performed by means of a precision hoe and the hoe conformed rolling harrow, equipped with an hand guidance system and elastic tines for in-row selective weed control. Finally a row ridging intervention was carried out with an on purpose made operative machine.

ACMS was characterized by the same operations described for ICMS plus the inclusion of hairy vetch (Vicia villosa Roth) as living mulch crop (Carlesi et al., 2009). In this case the ridging intervention was not performed.

In this work performances of innovative operative machines and hand labour requirements for weed control and transplanting were assessed. Moreover economic parameters were estimated. Data were not processed by ANOVA because are referred to operative characteristics and economical estimations. The innovative machines for physical weed control

Innovative weed control was carried out by means of two different machines: the rolling harrow and a precision hoe (Figures 1, 2, 3,4).

The rolling harrow is a patent of the University of Pisa and it can be easily use for false or stale seedbed technique and also for hoeing interventions properly adjusting the working tools (spike disks at the front and cage rolls at the rear) (Peruzzi et al., 2007a, 2007b and 2007d).

In this case it was used twice before crop transplanting and once as an hoeing machine. For this second passage the first version of rolling harrow was used, that is not equipped with the precision guidance system and elastic tines for selective in-row weed control (Figures 1 and 3).

Moreover two passes with a precision hoe were performed (the first and the third intervention after crop transplanting) (Figures 2 and 4). The machine was equipped with three elements, each one characterized by one rigid goose-foot tine, two lateral “L” shaped blades elements and a torsion weeder for intra-row selective weed control (Peruzzi et al., 2007a, 2007b and 2007d).


Figure 1. The rolling harrow during a false seedbed intervention before cauliflower transplanting.

Figure 2. First hoeing intervention carried out by means of the precision hoe.

Figure 3. Second hoeing intervention carried out by means of the “hoe-conformed” rolling harrow.


Figure 4. Late hoeing intervention carried out by means of the precision hoe.

Results and discussion Operative performances of all the innovative operative machines for physical weed control are

shown in Table 1. The rolling harrow during false seedbed technique interventions was characterized by the

highest values of working speed and capacity, so that it can be considered the cheapest machine. On the other hand, post-transplanting interventions carried out by means of the precision hoe showed the lowest working speed values (about 3.5 km h-1), while the “hoe-conformed” rolling harrow, for the same operation, allowed to reach nearly 7 km h-1 (Tab. 1). Table 1. Performances of operative machines adopted for physical weed control on cauliflower. Parameters Rolling harrow

(I° pass) Rolling harrow

(II° pass) Precision hoe

(I° pass) Hoe conformed rolling harrow

Precision hoe (II° pass)

Working width m 1.35 1.35 1.35 1.35 1.35

Woking depth cm 3.52 3.11 3.22 2.10 4.15 Working speed km h-1 6.84 10.34 3.53 6.63 3.65 Working capacity ha h-1 0.85 1.29 0.46 0.81 0.46 Working time h ha-1 1.17 0.78 2.20 1.24 2.15 N. of operators - 1.00 1.00 2.00 1.00 2.00 Tractor power kW 37.40 37.40 37.40 37.40 37.40 Engine load % 20.00 20.00 20.00 20.00 20.00 Fuel consumption kg ha-

1 2.37 1.57 4.44 2.50 4.34

Concerning working times for all the operation that differed among the three farming systems,

SCMS was characterized by the highest values, mainly because of manual transplanting. Furthermore manpower requirement for weed control was lower than 15 h ha-1 for ICMS and ACMS, so that the innovative technique appeared definitively economically sustainable (Tab. 2).

Gross marketable production was relevantly higher for ICMS and ACMS with respect to SCMS, because of a significant higher yield (Carlesi et al., 2009). Total costs, including all variable costs, were slightly lower for SCMS, just because a lower time was required for harvest, as a consequence of the very low yield obtained in this system. Gross income was negative for SCMS while was about 7000 € ha-1 for the two other systems (Tab. 3).

Finally it is possible to conclude that weed control treatments carried out by means of the innovative operative machines were economically sustainable, according to the high working speed reached and the low working time requirement.


Table 2. Total manpower requirement for transplanting and weed control for the three compared farming systems.

Manpower requirement (h ha-1) Farming system Transplanting Mechanical

weed control Vicia villosa interseeding

Mulching with MaterBi film

Total

SCMS 81.0 - - 8.0 89.0 ICMS 25.0 13,6 - - 38.6 ACMS 25.0 11,9 0.4 - 37.3 SCMS, standard crop management system; ICMS, intermediate crop management system; ACMS, advanced crop management system.

ICMS and ACMS showed a sensibly lower labour time requirement for weed control and transplanting and considerably higher economic parameters with respect to SCMS.

Moreover ACMS showed a slightly higher labour time requirement and a slightly lower gross marketable production with respect to ICMS.

SCMS, on the other hand, did not appear economically sustainable. Table 3. Gross marketable production, total costs and gross income for organic cauliflower cultivated with three different farming systems

Farming system Gross marketable production (€ ha-1)

Total costs (€ ha-1)

Gross income (€ ha-1)

SCMS 4220 4709 -489 ICMS 12250 4738 7513 ACMS 10901 4805 6095 SCMS, standard crop management system; ICMS, intermediate crop management system; ACMS, advanced crop management system.

References BÀRBERI P., BIGONGIALI F., ANTICHI D., CARLESI S., FONTANELLI M., FRASCONI C. & LULLI (2008). Innovative crop

and weed management strategies for organic spinach: crop yield and weed suppression. In: Proceedings of the 16th IFOAM ORGANIC WORLD CONGRESS, Cultivate the future, Modena 16-20, Volume 1 (Organic crop production), 252-255.

CARLESI S, BIGONGIALI F, ANTICHI D, FONTANELLI M, FRASCONI C, LULLI L, SORELLI F & P. BÀRBERI (2009) Effect of innovative crop and weed management systems on organic cauliflower in Central Italy. In: Proceedings of 8th EWRS Workshop on Physical and Cultural Weed Control, Zaragoza, Spain, 9-11 March.

CLOUTIER D C, VAN DER WEIDE R Y, PERUZZI A & LEBLANC M L (2007) Mechanical weed management. Chapter of the book: Non-chemical weed management, Principles, Concepts and Technology, Edited by Upadhyaya M K and Blackshaw R E, Cabi International, Wallingford, Oxfordshire, UK. 111-134.

DEDOUSIS A P, GODWIN R J, O’DOGHERTY M J, TILLETT N D, GRUNDY A C (2007) A novel system for within the row mechanical weed control. In: Proceedings of 7th EWRS Workshop on Physical and Cultural Weed Control, Salem, Germany, 11-14 March. 112.

DE RUVO (2008) I consumi biologici crescono ancora. L’Informatore Agrario 32, 27-29. FONTANELLI M., FRASCONI C., LULLI L., ANTICHI D., BIGONGIALI F., CARLESI S., BÀRBERI P. & PERUZZI A. (2008)

Innovative crop and weed management strategies in organic spinach: machine performance and cultivation costs. In: Proceedings of the 16th IFOAM ORGANIC WORLD CONGRESS, Cultivate the future, Modena, 16-20 June, Volume 1 (Organic crop production), 256-259.

PERUZZI A, GINANNI M, RAFFAELLI M & FONTANELLI M (2007a) Physical weed control in organic chicory cultivated in the Fucino Valley (South Italy). In: Proceedings 7th Workshop of the EWRS Working Group on Physical and Cultural Weed Control, Salem, Germany, 11-14 March, 22-31.

PERUZZI A, GINANNI M, RAFFAELLI M & FONTANELLI M (2007b) Physical weed control in organic fennel cultivated in the Fucino Valley (South Italy). In: Proceedings 7th Workshop of the EWRS Working Group on Physical and Cultural Weed Control, Salem, Germany, 11-14 March, 32-40.

PERUZZI A, GINANNI M, FONTANELLI M, RAFFAELLI M & BÀRBERI P (2007c). Innovative strategies for on-farm weed management in organic carrot. Renewable Agriculture and Food Systems. 22 (4), 246-259.


PERUZZI A, GINANNI M, RAFFAELLI M, FONTANELLI M, FRASCONI C & LULLI L (2007d) Physical weed control in organic carrot cultivated in the Catania Plain (South Italy). In: Proceedings 7th Workshop of the EWRS Working Group on Physical and Cultural Weed Control, Salem, Germany, 11-14 March, 41-52.

VAN DER WEIDE R Y, BLEEKER P O, ACHTEN V T J M, LOTZ L A P, FOGELBERG F AND MELANDER B (2008) Innovation in mechanical weed control in crop rows. Weed research 48, 215-224.

www.biobank.it www.sinab.it


Innovative operative machines for physical weed control on processing tomato in the Serchio Valley (Central Italy)

Fontanelli M.1, Raffaelli M.1, Ginanni M.2, Lulli L.1, Frasconi C.1, Sorelli F. 1 and Peruzzi A.1 1 MAMA - DAGA, University of Pisa. Via del Borghetto 80, 56124 Pisa, Italy

mail address: [email protected] 2 CIRAA “Enrico Avanzi”, University of Pisa. Via Vecchia di Marina, 6, 56122 San Piero a Grado

(PI), Italy. mail address: [email protected]

Abstract In ItalyProcessing tomato is the first Italian vegetable crop for harvested area (about 94000 ha). A three year “on-farm” open field research on processing tomato was carried out from 2006 to

2008 in a conventional farm in the Serchio Valley (Pisa, Central Italy), with the aim of testing innovative strategies for non-chemical weed control.

The innovative strategy was compared with the farm conventional technique (consisting in chemical treatments and rotary inter-row hoeing interventions). The innovative strategy was carried out by means of the stale-seedbed technique (consisting in a rolling harrowing and flaming treatments) and post-transplanting precision hoeing interventions (with innovative hoes equipped with rigid or rotating tools). All the operative machines were on purpose adapted or modified according to tomato space arrangement. The rolling harrow was equipped with spike disks (placed in the front) and cage rolls (placed at the rear), connected by an overdrive with a ratio equal to 2. The flaming machine was equipped with three 50 cm wide rod burners, for a total working width of 1,5 m. The precision hoe was equipped with rigid elements for inter-row cultivation, elastic elements for in-row selective weed control and a precision guidance system. With this machine it was possible to till soil and control weeds even inside the crop pairs, without removing the drip irrigation hoses. Furthermore, “V” shaped elements, that allow to “open” crop vegetation during late hoeing interventions, were appositely built. The operative machines performances, weed density during the crop cycle, dry weed biomass at harvest and crop fresh yield were recorded.

The innovative strategy allowed to reach significantly higher yield values, an efficient weed control and a relevant increase of gross marketable production with respect to conventional management in all the three years of experiment.

Introduction In Italy processing tomato is the most important vegetable crop, although a significant

reduction of tomato harvested area was observed in Italy in the last three years (-40%, from 113000 to 68800 ha). This trend is mostly due to political (uncertainty of CMO reform) and economical (high cultural fixed costs) reasons (Bazzana, 2008).

The production valorization could be a good strategy in order to follow the new policy trends and to guarantee accurate profits to farmers. This aim could be easily reached by means of cultivation that respect environment and consumer health safety.

At the moment physical weed control is mostly studied in Northern Europe, where strict pesticide action plans are commonly launched in order to reduce agrochemicals (about the 50% of which are herbicides) use in agriculture (Melander, 2007).

Otherwise processing tomato is a typical Mediterranean crop. Thus, with the exception of some recent Spanish field trials (Cirujeda et al., 2007), at the moment on this crop scientific knolled are not available.

Paired-rows transplanting is at the moment one of the most utilized spatial crop arrangement for processing tomato cultivation in Italy, and it’s the system adopted in the Serchio Valley (Tei et al., 2008). This system is surely characterized by some advantages with respect to the single row cultivation, as more contemporaneous fruit maturation, an easier field accessibility with operative


machines, the use of one irrigation line per paired-row instead of one irrigation line per row (Tei et al., 2008) but, at the same time, it is very difficult to control weeds into the pair space. The presence of the irrigation line in the middle of the pair makes the treatment even more difficult. For this reason conventional farmers usually carry out post-transplanting mechanical intervention (usually with PTO powered rotary hoes) just in the inter-pair space.

In this work, the results of a three year-long (2006-2008) “on-farm” open field research carried out by MAMA division of DAGA and CIRAA “E. Avanzi” of the University of Pisa are reported. The aim of the study was develop and improve innovative strategies and operative machines for an effective physical weed control on processing tomato.

Materials and methods

The experimental trial The experiment was carried out during three growing seasons (2006-2008) on processing

tomato in a conventional farm placed near Pisa. The tomato varieties utilized were two hybrids: “Leader” in the first year and “Reflex” in the second and third year. The crop was mechanically transplanted on paired rows at the density of 33000 plants ha-1 (1.10 m of inter-pair space; 0.4 m of intra-pair space and 0.25-0.30 m of intra-row space) (Figure 1). Crop was irrigated by drip hoses placed in the middle of the intra-pair space. Organic-mineral fertilizer was applied before crop-planting while fertirrigation was carried out in post-emergence. Soil was sandy-loam and a four year rotation was adopted (tomato, wheat, maize, sunflower).

The experiment consisted in the comparison between the traditional farm weed management system (FS) and an innovative physical weed control system (PWCS). FS was carried out by means of three different chemical treatments: one before crop transplanting (1 kg ha-1 of “Stomp” – a.i. Pendimetalin – and 1 kg ha-1 of “Ronstar” – a.i. Oxadiazon) and two after crop establishment (250 g ha-1 of “Sencor” – a.i. Metribuzin – and 40 g ha-1 of “Titus” – a.i. Rimsulfuron”). FS was performed also with two post-transplanting PTO powered rotary hoe interventions (not able to till the soil in the intra-pair space). PWCS was carried out applying the innovative strategies and machines developed by the University of Pisa.

Hand weeding was also performed when necessary in both the two different systems, in order to control well developed weeds in the in-row space that survived to the treatments.

Figure 1. Pair-rows transplanted processing tomato experimental field during the second year of trial. The irrigation line is placed in the middle of the pair. The innovative physical weed control strategy

PWCS was carried out by means of the stale-seed bed technique (realized by one rolling harrow pass followed by one flaming treatment) and two post-transplanting precision hoeing interventions. The aim of this strategy is to reduce decidely the surface weed seed-bank and consequently to reduce weed emergence during the crop cycle.


Precision hoeing interventions were performed in order to reduce weed presence during the crop cycle in the inter-pair space, intra-pair space and selectively in the in-row space.

The innovative operative machine for physical weed control

Three different innovative operative machines were used for physical weed management: a rolling harrow, a flaming machine and a precision hoe.

The rolling harrow was projected, built, tested and patented by Pisa University. It was set up both for pre-sowing (or pre-transplanting) and post-emergence hoeing (for inter-row and intra-row selective weed control) interventions (Figure 2). Working tools are spike disks (placed in the front) and cage rolls (placed at the rear), respectively mounted on two different parallel axles. The axles are connected by an overdrive with a ratio equal to 2. Spike discs till the soil very shallowly while cage rolls (rotating with a double peripheral speed) allow to separate weed seedling roots from soil (Peruzzi et al., 2008). In this case the treatment was carried out just before crop trans-planting with a working speed of 7 km h-1 and working depth of about 4 cm.

The flaming machine controls weeds by the use of open flame. In this experiment it was equipped with three 50 cm wide rod burners, for a total working width of 1,5 m. This treatment has the advantage of eliminating weeds without stimulating new emergence because the soil remains undisturbed (Figure 3).

The machine was equipped with three common 15 kg weight LPG tanks placed into an on purpose made hopper. Furthermore this machine was also equipped with an innovative heat exchange system, in order to avoid tanks cooling during the flaming. The treatments were performed just in the pre-transplanting phase, but if necessary, tomato may tolerate post emergence selective flaming interventions (with the flame directed to the crop collar) (Peruzzi et al., 2008).

Figure 2. The rolling harrow at work during the pre-transplanting intervention in 2008.


Figure 3. The flame weeder at work before tomato transplanting in 2008

The precision hoe is a machine 3 m wide. It is equipped with rigid elements for inter-row cultivation (a centrol “foot-goose” tool and two side “L” shaped sweeps) and elastic elements for intra-row selective weed control (torsion weeders and vibrating tines). The operative machine is also equipped with a seat, steering handles and directional wheels. By means of these tools, it was possible to till soil and control weeds even inside the crop pairs, without removing the drip irrigation hoses (Figure 4a). Furthermore, the precision hoe was equipped with on purpose made “V” shaped elements, that allowed to “gently-open” crop vegetation during late hoeing interventions(Figure 4b). Average working speed was about 2 km h-1 and working depth was about 4 cm.

a) b) Figure 4. The precision hoe during an early intervention on processing tomato in 2007 (a). Detail of the “V” shaped tool utilized for the late hoeing intervention carried out in 2006 (b). Experimental assessments, experimental desing and data analysis

During the trials, all data concerning the operative performances of the operative machines used for physical weed control were recorded: working depth, working speed, working capacity, working time, LPG working pressure, fuel and LPG consumption.

Furthermore, numerous weed parameters were recorded at repeated times. Weed density was measured before and after each physical weed control treatments on three 25 x 30 cm sampling areas plot-1. At harvest, weed and fruit samples were collected from a 1,2 m2 area plot-1 (corresponding to the surface covered by four tomato plants). Weed samples were then oven dried until constant weight, in order to assess dry biomass.

Some economic parameters were also considered. The gross marketable production as gross income was calculated for both the weed management systems compared


The experimental design was a randomized block with four replicates. Data were analyzed by ANOVA.

Results Innovative machine operative performances

Innovative machines performances are shown in table 1. All the machines with the working width of was 1.5 m, in accordance with the traditional traffic

lanes system adopted in the Serchio Valley and in most of the vegetable production contexts in Central and Southern Italy.

Tractor used in the trials engine power ranged from 44 up to 60 kW, power values were absolutely an excess considering the real operative machine power requirement.

Both the machines for soil tillage reached very few working depths, with a maximum value of about 4 cm.

The rolling harrow was characterized by the best operative performances, reaching the highest working speed and working capacity values (about 7 km h-1 and 1 ha h-1 respectively), while precision hoeing (that is surely a more difficult operation) appeared the most expensive treatment because of its low working speed (about 2 km h-1) and the presence of a second back seated operator. Before the precision hoeing, the rolling harrow, conformed for post-emergence treatments, was used in the last year of trials, its working speed value was 1,94 km h-1 and its working capacity was 0,29 ha h-1.

Flaming was performed only in two years 2006 and 2008. In the first year, working speed was about 3,5 km h-1 and LPG consumption was about 35 kg ha-1, but in the last year the working speed value was 7 km h-1 and LPG consumption was about 23,81 kg ha-1,

Fuel consumption was very low for all the tested machine. However, precision hoeing, a result of a low working speed, showed sensibly higher fuel consumption values with respect to flaming and rolling harrowing. Table 1. Mean operative performances of the innovative machines for physical weed control during the three year of experiment.

2006 2007 2008

Characteristics

Rolling harrow

Flamer Hoe Rolling harrow

Hoe Rolling harrow

Flamer Rolling harrow

hoe conform

ed

Hoe

Working width (m) 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5

Working depth (cm) 3.5 - 3.8 2.8 3.1 2.8 - 2.7 4.2

Working speed (km h-1) 6.8 3.4 1.9 6.4 1.3 6.5 7.0 1.9 1.4 Working capacity

(ha h-1) 0.9 0.5 0.3 0.8 0.2 1.0 1.1 0.3 0.2

Working time (h ha-1) 1.1 2.2 3.9 1.3 5.7 1.15 1.2 3.6 5.0

Operators (No.) 1.0 1.0 2.0 1.0 2.0 1.0 1.0 1.0 2.0

Tractor engine power

(kW) 55.0 55.0 55.0 44.2 44.2 58.3 58.3 58.3 58.3

Fuel consumption

(kg ha-1) 3.3 6.5 11.6 3.1 13.6 3.63 3.9 11.5 15.7

LPG pressure (MPa) - 0.2 - - - - 0.4 - -

LPG consumption

(kg ha-1) - 35.9 - - - - 23.8 - -


Weed control Weed density trend for the PWCS in the three years of experiment is shown in Figure 5.

Innovative machines always controlled weeds very well. The effectiveness of rolling harrow and flamer was 100%, while precision hoeing efficiency varied from 50% up to 100% depending on tomato and weeds development stage.

FS weed average density registered during the three years ranged from 2 up to 10 plants m-2 (data not shown in the graph). Yield, weed dry biomass at harvest, labour time and crop economy

In the first and in the second years, the innovative strategy allowed to reach significantly higher yield values (from 15 to 20%), a good weed control and a relevant increase of gross income with respect to conventional management (Table 2). Yield increase could be probably due to the positive agronomical effects of the intra-pair hoeing (crust breaking, soil oxygenation, etc.).

Furthermore PWCS was characterized by sensibly higher total labour time values in 2006 and 2007 (+260% and +110% respectively) caused by the higher need of working time for in-row hand weeding, but the gross income increase completely justify by produce characteristic obtained.

During the last year, significant differences between PWCS and FS were not recorded for yield and total labour time, because the weather was characterized by more rain and higher humidity, so that huge amount of hand weeding with a conserve productivity drop. Table2. Yield, weed biomass at harvest, total labour time was requirement and gross income registered during the three years of activity.

Weed management system Yield (Mg ha-1)

Weed biomass (g m-2)

Total labour timea

(h ha-1) Gross incomea

(€ ha-1) 2006

Conventional system 59.4 b 102.9 ns 15.0 2600.8

Innovative system 72.1 a 126.1 ns 54.1 3298.8

2007

Conventional system 54.1 b 2.1 ns 11.3 2649.9

Innovative system 61.9 a 21.9 ns 24.0 3528.9

2008

Conventional system 51.7 ns 5.1 b 50.0 2077.2

Innovative system 52.1 ns 56.0 a 61.0 2074.9 Different letters on the same column and for the same year mean significant differences for P≤0,05 (LSD test). aData were not analyzed by ANOVA.

Conclusions The innovative physical weed control strategy allowed to reach higher yields and gross

marketable production values during the three years of experiment. Furthermore, innovative operative machines for physical weed control appeared very versatile,

suitable and adaptable to the processing tomato crop. Moreover, these machines can be easily utilized for weed control in organic farming, where herbicides use is not permitted. The results of this three years experiment showed that the alternative cultural strategy could be suitable not only for environment and consumers health but also for farmers gross income.


a)

b) 2008

0

200

400

600

800

1000

1200

5-05-08 12-05-08 19-05-08 26-05-08 2-06-08 9-06-08 16-06-08

Weed

den

sity

(p

lan

ts m

-2)

rolling harrow hoe conformed

precision hoe

rolling harrow

c) Figure 5. Weed density trend during the crop cycle in 2006 (a), 2007 (b) and 2008 (c) for the innovative weed control system (arrows indicate physical weed control treatments).

References BAZZANA L (2008). Superfici a pomodoro da programmare meglio. L’Informatore Agrario, 40, 12. CIRUJEDA A, ANZALONE A, PARDO G, LEON M & ZARAGOZA C (2007). Mechanical weed control in processing tomato.

Proceedings of 7th EWRS Workshop on Physical and Cultural Weed Control, Salem, Germany, 11-14 March, 2007.

MELANDER B (2007). Status on physical and cultural weed control methods for field crops in Europe. Proceedings of the Conference on Novel and Sustainable Weed Management in Arid and Semi-Arid Ecosystems, Rehovot, Israel,October 7-12, 2007.

PERUZZI A, RAFFAELLI M, GINANNI M, LULLI L, FRASCONI C & FONTANELLI M (2008). Innovative operative machines for physical weed control on processing tomato in the Serchio Valley (Central Italy). Proceedings of International


Conference “Innovation Technology to Empower Safety, Healt and Welfare in Agriculture and Agro-food Systems”. Ragusa - Ibla Campus: 15-17 September, Italy 71.

TEI F, NATALINI G & BRUNI R (2008). Manuale di corretta prassi per la produzione intergrata del pomodoro da industria. Progetto per la Valorizzazione delle Produzioni Agroalimentari Umbre.

www.parco3a.org.


Non-chemical weed control on open-field fresh market tomato in the Serchio Valley (Central Italy)

Fontanelli M.1, Raffaelli M.1, Ginanni M.2, Lulli L.1, Frasconi C.1, Sorelli F. 1 and Peruzzi A.1 1 MAMA - DAGA, University of Pisa. Via del Borghetto 80, 56124 Pisa, Italy

mail address: [email protected] 2 CIRAA “Enrico Avanzi”, University of Pisa. Via Vecchia di Marina, 6, 56122 San Piero a Grado

(PI), Italy. mail address: [email protected]

Abstract Fresh-market tomato is one of the most widespread Italian open-field vegetable cultivation,

involving over than 23000 ha. A three year on-farm experimental trial (2006-2008) was carried out by the MAMA Division of

DAGA Department and CIRAA “Enrico Avanzi” research centre of the University of Pisa in cooperation with San Giuliano Terme and Vecchiano municipalities (Serchio Valley, Pisa, Central Italy).

The experiment was run in two different “pilot” integrated farms (the first one in 2006 and 2008 and the second one in 2007) placed in the Serchio Valley. Tomato was transplanted following a single row space arrangement. Crop density was about 10000 plants ha-1 (1,5 m x 0,7 m). Crop was not irrigated.

The aim of the research was to compare three different non-chemical weed control methods: 1) biodegradable black plastic mulching film; 2) physical weed control on bare soil; 3) physical weed control plus straw dead mulch (only during 2007 and 2008).

Biodegradable plastic mulch was placed just before crop transplanting. In this case just few hand weeding interventions were performed in addiction. No false or stale seedbed technique was carried out. This is the developing ordinary cultivation technique adopted in the area.

Physical wed control on bare soil was performed with innovative operative machines built by the University of Pisa. Stale-seedbed technique was covered at by means of the rolling harrow (a patent of the University of Pisa) and an open flame operative machine before crop transplanting. Hoeing interventions were carried out with a precision hoe and the hoe conformed rolling harrow, both of them equipped with a hand guidance system and elastic tines for selective in-row weed control. In-row hand weeding was also performed when necessary.

The third method consisted in the use of straw dead mulch together with the physical weed control strategy previously described. The straw mulch was placed just after the last hoeing intervention in 2007 and after crop transplanting in 2008. The purpose of the straw mulch use was to increase weed control effectiveness and water supply for the crop.

Weed density, weed dry biomass at harvest, fruit fresh yield, operative machine performances and economic parameters were assessed.

During the first year of experiment, the use of plastic biodegradable mulch allowed to reach high yields and lower weed dry biomass values at harvest with respect to the physical weed control method applied to bare soil. In 2007 the use of the straw mulch allowed to reach similar yield and economical parameters with respect to the biodegradable plastic mulch utilization, while physical weed control methods applied to bare soil gave again worse results. In 2008 the two innovative techniques allowed to reach higher yield and weed control levels with respect to biodegradable plastic film application.

Introduction Tomato is one of the most important vegetable crop production in Italy. In 2008, over 500000

Mg of fresh-market tomato were produced in Italy, involving over than 24000 ha (Istat, 2008).


In Tuscany, fresh tomato total production is actually about 16000 Mg (Istat, 2008), and the Serchio Valley represents one of the most important production areas of Central Italy.

The production valorisation (for example by organic cultivation) could be a good strategy in order to follow the new policy trends and to guarantee an appropriate iincome to the farmers. This goal could be easily reached by means of environment and consumers health “friendly” cultural practices.

Moreover, tomato is a “Minor crop”, as most of vegetable crops, and the process of labelling new herbicides for specialty crop is often very difficult and expensive, so that nowadays, hand weeding is still an important and widespread weed control method, especially in organic and integrated agriculture (Fennimore and Doohan, 2008). For this reason, at the moment, new operative machines for physical weed control development is one of the most important research topics for weed science (Fennimore and Doohan, 2008; van der Wade et al., 2008).

In this work, the results of a three year long (2006-2008) “on-farm” open field research are reported. The experiment was carried out by MAMA Division of DAGA department and CIRAA “E.Avanzi” research centre of the University of Pisa in cooperation with San Giuliano Terme and Vecchiano municipalities (Serchio Valley, Pisa, Central Italy) with the aim to compare three different non chemical weed control methods on fresh tomato.

Material and methods The experiment was run in two different “pilot” integrated farms (the first one in 2006 and

2008, the second one in 2007) placed in the Serchio Valley. The tomato varieties utilized were two different hybrids: “Italpeel” the first and last year, “Incas” the second year. The crop was hand transplanted on single row at the density of about 10000 plants ha-1 (1,5 m of inter-row space; 0,75 m of intra-row space). Crop was not irrigated. Soil primary tillage was chiseling at about till 40 cm of depth. Seedbed preparation was carried out with a rotary harrow till 15 cm of depth. Fertilization was performed by distributing Unimer (4-8-16) at the rate of 700 kg ha-1 in the first and last year, and Nitrophoska at the rate of 600 kg ha-1 in the second year (8-24-2) before crop transplanting. Post transplanting fertilization was carried out using “Nutri Leaf” (leaf fertilizer 20-20-20) only during the second year, at the rate of 12,5 kg ha-1. Soil was sandy-loamy and the rotation adapted was characterised by the cultivation of spinach and black cabbage before cauliflower.

The experiment consisted in the comparison three different non-chemical weed control methods: biodegradable black plastic mulching film utilization among (BPMF), physical weed control on bare soil (BS) and physical weed control plus straw dead mulch (only during 2007 and 2008) (SDM).

Hand weeding was also performed, when necessary, in all the three tested systems, in order to remove from the intra-row space all the weeds that were not controlled by the previously cited means. The biodegradable black plastic mulching film weed control strategy (BPMF).

Biodegradable plastic mulch was placed just before crop transplanting. In this case just few hand weeding interventions were performed in addiction. No false o stale seedbed technique was carried out. It is important to stress that the intervention with rotary harrow promotes the fragmentation of taproots and stolons and the diffusion of weed seeds along the tilled soil profile. This is the developing ordinary cultivation technique adopted in the area. Physical weed control on bare soil (BS)

Physical weed control on bare soil consisted in the application of stale-seedbed technique, as preventive method and post-transplanting precision hoe interventions as direct method. The stale seedbed technique consists in one or more shallow tillage (3-5 cm of depth) intervention, carried out, by means of an on purpose made operative machine, before crop planting. These interventions have substantially two main issues: actual flora control and weed emergence stimulation. This


technique is also characterized by the use of the rolling harrow and the flamer before crop planting. Hoeing interventions were carried out by means of a precision hoe and the hoe conformed rolling harrow, both of them equipped with a hand guidance system and elastic tines for selective in row-weed control. In row hand weeding was also performed when necessary. Physical weed control plus straw mulch application (SDM)

The third method consisted in the use of straw dead mulch together with physical weed control strategy (Figure 1). The straw mulch was utilized in order to increase weed control and preserve soil water content, thus, drought problems could be reduced during summer time. The straw mulch was applied in 2007 after the last hoeing intervention in July, while in 2008, the straw was applied immediately after crop transplanting. Obviously, also in this latest innovative strategy the innovative operative machines built by the University of Pisa were used.

Figure 1. Tomato covered with straw mulch in 2008. The innovative operative machines for physical weed control

During the experiment, three different machines were utilized: the rolling harrow, the flaming machine and the precision hoe.

The rolling harrow was projected, built, tested and patented by the University of Pisa. It was set up both for pre-transplanting and post emergence hoeing (for inter-row and intra-row selective weed control) interventions (Figure 2). Working tools are spike disk (placed in the front) and cage rolls (placed al the rear), respectively mounted on two different parallel axles, connected by an overdrive with a ratio equal to 2. Spike discs till the soil very shallowly while cage rolls (rotating with a double peripheral speed) allow separating weed seedling roots from soil (Peruzzi et al., 2008, 2007a, 2007b and 2007c). In this case the treatment was carried out just before crop transplanting with a working speed of 7 km h-1 at a depth of about 4 cm.

Figure 2. The rolling harrow at work during the pre-transplanting intervention in 2008.


The flaming machine controls weeds by the use of an open flame. In this experiment it was equipped with five 25 cm wide rod burners, for a total working width of 1,5 m. This treatment is able to significantly reduce weed density without stimulating new seedling emergence, because in this case soil remains undisturbed (Figure 3).

The machine was equipped with three common 15 kg weight LPG tanks placed into an on purpose made hopper. Furthermore, this machine was also equipped with an innovative heat exchange system, in order to avoid tanks cooling during the flaming treatment. The treatment was performed just in the pre-transplanting phase, but if necessary, tomato may tolerate post emergence selective flaming interventions (with the flame directed to the crop collar) (Peruzzi et al., 2008, 2007a, 2007b and 2007c).

Figure 3. The flamer at work before tomato transplanting in 2006.

The precision hoe is characterized by a 3 m wide frame. It is equipped with rigid elements for inter-row cultivation (a central “foot-goose” tine and two side “L” shaped sweeps) and elastic elements for intra-row selective weed control (torsion weeders or vibrating tines) (Figure 4). The operative machine is also equipped with a seat, steering handles and directional wheels. By means of these tools, it was possible to till soil and control weeds very close to crop row. Average working speed was about 2 km h-1 and working depth was about 4 cm (Peruzzi et al., 2008, 2007a, 2007b and 2007c).

Figure 4. The precision hoe during an early intervention on fresh tomato in 2007. Experimental assessments, experimental designs and data analysis

The experimental design was a randomize block with four replicates. Data were analyzed by ANOVA.

Innovative operative machines performances (working depth, speed, capacity and time, engine load, LPG working pressure, fuel and LPG consumption), weed density during the crop cycle, dry weed biomass at harvest and crop fresh yield were recorded. Weed density was measured before and after each physical weed control treatments on three 25 x 30 cm sampling areas per plot. At harvest, weed and fruit samples were collected from a 1 m2 area per plot. Some economic


parameters were also considered. The gross marketable production and gross income were calculated for the three compared weed management systems.

Results and Conclusions Experimental assessments, experimental designs and data analysis

Innovative machines performances are shown in table 1. All the machines were characterized by a working width of 1,5 m, in accordance with the

traditional traffic lanes adopted in the Serchio Valley. Tractor used in the experiment had engine power ranging from 40 up to 55 kW, but this values

were absolutely excessive considering the power requirements of the operative machines. Both the machines for soil tillage reached very few working depths, with a maximum value of about 4 cm.

The rolling harrow was characterized by the best operative performances, reaching the highest working speed and capacity values (about 7 km h-1 and 1 ha h-1 respectively), while precision hoeing (that is surely a more difficult operation) appeared the most expensive treatment because of its low working speed (1,5 km h-1) and the presence of a second back seated operator.

In 2007 weed density before transplanting was not high enough to justify a thermal treatment. As a matter of fact, flaming was performed only in 2006 and 2008. In the first year, working

speed value about 3,4 km h-1 and LPG consumption was about 36 kg ha-1, but in the last year the working speed value was 6 km h-1 and LPG consumption was about 27 kg ha-1.

In 2007, the hoe conformed rolling harrow showed a sensibly higher working speed with respect to the precision hoe (3.7 km h-1 vs 1.5 km h-1 respectively), thus the first machine was characterized by lower working time and fuel consumption values.

Moreover, fuel consumption was very low for all the tested machine. However, precision hoe, as a result of its low working speed, showed sensibly higher fuel consumption values with respect to flaming and rolling harrow. Table 1. Mean operative performances of the innovative machines for physical weed control during the three-year of experiment.

2006 2007 2008

Characteristics Rolling harrow

Flamer Hoe Rolling harrow

Rolling harrow hoe conformed

Hoe Rolling harrow

Flamer Hoe

Working width (m) 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 Working depth (cm) 3.5 - 3.8 3.2 1.8 3.5 3.6 - 3.1 Working speed (km h-1) 6.8 3.4 1.9 6.1 3.7 1.5 7.2 6.0 1.1 Working capacity (ha h-1) 0.9 0.5 0.3 0.7 0.5 0.2 0.9 0.7 0.2 Working time (h ha-1) 1.1 2.2 3.9 1.3 2.0 4.6 1.1 1.4 6.2 Operators (No.) 1.0 1.0 2.0 1.0 2.0 2.0 1.0 1.0 2.0 Tractor engine power

(kW) 55.0 55.0 55.0 40.4 40.4 40.4 40.6 40.6 46.6

Fuel consumption (kg ha-1) 3.3 6.5 11.6 2.9 4.4 10.1 2.3 3.0 13.7 LPG pressure (MPa) - 0.2 - - - - - 0.4 - LPG consumption (kg ha-1) - 35.9 - - - - - 27.8 -

Weed control

Weed density trend during the crop cycle is shown in figure 4: BS for all three years, SDM and BPMF only for the last year. Weed densities of 2007 for SDM are not present because the straw was distributed after the end of all interventions of physical weed control. BPMF weed average density registered during the first and the second year ranged from 2 up to 10 plants m-2 (data not shown in the graph), weed values were surprising high during 2008 because of the bad weather conditions registered especially during spring time (very high levels of rainfall).


Innovative machines controlled weeds very well, reaching the 100% of effectiveness in the case of rolling harrow and flamer, while precision hoe efficiency varied from 50% up to 100% depending on tomato and weeds development stage.

2006

0,00

5,00

10,00

15,00

20,00

25,00

30,00

35,00

40,00

31-05-06 10-06-06 20-06-06 30-06-06 10-07-06

Weed

den

sity

(p

lan

ts m

-2)

Rolling harrow

Flamer Precision hoe

a) 2007

0,0

100,0

200,0

300,0

400,0

500,0

600,0

700,0

800,0

18-05-07 28-05-07 7-06-07 17-06-07 27-06-07 7-07-07

Weed

den

sity

(p

lan

ts m

-

Rolling harrow

Precision hoe

Rolling harrow

Precision hoe

b) 2008

0,00

100,00

200,00

300,00

400,00

6-05-08 16-05-08 26-05-08 5-06-08 15-06-08 25-06-08 5-07-08

Weed

den

sity

(p

lan

ts m

-

BS

SDM

BPMF

Rolling harrow

Flamer

Precision hoe

Precision hoe

c) Figure 4. Weed density trend registered during the crop cycle in 2006 (a), 2007 (b) and 2008 (c) (arrows indicate physical weed control treatments). Yield, weed dry biomass at harvest, labour time and economical evaluation

The weed control system with biodegradable black plastic film showed excellent results during the first and second year of the trials. This strategy was characterised by good yields, low weed biomass values and a good gross income, while during the last year, BPMF appeared the worst solution, because of the bad weather conditions (rainy spring and dry summer).

BS allowed to obtain the best results covering yield and gross income during the last years, when weather conditions were not ordinary, while it was the worst solution during 2006 and 2007, where soil mulching appeared necessary in order to guarantee good yields.


Finally, the physical weed control combined with the distribution of straw as dead mulch allowed to obtain excellent results, in both the two years of use (similar to BMPF in 2007 and similar to BS in 2008). Table 2. Yield, weed biomass at harvest, total labour time and gross income registered during the three years of activity.

Weed management

system

Yield (Mg ha-1)

Weed biomass (g m-2)

Total labour timea

(h ha-1) Gross incomea

(€ ha-1)

2006

BPMF 38.2 a 25.8 ns 16.0 7087.9

BS 29.4 b 49.7 ns 14.1 5479.9

2007

BPMF 26.0 a 19.9 ab 36.0 4660.5 BS 17.2 b 34.7 a 25.89 1243.5

SDM 25.2 a 10.2 b 25.89 4653.5 2008

BPMF 12.9 b 84.9 ns 26.0 948.97

BS 22.9 a 44.1 ns 21.9 8838.1

SDM 19.6 a 67.8 ns 29.4 5906.1

Different letters on the same column and for the same year mean significant differences for P≤0,05 (LSD test). aData were not analyzed by ANOVA.

The use of non-chemical weed control allowed to obtain very good positive results during the three years of experiment. The three systems tested gave different results depending on weather condition. Furthermore, innovative operative machines for physical weed control appeared very versatile and suitable to fresh tomato crop.

Finally, the use of non-chemical weed control guarantees higher market values of the product.

References FENNIMORE S A & DOOHAN D J (2008). The challenges of specialty crop weed, future direction. Weed technology 22,

364-372. PERUZZI A, RAFFAELLI M, GINANNI M, LULLI L, FRASCONI C & FONTANELLI M (2008). Innovative operative machines

for physical weed control on processing tomato in the Serchio Valley (Central Italy). In: Proceedings of International Conference “Innovation Technology to Empower Safety, Healt and Welfare in Agriculture and Agro-food Systems”. Ragusa - Ibla Campus: 15-17 September, Italy 71.

PERUZZI A, GINANNI M, RAFFAELLI M & FONTANELLI M (2007a) Physical weed control in organic chicory cultivated in the Fucino Valley (South Italy). In: Proceedings 7th Workshop of the EWRS Working Group on Physical and Cultural Weed Control, Salem, Germany, 11-14 March, 22-31.

PERUZZI A, GINANNI M, RAFFAELLI M & FONTANELLI M (2007b) Physical weed control in organic fennel cultivated in the Fucino Valley (South Italy). In: Proceedings 7th Workshop of the EWRS Working Group on Physical and Cultural Weed Control, Salem, Germany, 11-14 March, 32-40.

PERUZZI A, GINANNI M, RAFFAELLI M, FONTANELLI M, FRASCONI C & LULLI L (2007 c) Physical weed control in organic carrot cultivated in the Catania Plain (South Italy). In: Proceedings 7th Workshop of the EWRS Working Group on Physical and Cultural Weed Control, Salem, Germany, 11-14 March, 41-52.

VAN DER WEIDE R Y, BLEEKER P O, ACHTEN V T J M, LOTZ L A P, FOGELBERG F & MELANDER B (2008) Innovation in mechanical weed control in crop rows. Weed research, 48, 215-224.

www.istat.it


Physical weed control on cabbage in the Serchio Valley (Central Italy)

M. Fontanelli 1, M. Raffaelli 1, M. Ginanni 2, L. Lulli 1, C. Frasconi 1, F. Sorelli 1 and A. Peruzzi 1

1 MAMA - DAGA, University of Pisa. Via del Borghetto 80, 56124 Pisa, Italy Email: [email protected]

2 CIRAA “Enrico Avanzi”, University of Pisa. Via Vecchia di Marina, 6, 56122 San Piero a Grado (PI), Italy. Email: [email protected]

Abstract Cabbage is a common open-air vegetable crop in Italy (nearly 40000 ha), and although it is a

very competitive species, weed management still represents a key operation during the crop cycle. One year “on-farm” experiment was carried out during 2006-2007 season by the MAMA

Division of DAGA Department and CIRAA “Enrico Avanzi” researches centre of the University of Pisa in cooperation with San Giuliano Terme and Vecchiano municipalities (Serchio Valley, Pisa, Central Italy).

The innovative weed control strategy was performed by means of the false seedbed technique application, two precision hoeing interventions and in-row hand weeding as complementary operation.

False seedbed technique was performed by the rolling harrow (a patent by the University of Pisa), while hoeing passes were carried out with the “hoe conformed” rolling harrow and a precision hoe, both of them equipped with hand guidance systems and elastic tines for selective in-row weed control.

Weed density, weed biomass at harvest, crop fresh yield, innovative operative machines performances and economic parameters were assessed during the experiment.

The two compared weed management systems gave very similar results concerning with all the parameters involved in 2006. Another year of experiment was carried out from the autumn of 2008. The results of this year of experiment currently are being processed and will be presented at the workshop.

Introduction Previous experimental researches carried out on organic spinach in the municipalities of San

Giuliano Terme (PI) and Vecchiano (PI) (Peruzzi et al. 2005a) allow defining strategy and setting up operative machines for physical weed control. The experimental research, was carried at since 2006 on cabbage and tomato (Peruzzi et al. 2008) in order to transfer these techniques on other crops. The botanic varietes object of this research were Brassica oleracea L. var botrytis L. (cauliflower) and Brassica oleracea L. var sabauda L. (savoy cabbage).

The on farm trials aimed at comparing the traditional and ordinary weed management to a physical weed management carried out by means of innovative machines designed, realized and tested by the MAMA-DAGA of the University of Pisa. Owing to the rapid development, and the good competitiveness of cabbage, the weed flora represents an adversity of this cultivation, only after transplanting during the end of summer and the first months of fall (Bianco & Pimpini 1990). The cultivation of cabbage in the province of Pisa is very important; the estimated production in the last three-year period is approximately the 23% of the whole regional production of Tuscany (ISTAT, 2008). The trials were carried out in both conventional and integrated farms. The aim of this research was a comparison between the farm and an innovative strategy for weed control. The trials were carried out in integrated and conventional farms with the aim to demonstrate that the physical management of weed flora can be successfully used, not only in organic farming and it can substitute herbicides based techniques. This substitution is connected with the reduction of the use chemicals and sometimes with a reduction of the labour time for hand weeding. Thus the use of


non-chemical weed control is a key strategy for environment protection and operators and consumers safety.

Materials and methods The experiment was carried out in two different farms placed in the Serchio Valley

(municipalities of San Giuliano Terme 43°45′45″N 10°26′29″E and Vecchiano 43°47’0”N 10°23’0”E) that is a typical cabbage production area of Tuscany.

The aim of the trial was to compare the ordinary farming weed management system with an innovative physical weed control strategy performed with on purpose made operative machines. The innovative equipments were: a rolling harrow and a precision hoe.

The rolling harrow is equipped with spike discs placed on an axle and gage rolls mounted on a second axle placed at the rear. The front and rear tools are connected by an overdrive with a gear ratio equal to 2. The disc and the rolls can be placed differently on the axles: in close arrangement, in order to perform a very shallow tillage (3-4 cm) of the whole treated area (false seed bed tecnique); and in spaced arrangement, with the possibility to use also elastic tools in order to per form precision hoeing between and in the rows (Peruzzi et al. 2005b, 2008) (fig. 1).

Figure 1. Rolling harrow used during the experimental trials on cabbage in 2006, characterized by a 2 m working width, hand steering system and elastic tines.

Precision hoe was equipped with an hand guidance system and several working units connected to the frame by means of articulated parallelograms. Each working unit is equipped with rigid elements (one goose foot tine and two side “L” shapd blades) and with couple of elastic tools (working as vibrating tines or torsion weeder) in order to perform, once again, weed control both between and in the rows.

The first farm was a conventional farm. In this case the cabbage botanic variety was savoy, the second farm was an integrated farm. In this case the cabbage botanic variety was cauliflower. Savoy cabbage

The seedbed preparation was carried out after a 40 cm deep ploughing by means of a rotary hoe

800 kgha-1 of chemical ternary fertilizer “Superalba” (title 9-12-21) were distributed. Savoy was mechanically transplanted on the 20/08/2006 following a single row space arrangement for a crop density of about 30000 plants ha-1 (0,75 m x 0,45 m). The savoy cultivar was “Barbosa” cicle of 120 days.


20 days after the transplanting 300 kg ha-1 of urea were distributed (equivalent to 150 kg ha-1 of nitrogen). The hand harvesting started at the beginning of February.

The ordinary farm weed management technique was characterized by one post-transplanting chemical intervention (1 kg ha-1 of “Butisan”, a.i. Metazaclor) and two post transplanting mechanical interventions (carried out with a PTO powerd hoe and a conventional static hoe).

The innovative weed management consisted in false seedbed technique carried out with a passage of rolling harrow and two post transplanting hoeing. The first one carried out with the rolling harrow hoe conformed (space arrangement) (fig. 2), and the second with the precision hoe (fig. 3). Both operative machines worked on “strip” 1,5 m wide. A hand weeding intervention was also performed.

Figure 2. Effect of the first hoeing intervention performed on savoy cabbage with the rolling harrow.

Figure 3. Precision hoe equipped with 4 elements, performing the second intervention on savoy cabbage.


Cauliflower Also in this case the seedbed preparation was carried after a 40 cm deep ploughing by means of

a rotary hoe. Before transplanting 700 kg ha-1 of mixed mineral organic fertilizer “Unimer” (title 4-8-16)

were distribuited. Cauliflower cultivar adopted was “Talbot” that was mechanically transplanted in single rows for a crop density of about 22000 plants ha-1 (0,75 x 0,6 m) the 16/08/2006. 15 day after transplanting 250 kg ha-1 of chemical fertilizer “Idro” (title 11-22-16) were distribuited. The hand harvesting started the 27/12/2006.

The conventional weed control system consisted in one ordinary hoeing intervention carried out with a static operative machine, followed by a ridging intervention.

The innovative weed management included false seedbed technique, two precision hoeing and a row-ridging intervention. The false seedbed technique was carried out by means of the rolling harrow before transplanting. In this case hoeing was always performed by means of the precision hoe (fig. 4). The row-ridging intervention was carried out with the same operative machine adopted in the farm system. Also in this case a hand weeding intervention was performed.

Figure 4. Precision hoe with 4 element working on cauliflower

In order to evaluate the effectiveness of the different compared streategies, weed inter-row and intra-row density was calculated in a 0,25 m2 sample area. Operative performances of the machines were recorded for all the interventions performed.

At harvest crop production and weed biomass were determined. Experimental design was a randomized block with three replicates. Data collected were analysed by ANOVA. Treatment means were separated by Fisher least square difference at P< 0,05 (Gomez e Gomez 1984).

Gross marketable production and costs of weed control were estimated for both the two compared different managements.

Results and discussion In table 1 are reported the mean operative performances used machines.All the operative

machines worked at a very reduced depth and with low engine load (20%) of the tractor. The false seedbed treatment was very fast (6,7 km h-1) and consequentely characterized by a

high working capacity (0,9 ha h-1). The use of the rolling harrow for precision hoeing on savoy, was characterized by a high working speed (5 km h-1) allowing to perform the treatment with a working capacity more than two times higher with respect to the precision hoe. However rolling harrow can be used for hoeing only in the first stage of development of cabbage, as later interventions could damage crop leaves.


Table 1. Mean characteristics of the operative machines used for physical weed control on cauliflower and savoy cabbage. Working characteristics Rolling harrow Rolling harrow hoe

conformed* Precision hoe

Effective Working width m 1,5 1,5 1,5 Working depth cm 3,2 3,2 5,1 Working speed km h-1 6,7 5,0 2,4 Working capacity ha h-1 0,9 0,7 0,3 Working time h ha-1 1,1 1,5 3,1 Number of operators 1 2 2 Tractor power kW 55 55 55 Engine load % 20 20 20 Fuel consumption kg ha-1 3,3 4,4 9,2 * Only used for savoy Savoy cabbage

Weed flora initially observed on cauliflower was characterized by summer species like Amaranthus retroflexus L., Chenopodium album L., Cyperus spp. with an overall relative density value ranging from 15% to 25% and by summer grasses (60%). Figures 5 and 6 show that the first hoeing treatment performed with the rolling harrow allowed a good intra and inter-row control, although before the intervention weed density was about 100 plants m-2. After the two hoeing interventions weed density in the innovative systems decreased to levels similar to those of the ordinary system. At harvest no weed dry biomass was recorded for both the compared strategies.

Figure 5 Trend of the inter-row weed density for the two management systems recorded on savoy cabbage.


Figure 6 Trend of the intra-row weed density for the two management systems recorded on savoy cabbage. Yield results highlighted that there was no significant difference between the two compared strategies (fig 7).

a a

0

5

10

15

20

25

30

35

40

Ordinary weed management Innovative weed management

Yie

ld (

t/h

a)

Figure 7. Yield results obtained in savoy cabbage for the two weed management systems tested. Different letters indicate significant differences (LSD =0,05). Cauliflower

The prevalent weed species during the first development stage after transplanting was Portulaca oleracea L. with a relative density of 80% (fig. 8). Other weeds like Poligonum spp. and Solanum nigrum L. were also present with relative density values lower than 20%. The effectiveness of post transplanting mechanical treatments is shown in figures 9 and 10. It’s evident that hoeing reduced the intra-row and inter-row weed density respectively by the 70% and 80%. After the second hoeing and the following row-ridging, weed density in the innovative management was similar to that observed in the conventional system. At harvest no weed dry biomass was recorded for both the compared weed control managements.


Figure 8. Infestation of Portulaca oleracea L on cauliflower.

Figure 9. Trend of the inter-row weed density for the two compared strategies used in cauliflower.


Figure 10. Trend of the intra-row weed density for the two compared strategies used in cauliflower.

Also in this case, yeld results highlighted that there was no significance difference between the two compared strategies (Fig 11)

aa

0

5

10

15

20

25

30

35

40

Ordinary weed management Innovative weed management

Yie

ld (

t/h

a)

Figure 11 Yields results obtained in cauliflower for the two weed management systems tested. Different letters indicate significant differences (LSD =0,05).

In the ordinary system hand weeding was not performed. However in the innovative strategiey hand weeding use was low for both savoy and cauliflower (Table 2) Table 2 Hour of manpower spent for manual weeding on savoy and cauliflower in the two different weed control managements. Hand weeding (h/ha)

Crop Innovative weed control Savoy 10

Cauliflower 15

Costs of weed control in savoy were quite similar for both the tested system, while in cauliflower higher cost of 290 € ha-1 was observed for the innovative management (fig 13).


0

50

100

150

200

250

300

350

400

450

Cauliflower Savoy

Co

sts

for

weed

co

ntr

ol

€/

ha

Ordinary weed control

Innovative weed control

Figure 13 Costs for weed control in savoy and cauliflower for the two compared strategies.

Gross marketable productions net of the costs for the two weed control managements (GMPnwc) were also calcualted (fig 14). In cauliflower the innovative management reached a value of GMPnwc slightly higher (+133 € ha-1 with respect to the conventional technique equal to 1,2%). On the contrary in savoy cabbage, the ordinary management was connected with a higher value of GMPnwc (about + 500 € ha-1) with the respect to the innovative strategy.

9500

10000

10500

11000

11500

12000

12500

13000

Cauliflower Savoy

GM

P n

et

of

Co

sts

for

weed

co

ntr

ol €

/h

a

Ordinary weed control

Innovative weed control

Figure 14 Estimated GMP net of the costs for two strategies used in savoy cabbage and cauliflower.

Conclusions In this paper are reported the results of the first year of the experiment on the use of physical

weed control in cabbage. The on farm research is still in progress. The obtained results are quite promising, but there is a clear need to carry on the research to improve the performances of the strategies for physical weed control and consequently, to reduce total working times and costs and increase yield results and farmer income.


Acknowledgments The authors would like to thank Mrs Roberta Del Sarto, Mr Alessandro Pannocchia, Mr Paolo

Gronchi, and Mr Calogero Plaia of the M.A.M.A.-D.A.G.A. and C.I.R.A.A. of the University of Pisa.

References BIANCO VV & PIMPINI F (1990) Cavolfiore (Brassica oleracea L var botrytis L. In: Orticoltura. (eds Bianco VV &

Pimpini F) 1th edn,359-376. Patron, Bologna, Italy. GOMEZ KA & GOMEZ AA (1984) In: Statistical procedures for agricultural research. (eds Gomez K.A., Gomez A.A)

2nd edn, John Wiley & Sons U.S.A ISTAT 2008 dati congiunturali sull’agricoltura www.istat.it PERUZZI A, GINANNI M, RAFFAELLI M & FONTANELLI M (2005) Physical weed control in organic spinach in the Serchio

Valley (Italy). In: Proceedings 2005 13 th EWRS Symposium, Bari, Italy. PERUZZI A, GINANNI M, RAFFAELLI M, DI CIOLO S (2005 b) The rolling harrow: a new implement for physical pre and

post emergence weed control.In: Proceeding 200513 th EWRS Symposium, Bari Italy. PERUZZI A, RAFFAELLI M, GINANNI M, FONTANELLI M, LULLI L & FRASCONI C (2008). The rolling harrow: a new

operative machine for physical weed control. In: Proceedings 2008 International Conference on Agricultural Engineering, Hersonissos, Crete - Greece, 23-25 June. Paper n. 1177932.

PERUZZI A, RAFFAELLI M, GINANNI L, LULLI L, FRASCONI C & FONTANELLI M (2008) Innovative strategies for physical weed control on processing tomato in the Serchio Valley (Central Italy) In: Proceedings 2008 of the 16 TH IFOAM ORGANIC WORLD CONGRESS Cultivate the future Modena Italy .


Managing weed seed rain to enhance physical weed control efforts

E.R. Gallandt Sustainable Agriculture Program, University of Maine, 5722 Deering Hall, Orono, ME, 04469-

5722, USA Email: [email protected]

Modern weed management continues to be focused on controlling weed seedlings, most often using herbicides or cultivation. This focus on early season weed control effectively minimizes yield loss due to weeds, but weed seed rain from individuals that survive may be extraordinary, resulting in a recurring weed problem. Furthermore, growers are frequently encouraged to plant fall cover crops, an operation that usually involves tillage and, consequently, burial of the new seed rain. Burial offers protection from predators, and fewer environmental cues to encourage germination, a fatal process for summer annual weeds when it occurs late in the year. In herbicide-based cropping systems, it is possible to largely ignore such critical seedbank processes because seedling mortality is generally very high. However, as the efficacy of most cultivation operations may be comparatively less, with increasing seedbank densities, the weed pressure in crops increases. Satisfactory weed control thus requires a systematic effort focused on the seedbank, not only weed seedlings (Gallandt, 2006).

We have completed two years of research on fall weed management strategies including: preemption of seed rain by hand hoeing/pulling, tillage, or mowing followed by fall cover cropping; managing seed rain by flail mowing for maximal presentation to post-dispersal seed predators (exclosures installed to demonstrate the contribution of various weed seed predator guilds); and no-till seeding of fall cover crops to maintain weed seed on the soil surface where they are exposed to weed seed predators.

Evidence from both on-farm and on-station experiments support our original hypothesis that seed predators may contribute to significant fall/overwinter/spring seed removal. At two on-farm sites, and at the the replicated on-station trial, there were an average of 60% more germinable weed seeds within predator exclosures, suggesting that predators were responsible for a 40% reduction in the seedbank. For example, in the replicated trial, the germinable weed seedbank averaged 62,000 seeds per square meter within exclosures, but 36,000 outside the exclosures. Inexplicably, we failed to detect a significant difference in the germinable seedbank between fall no-till and tilled cover crop treatments. For example, the germinable seedbank of common lambsquarters, measured in the spring following either no-till or tilled cover cropping, was 11,000 and 13,000 germinable seeds per square meter, respectively. The impact of the weed-free control was dramatic: 600 germinable common lambsquarters seeds per square meter. While it may not be surprising that eliminating weed seed rain causes a reduction in the germinable seedbank, the magnitude of this single-season effect is noteworthy. It demonstrates that, even for a species with complex seed dormancy such as common lambsquarters, benefits of seebank management practices may be realized within a short time frame.

References Gallandt, E. R. (2006). How can we target the weed seedbank? Weed Science 54, 588-596.


Performance of a min-till rotary hoe in field pea (Pisum sativum L.)

E. N. Johnson1 and S.J. Shirtliffe2

1Agriculture and Agri-Food Canada (AAFC), Box 10, Scott, SK, Canada S0K 4A0 email: [email protected]; 2Department of Plant Sciences, College of Agriculture and Bioresources,

University of Saskatchewan, Saskatoon, SK, Canada.

Organic farming is growing in Western Canada. In 2005, there were 530,000 hectares of land in certified organic production in Canada, 85% of which is in the Canadian Prairies. Excessive soil tillage for weed control is undesirable in the semi-arid Prairie region. Maintenance of crop residues on the soil surface is necessary to reduce the potential for wind and water erosion. Tillage tools that can provide some level of weed control but maintain surface residues are badly needed. Two experiments were conducted at the Scott Research Farm to investigate the performance of a min-till rotary hoe in field pea (Pisum sativum L.). The first experiment was conducted in 2004 to 2006 and investigated the tolerance of field pea to rotary hoeing at various growth stages, as well as the effect of hoeing on crop residues. Up to six passes of the rotary hoe were conducted at pre-emergence, ground-crack, 5-node, 8-node, and the 11-node stage of field pea. The second experiment was conducted in 2007 and 2008 to evaluate the weed control efficacy of rotary hoeing. Single, double, and triple passes with the hoe were conducted sequentially at the ground-crack, 3-node, and 5-node stage of field pea. In the first experiment, surface residues were maintained on the soil surface even with 6 passes of the rotary hoe. Field peas exhibited the ability to tolerate rotary hoeing based on plant counts conducted 3 to 4 weeks after harrowing. There was less plant mortality when rotary hoeing was conducted at later crop stages, indicating a higher ability to resist injury at later growth stages. The number of passes had a significant effect on field pea yield which was independent of crop stage. This indicates that the crop’s ability to recover from post-emergence harrowing was similar at all growth stages. Under weed-free conditions, field pea yields declined by 5 to 6% after 2 passes were conducted. In the second experiment, the lowest weed density and biomass and highest field pea yield were recorded when 6 passes were conducted (double pass at ground crack, 3- and 5-node or triple pass at ground crack and 3-node stage); however, it was only marginally better than a 4-pass system (double pass at ground-crack and 3-node stage). Considering the large size of farms in Western Canada, it may not be practical to conduct 4 to 6 passes in order to optimize weed control with a rotary hoe. Future studies will investigate the use of more efficacious implements at the pre-emergence stage (stage of high selectivity) followed by post-emergence rotary hoeing to determine if fewer passes are required for efficacious weed control.


Current weed management and the problem of the highly adaptive, cosmopolitan weeds

L. Radics1 and M. Glemnitz² 1Corvinus University Budapest (BCE), Department of Ecological and Sustainable Production

Systems, Budapest , Hungary, Email: [email protected] ² Leibniz-Centre for Agricultural Landscape Research (ZALF), Institute of Land Use Systems,

Müncheberg, Germany, Email: [email protected]

Weed management made great progress within the last three decades. The technological development in weed management is regarded as one key factor for yield increases, e.g. of 20-25 % in winter wheat, in the last 20 years in Germany (Farack 2008). Modern weed management prepared the way for simplifications in crop rotations, for the spread of reduced soil tillage systems and for yield increases in organic farming. One main trend in weed management was the search for simple, single solutions with low manpower requirement and high share of provisional management.

As a result of this development weed flora changed. Not only the average weed coverage has been reduced, but also the weed flora diversity. Data from field and expert surveys within European networks show clearly that the weed flora is dominated by a quite small number of highly frequent and abundant species. Most of them are highly flexible in their ecological behavior, can re-grow or re-germinate after weed control treatments and have a high dynamic. Moreover the invasion of some new species becomes economically important.

The problem of the highly flexible, well adapted noxious and invasive weed species require modifications in the future weed management strategies. From many research projects on this topic it is known, that the control of flexible and invasive species requires the application of multiple tools to be successful. This need is contrary to attempts of further simplification in weed control and can not be met without changes in weed management strategy especially for organic agriculture but not only for them. The physical, or more in general the non-chemical weed management should reflect the changes in weed flora composition and develop new strategies starting from an expanded base of management tools. Maintaining biodiversity under modern agricultural practice can only be achieved if new strategies will be found, which are more selective, targeting the dominant frequent species in particular and taking ecological spillover effects into account. Many is already known about the effects of particular measures on single weed species. This knowledge needs to be systemized before ist get lost, deepened through research and supplemented with new solutions as well as specific solutions for particular problems („curing systems”). We call this complex system agro-ecological weed management.

Fernandez-Quintanilla et al. (2008) provided a paper spotlighting the need for new orientation in weed management and potential future paths in weed research. This paper is a breakthrough regarding current concepts. With our paper we want to contribute to this discussion with some supplementary comments especially for organic farming in Europe.

References Farack M (2008) Entwicklungstendenzen und Faktoren der Ertragsbildung bei Getreide in Thüringen. Thüringer

Landesanstalt für Landwirtschaft, Themenblatt-Nr.: 51.19.600. Fernandez-Quintanilla C, Quadranti M, Kudsk P and Báberi P (2008) Which future for weed science?. Weed Research

48(4), 297-301.


Physical weed control in protected leaf-beet in Central Italy

M. Raffaelli1, M. Fontanelli1, C. Frasconi1, L. Lulli 2, M. Ginanni2, F. Sorelli, A. Peruzzi1 1 MAMA - DAGA, University of Pisa. Via del Borghetto 80, 56124 Pisa, Italy

mail address: [email protected] 2 CIRAA “Enrico Avanzi”, University of Pisa.

Abstract In Central Italy leaf-beat is a typical and very important protected cultivation. In leaf-beet

protected cultivation weed control is one of the most important problems, because of it’s quite long crop cycle (about 4-5 months).

The aim of this research was to set up an efficient non-chemical weed control strategy performed with innovative machines built and set up by the University of Pisa.

A two-year (2006-2007) “on-farm” experimental trials were carried out in Crespina (PI). A conventional weed management technique (consisting in one pre-transplanting chemical treatment) was compared to an innovative physical weed control strategy (consisting in stale seedbed technique, in some post emergence precision hoeing and in-row hand-weeding treatments). In the conventional technique the leaf-beat was manual transplanted, while in the innovative strategy it was sowed with a precision pneumatic planter. All the innovative machines for physical weed control were adjusted and set up for the protected cultivation. In the two year trials similar yields were recorded for the two systems in comparison. Total labour time (for weed management and crop planting) was appreciably lower in the conventional system in the first year of experiment (-67%), while, in the second year, some improvement in the innovative technique allowed to reach lower values with respect to the conventional technique (-40%). Weed dry biomass at harvest was significantly lower for innovative system (on average -50%).

Introduction Recently, integrated and organic vegetable production systems has gained a great deal of

attention in agreement with EU agricultural policy reorientations, furthermore, this is in line with mounting public concern for environmental issue, workers safety and the growing consumer demand for high quality food products (Peruzzi et al., 2007). One of the major technical problems that arise in vegetable cropping when decreasing use of agrochemicals is weed control (Bàrberi, 2002). This is a very important problem in protected cultivation in which an inevitable intensification of cultivation involves even more difficulties. Protected cultivation has many commercial advantages but it has many agronomic and crop protection problems, including weed control. This problem, very important for horticultural crops, can be tackled and solved in sustainable way using and optimising physical weed control.

A series of techniques and purpose-designed operative machines have been devised to perform efficient and economically viable non chemical weed control in the open field. Numerous interesting trials have been carried out with promising results on spring-summer crops (Raffaelli et al., 2004 e 2005), on winter cereals (Bàrberi et al., 2000; Rasmussen, 2004) and on horticultural crops (Peruzzi et al., 2007).

In contrast research on physical weed control in protected cultivations was not developed to the same extent. For this reason, technical and scientific knowledge available on this topic is lacking. Moreover, the specifically designed techniques and operative machines are not usable on different crops and operative conditions. Therefore, to obtain a sustainable weed control in protected cultivation it is needed to study with attention the problem, to have specific machines devised for the different operative conditions and to analyse into deep the interactions among operative parameters (crop typology and management practices, as well as weed density, developmental stage


and competitiveness, soil conditions, protection typology, etc.) (Bàrberi et al., 2000; Peruzzi et al., 2007; Vanhala et al., 2004).

With the aim to set up and evaluate strategies in order to reduce or eliminate herbicide use in protected cultivation an experiment was carried out on leaf-beat (Beta vulgaris L. var. cycla (L.) Ulrich). In Central Italy leaf-beat is a typical and very important protected cultivation and for this crop weed control is one of the most important problems, because of it’s quite long crop cycle (about 4-5 months). In the experimental trials a conventional weed management technique was compared to innovative physical weed control strategies.

Materials and methods A two-year (2006-2007) “on-farm” experimental trials were carried out in Crespina (PI). Two

different farms, conventional and organic, with similar sandy soil (sand 67%, silt 26%, clay 7% and organic matter 2%) and climatic conditions were involved.

In organic farm sowing was performed with a precision planter, in August at a seeding rate of 30 seeds m-2 (20 X 12 cm), on ridges 1.4 m wide (with 5 rows ridge-1). In conventional farm leaf-beet was sowed in seed-bed in August and after was manual transplanted in tunnel at the end of September at a rate of 12 plant m-2, on ridges 1 m wide (with 3 rows ridge-1). First year

In the conventional strategy weed control was carried out with transplanting technique and with one pre-transplanting chemical treatment (8 kg ha-1 Kerb, a. i. propizamide).

In the non chemical strategy weed control was carried out with false seed bed technique (by means rolling harrow), three post-emergence precision hoeing and two in-row hand-weeding treatments performed between hoeing treatments. The machines used for physical weed control were studied, built and set up by Research Unit involved in this trial, over the years, to perform effective and efficient treatments.

The rolling harrow was realized to perform a very shallow tillage and an efficient weed control both in “false seedbed” technique and in precision hoeing treatments in post-emergence of the crop. The machine is modular, so it can be built with different working widths adapted to operative conditions (Fig. 1). Apart from working width, the rolling harrow is structured on a square draw piece frame bearing working tools and three points linkage. The tools are spike discs with diameter of 30-35 cm (placed in the front) and gage rolls with diameter of 27-33 cm (placed in the rear), that are inserted in two axles connected one another by means of a chain drive with a ratio easily adjustable. The discs and the rolls are placed differently on the axles and changed with elements of different sizes with a very simple blocking system. The discs and the rolls can be placed differently on the axles (Fig. 1): close arrangement in order to perform a very shallow tillage (till 3-4 cm) of the whole treated area for seed-bed preparation and non-selective mechanical weed control after false sowing and with spaced arrangement in order to perform efficient selective mechanical weed control treatments in post-emergence for precision inter row weeding. In precision weeding it is possible to work on very different inter row distances from a minimum value of 15 cm. The action of the rolling harrow is characterized by the passage of the spike discs that till the soil at 3-4 cm of depth followed by the passage of the gage rolls that work at high peripheral speed as the rear axle is powered by the front axle by means of an overdrive tilling and crumbling the soil till a depth of 1-2 cm. The rolling harrow can be equipped with couples of elastic tines (working as both vibrating teeth and torsion weeders) in order to perform a mechanical weed control also in the rows. For precision weeding a version of rolling harrow with a steering handle system was set up. In these trials a machine 1,4 m wide was used.


Figure 1. a) Scheme of the rolling harrow: (A) frame; (B) front axle with spike discs; (C) rear axle with cage rolls; (D) chain drive; (E) three points hitch. b) Arrangement for treatments on the whole surface (left) and of hoeing (right).

The precision hoe utilized in this experimental trial is a machine 2 m wide (Fig. 2), designed to

perform selective weed control in the horticultural row crops with very low inter row distance (in this trial 20 cm). The precision hoe is structured on a square draw piece frame bearing working tools and three points linkage. The working tools can be 11 and each is placed on articulated parallelogram equipped with a small wheel for the working width adjustment. The machine was equipped with rigid elements bearing a 9 cm wide triangular horizontal blade and two kinds of elastic tines (torsion weeders and vibrating tines). The elastic tines are able to perform a selective weed control on the row crop. A back-seated operator can adjust the actual position of the working tools by operating a steering handle. This precision hoe is a very interesting innovation for the region trial because it is able to work on 5 rows on a “standard” ridge 1,4 m wide.

Figure 2. a) Scheme of precision hoe: A) hoe operator seat; B) steering handle; C) steering wheel D) articulated parallelogram; E) working tool; F) lateral disc; G) support wheel; H) elastic tines. b) Vibrating tines (left) and torsion weeders (right).

In conventional farm the machine utilized for chemical treatments was a sprayer of firm Projet srl, model sprayer mix, with a tank capacity of 300 dm3. The treatments was performed with an hand lance equipped with a turbulence full cone spray nozzle and with manual valve for flow adjustment. The hand lance was equipped with a tube 100 m long that is reeled by an on purpose manual tube reel.

In the trials data concerning soil (physical and mechanical characteristics), machine operative characteristics (working width, depth, speed, capacity, time and fuel consumption), weeds (density before and after each weed control treatments and dry biomass at hand weedings and harvest) and crop production (total fresh yield) were measured or calculated. Second year

In the conventional farm weed control was carried out with transplanting technique and with one pre-transplanting chemical treatment (3.5 kg ha-1 Kerb, a. i. phenmediphan) performed with the same machine of the first year.

In organic farm, in the light of the first year results, weed control was carried out with a modified strategy performed with false seed bed technique (by means rolling harrow), one pre-

a b

a b


emergence of the crop flaming, two post-emergence hoeing (the first with rolling harrow, the second with precision hoe) and one final in-row hand-weeding treatments. For physical weed control, besides the machines used in the first year, a flaming machine realized and set up in previous experiment by Research Unit was used.

The implement allows to perform both pre-emergence and post-emergence flame weeding (Fig. 3). The flaming machine can be equipped with 8 rod burners 25 cm wide that have a good flame shape and with 4 LPG tanks. Any couple of burners is placed on a control board and is connected to a 25 kg LPG tank on which a pressure regulator and a manometer are placed. The LPG tanks are placed inside a hopper which contains warm water, thus allowing good heat exchange. The exhausted gas of the tractor engine are used to heat the water by means of a flexible pipe connected to both the exhaust head and a copper tube placed inside the hopper. Any couple of burners is connected to an articulated parallelogram in order to maintain the set out adjustments (height and inclination with respect to soil surface) when the flamer is working. Any burner is also equipped with one valve, one safety tap and an electronic control system which allows the tractor driver to adjust the LPG feed (high or low levels) and to control if the burners work appropriately directly from his seat.

Figure 3. Scheme of the new flaming machine: (a) burner; (b) articulated parallelogram; (c) hopper containing water; (d) LPG tank; (e) shelf on which the inflow LPG control system is located; (f) control panel; (g) flexible pipe that pipes the exhausted gas of the tractor engine to the heat exchanger in the hopper; (h) heat exchanger.

During the trial assessments were the same of the first year with the addition of a final weed

sampling carried out by means of the Braun-Blanquet ordinal scale that is able to give good informations on weed canopy assessment, biodiversity and aggressiveness.

Results Operative characteristics

The operative characteristics of the machines used for physical weed control in the first year trial are presented in Table 1.

The rolling harrow, utilized only for pre sowing treatment, was used with high speed (about 6 km h-1) and therefore its working time was low (1.47 h ha-1). The precision hoe instead, given the more “gentleness” of intervention, was utilized on average with a working speed of 1.2 km h-1 and consequently the working time for each treatment was higher to 6 h ha-1. The working depth was for all treatments lower to 4 cm in order to avoid soil disturbance, that could cause a new high level of infestation. Fuel consumption was about 3 kg ha-1 for false seed bed treatment and on average 13 kg ha-1 for each hoeing.

Total working time for physical weed control was 283.96 h ha-1, (20.91 h ha-1 for physical weed control, 4.34 h ha-1 for sowing and 258.71 h ha-1for hand weeding) that were necessary largely for two expensive treatments of hand weeding that required respectively more than 80 h ha-1 the first and more than 170 h ha-1 the second.


Table 1. Performances of the machines used for physical weed control in 2006. Characteristics Har Hoe 1 Hoe 2 Hoe 3 Working depth cm 3.6 2.6 2.7 2.8 Working speed km h-1 5.9 1.2 1.1 1.2 Working productivity ha h-1 0.68 0.16 0.15 0.16 Working time h ha-1 1.47 6.21 6.87 6.36 Operators 1 2 2 2 Fuel consumption kg ha-1 2.9 12.4 13.7 12.7 Har=harrowing, hoe=hoeing (1, 2, 3 first, second or third pass)

In conventional strategy for weed control the time for manual transplanting (that can be considered a technique giving an advantage to the crop with respect to weeds) was 84.51 h ha-1, while working time for chemical treatment was 7.81 h ha-1 (total time was 92.32 h ha-1)

In the first year trial, in all, physical weed control strategy needed a total labour employed very higher than in conventional strategy (284 h ha-1 vs 92 h ha-1).

The operative characteristics of the machines used for physical weed control in the second year trial are presented in Table 2. Table 2. Performances of the machines used for physical weed control in 2007. Characteristics Har Fla Har Hoe Working depth cm 3.5 - 2.7 2.8 Working speed km h-1 6.1 3.5 1.7 3.0 Working productivity ha h-1 0.73 0.42 0.22 0.37Working time h ha-1 1.36 2.39 4.47 2.71Operators 1 1 1 2 Fuel consumption kg ha-1 2.7 4.8 8.9 5.4 Har=harrowing, Fla=flaming, hoe=hoeing

For false seed bed treatment, the rolling harrow worked at very high speed and as consequence its working time was very low 1.36 h ha-1. In the first hoeing treatment, for leaf beat small size, the rolling harrow was used with slow speed with consequent increasing in operative time (4.47 h ha-1). This parameter, however, was lower than that recorded in previous year for precision hoeing; this result was possible for a better overall strategy in which a flaming treatment in pre emergence of the crop was carried out. The flaming treatment was performed with forward speed of 3.5 km h-1 (2.39 h ha-1) and working pressure of 0.25 MPa with a LPG consumption of roughly 40 kg ha-1. Total working time of the last hoeing too, carried out with precision hoe with static tools, was very lower than that recorded for the hoeings of the previous year. Together with working time reduction, fuel consumptions reduced decidedly; total fuel consumption was about half of first year. The working time reduction trend was even more evident, in consequence of higher values, for the final hand weeding (37 h ha-1); this time was nearly seven times lower than first year.

In the second year the set up of strategy of physical weed control reduced heavily labour employed for an hard improvement of performances of all treatments, but specially for the reduction of hand weeding working times.

In conventional strategy for weed control the time for manual transplanting was nearly the same as that the previous year , while the time for chemical treatment was slightly lower.

For physical weed control in organic farm total working time was 52.31 h ha-1 (10.93 h ha-1 for physical weed control, 4.34 h ha-1 for sowing and 37.04 h ha-1for final hand weeding), while in conventional farm it was 89.86 h ha-1 (83.56 h ha-1 for transplanting and 6.30 h ha-1 for spraying).


Weed control and yield

In the first year trial physical weed control allowed a progressive depletion of seed-bank in the first centimetres of soil layer. Weed flora was at first composed by Picris echioides L. (30% of relative density), Veronica persica Poiret (20%), Rumex spp. L. (20%) and winter annual and perennial grasses (17%). Weed density value was 350 plants m-2 before stale-seedbed technique realization, 250 plants m-2 before first hoeing intervention, 100 plants m-2 before second hoeing intervention and 120 plants m-2 before third hoeing pass. Weed control efficiency was 100% for rolling harrow intervention and over 90% for hoeing passes (taking into account in-row and inter-row space). Weed dry biomass value registered during the second hand-weeding intervention was triple with respect to each sampled during the first one (12 vs 4 g m-2).

Weed density registered in the conventional farm before chemical treatment was about 150 plants m-2 and the most relevant species was Stellaria media (L.) Vill. (over 90% of relative density).

Weed dry biomass registered before the last crop leaf harvesting was 6 g m-2 and 13 g m-2 for the organic and the conventional farm respectively (Table 7). In this case, the most relevant species observed were Rumex spp. (13% of relative density), P. echioides (39%), Conyza canadiensis (L.) cronq. (19%), V. persica (7%), Anagallis arvensis L. (6%), Cerastium holosteoides Fries. ampl. Hylander (6%) for the organic farm and only S. media (almost 100% of relative density).

The two cropping systems didn’t show significant differences, at the end of the first year of the experimental trial, concerning with total fresh yield (Table 3). However, conventional system yield was slightly higher with respect to the organic one (on average about 37 Mg ha-1 vs 33 Mg ha-1). Table 3. Yield and weed biomass at harvest determined in 2006. Weed management system Yield

(Mg ha-1) Weed dry biomass

(g m-2) Conventional system Organic system

36.9 ns 33.4 ns

12.8 a 5.9 b

Different letters on the same column mean significant differences for p< 0,05 (LSD test)

Concerning with the second year of organic cropping system, weed density observed before stale-seedbed technique treatment was about 400 plants m-2. Weed flora was mainly composed by Solanum nigrum L. (52% of relative density), P. echioides (22%), C. canadiensis (22%) and Portulaca oleracea L. (8%). However the rolling harrow intervention, carried out before crop sowing, was characterized by a total weed control effectiveness. Furthermore, very few weeds re-grew from this treatment to crop emergence (about 10 plants m-2). Pre-emergence flaming treatment was carried out just in order to control few weed species (for example Cyperus spp.) that were fairly developed (4-6 true leaves) so that they could compromise crop emergence. Before the first hoeing intervention, carried out by means of the hoe conformed rolling harrow, weed density was about 200 plants m-2. This treatment controlled about the 90% of weed in the inter-row space and the 30% in the in-row space. The second hoeing treatment, carried out by means of the precision hoe, was characterized by similar levels of weed presence reduction. Weed density before this intervention was about 100 plants m-2. Moreover, one hand weeding intervention was carried out in order to reduce weed presence in the intra-row space. Weed dry biomass in that phase was about 4 g m-2 and Amaranthus retroflexus and Chenopodium album were the most widespread and developed weeds.

In the conventional farm, weed density before herbicide application was about 180 plants m-2, and Stellaria media was almost the only species emerging in the field.

Weed dry biomass and weed canopy data collected before the last crop leaf harvest showed significant differences between the two cropping systems. Organic plots were characterized by a significantly lower weed biomass (-50%) and weed canopy percentage (-65%) values with respect


to the conventional ones (Table 4). Concerning with weed canopy assessments, carried out by means of the Braun-Blanquet ordinal scale, 16 different weed species were observed in the organic farm and only three on the conventional plots. This probably means that chemical weed management could more easily bring to a sensible weed selection action with respect to organic weed management. Moreover, a strictly selected weed flora could be very aggressive. In this case the most widespread species was Stellaria media for both the cropping system. Its canopy percentage value was about 87% for the conventional farm and 57% for the organic one. Moreover, other two weed species reached relevant relative percentage values before the last harvest in the organic cropping system: Conyza canadensis and Chenopodium album (about 40% of relative density together).

Table 4. Yield, weed biomass and canopy at harvest determined in 2007. Weed management system Yield

(Mg ha-1) Weed dry biomass

(g m-2) Weed canopy

(%) Conventional system 30.6 ns 7.8 a 12.9 a Organic system 30.8 ns 3.5 b 4.4 b Different letters on the same column mean significant differences for p< 0,05 (LSD test)

Concerning with total crop fresh yield, no significant differences were registered during the second year too. The observed value was about 31 Mg ha-1 for both the cropping system and it was similar, even if slightly lower, with respect to the one observed in 2006 (Table 4).

Conclusions These experimental trials show that the physical weed control strategy set up in the two year

trials allowed an efficient cultivation of leaf-beat in protected cultivation. In particular in the first year the yield of the two compared cropping system was very similar but physical weed control strategy needed a total labour employed very higher than in conventional.

In the second year physical weed control strategy was improved and its working times reduced positively; in the two compared cropping system practically equal yields with similar working times were obtained. In a global evaluation of the different cropping systems must also take into account that physical weed control strategy, allows to obtain a produce with higher quality and price (on average in the two years the price on Central Italy market was 1.5 € kg-1 and 0.5 € kg-1 for organic and conventional leaf-beat respectively).

References BÀRBERI P, SILVESTRI N, PERUZZI A & RAFFAELLI M (2000) Finger-harrowing of durum wheat under different tillage

systems. Biological Agriculture and Horticulture 17, 285-303. PERUZZI A, GINANNI M, FONTANELLI M, RAFFAELLI M & BÀRBERI P (2007) Innovative strategies for on-farm weed

management in organic carrot. Renewable Agriculture and Food Systems 22 (4), 246-259. RAFFAELLI M, BÀRBERI P, PERUZZI A & GINANNI M (2004) Options for mechanical weed control in string bean.

Agricoltura Mediterranea 134 (2), 92-100. RAFFAELLI M, BÀRBERI P, PERUZZI A & GINANNI M (2005) Mechanical weed control in maize: evaluation of weed

harrowing and hoeing system. Agricoltura Mediterranea 135 (1), 33-43 RASMUSSEN J (2004) The effect of sowing date, stale seedbed, row width and mechanical weed control on weeds and

yields of organic winter wheat. Weed Research 44, 12-30. VANHALA P, KURSTJENS D, ASCARD J, BERTRAM A, CLOUTIER D, MEAD A, RAFFAELLI M & RASMUSSEN J (2004)

Guidelines for physical weed control research: flame weeding, weed harrowing and intra-row cultivation. 6th EWRS Workshop on Physical and Cultural Weed Control, Lillehammer, 8-10 March, 194-225.


Sensor based selective weed harrowing in cereals in Germany

Victor Rueda-Ayala and Roland Gerhards (Department of Weed Science (360b), University of Hohenheim, 70599 Stuttgart, Germany.

Email:[email protected], Tel.: 0711 - 459-23165, Fax.: 0711 - 459-22408)

Abstract Three field experiments were installed to investigate whether intensity, timing, and direction of

post-emergence weed harrowing in winter and spring cereals influenced the selectivity. Bi-spectral cameras and a soil resistance sensor were used to work out algorithms for site-specific weed control and real-time automatic set up of harrowing aggressivity. Selectivity is the relationship between weed control (\%) and crop soil cover (\%), as originally defined in Denmark. Each experiment was designed to create various intensities by increasing number of passes, angle tine and driving speed, applied at varying crop growth stages (BBCH). Objective estimation of leaf cover through differential image analysis was used. A recently proposed statistical procedure was used to describe the effects. Selectivity was in general influenced by timing of harrowing, except in wheat. It was higher at later crop growth stages in summer barley and lower for winter barley. No clear relationship between leaf cover and weed density at increasing harrowing intensities could be determined. That caused an increment of crop soil cover, although not always improving weed control. Harrowing across crop rows did not cause impacts on selectivity while along rows seemed to improve it at early growth stages, yet more research is needed to prove the results. Crop tolerance to soil covering was higher for wheat than for barley. Harrowing at low weed density might cause yield reduction. Intensities which generate the crop soil cover percent associated with the higher selectivity will be taken as the basis to develop algorithms to automatic adjustment of harrowing intensities.


Comparison of three tillage intensities on grass weed occurrence in cereal rotation

J. Salonen MTT Agrifood Research Finland, FI-31600 Jokioinen, Finland. Email: [email protected]

Direct drilling has become more common in Finland, particularly in cereal cropping. Currently, some 20-25% of cereal area is sown, either in autumn or in spring, with direct drilling system, in most cases without any soil cultivation. Such a fundamental change from traditional autumn ploughing and seed-bed preparation with shallow harrowing evidently affects the composition of weed flora.

A field study was carried out in Jokioinen in 2005-2007 with the aim to compare the impact of tillage intensity on the composition of weed flora and particularly on the establishment of grass weeds. In practice, Elymus repens is the main grass weed species in terms of frequency, abundance and economic importance in Finland. Otherwise, the Top-10 weed flora in cereal fields consists of broad-leaved species.

Three tillage systems were compared in clay soil: 1) conventional autumn ploughing at the standard depth of 20-25 cm, 2) stubble cultivation at the depth of 10-15 cm in the autumn and 3) direct drilling without any soil “disturbance”. Elymus repens (both seeds and rhizome pieces), Poa annua and Phleum pratense were sown at soil surface as “model weeds” in small quadrats in each experimental plot just before the primary tillage in the autumn 2005 or 2006. Before sowing, the ploughed and stubble cultivated plots were harrowed. The crop rotation in 2005-2007 was spring barley – winter wheat – spring wheat. The results from the spring wheat in 2007 are presented here.

In general, a rapid infestation of grass weeds was demonstrated in the plots with direct drilling. Not only the sown model weeds but some grass species, particularly Poa pratensis, from natural seed bank thrived in direct-drilled plots. On the contrary, hardly any grass weeds infested the ploughed plots. In regard to model weeds, Elymus repens sprouting from rhizomes succeeded best to establish new plants followed by Phleum pratense and E. repens from seeds. The establishment of Poa annua failed almost completely in all tillage options.

The proportion between grass weeds and broad-leaved weeds differed in three tillage systems. The weed biomass in direct-drilled plots was produced almost completely by grass weeds. On the other hand, Fumaria officinalis was the most abundant species in ploughed plots and Taraxacum officinale in stubble cultivated plots of spring wheat.

As no direct grass weed control was included in the trial protocol, grass weeds caused heavy yield reductions particularly in direct drilling. In practice, most of the farmers stick to annual use of glyphosate as a curative weed control in direct drilling. For the moment (2007-2009), a national survey of weeds in spring cereal fields is carried out by the MTT providing some new information about the impact of direct-drilling on weed flora at farm level.


Evaluation of finger and torsion weeders for cultivating cool season vegetables in Salinas, CA, USA

R.F. Smith and M. Silva Ruiz University of California Cooperative Extension, 1432 Abbott St., Salinas, CA 93901, USA Email:

[email protected].

Standard cultivation of lettuce and cole crops cultivates >80% of the bed but typically leaves a

3 to 5-inch wide uncultivated band around the seedline of the crop. The majority of hand labor expended in removing weeds from vegetables occurs in the seedline. Finger and torsion weeders are designed to remove weeds in the seedline on transplanted crops. These tools take advantage of the larger size of the transplanted crop and the smaller size of recently germinated weeds. They can be adjusted to aggressively cultivate the seedline and impact the crop plants or adjusted farther apart to avoid impacting crop plants. In addition, finger weeders from the Kress Company, Germany come in different levels of hardness which allow flexibility in selecting the right implement for the crop being cultivated. Torsion weeders from Frato Corp, Netherlands come in two diameters which also allows for flexibility in selecting the right tool. These tools require careful adjustment and have the reputation of being too slow for large scale vegetable production that is practiced in Salinas and therefore are not currently used in this region. Trials were conducted in 2007 and 2008 to evaluate 24.1 and 35.5 cm in diameter orange (softer) and yellow (harder) finger weeders (Kress Co.) and 7 and 9-mm torsion weeders (Frato Co.). Impact on weed removal, weeding time and yield of transplanted lettuce and leeks were evaluated.

Trials were conducted by mounting the cultivators on the back of the grower’s standard cultivators to supplement the cultivation operation. In the lettuce trial, standard cultivation removed 51% of total weeds while all finger and torsion weeders removed 74 to 87% of total weeds. Finger and torsion weeders removed a higher proportion Urtica urens but not Malva parviflora. Weeding time was significantly lower in the orange and yellow 24.1 cm finger weeders. Yield was significantly reduced in the 9-mm torsion weeder treatment. In the leek trial standard cultivation removed 28% of the weeds while finger weeders removed 73-88% and reduced hand weeding times 36 to 51%. All cultivations were conducted at standard cultivation speeds of 1.2 to 1.8 km/hr indicating that these tools have the potential for being added to standard cultivation rigs without sacrificing speed of cultivation operations. These tools are most appropriate for transplanted crops and work best if the crop is firmly rooted in the soil and the weeds are small (eg cotyledon to first true leaf stage).


Using a spring-tine harrow to weed a navy bean crop (Phaseolus vulgaris var. Tabella Brisa) destined for human consumption.

Taberner A.1, Llenes JM.1, Roque A. 1, Orri J2. and Miralles M.2 1Servei Sanitat Vegetal Malherbologia Rovira Roure 191, 2598 Lleida. 2Associació de Cultivadors de Fesols de Santa Pau, 17811 Santa Pau.

[email protected]

The majority of bean crops cultivated in Catalonia are marketed under a Quality Label. This assures quality for the consumer and a high price for the farmer.

One of these denominations of origin is the Tabella Brisa navy bean from Santa Pau (Girona). The specific characteristics of the local soil, which is volcanic in origin and of the local climate combine to produce a high-value culinary bean. Weed control is one of the main problems affecting this crop. Given the very specific culture conditions, growers prefer not to use herbicides. In the early stages of the crop growth, weeding is conducted between rows, but at later stages of crop development, it is necessary to weed more thoroughly and within rows, using more tools in order to ensure that the crop remains free from weeds until the end of its growth cycle. To date, little work has been done on the mechanical control of weeds in navy beans, although the works cited in the references were useful for checking and confirming the results shown in the poster.

In 2008, a demonstration was conducted with farmers to show the efficacy of controlling weeds in a bean field in addition to applying a pre-emergence herbicide. Three flex tine harrow adjustments were used in a single pass treatment.

The results obtained were satisfactory for controlling dicotyledonous weeds, despite the fact that the field was not excessively infested, while no significant differences were found between the 3 settings of the machine in terms of their effectiveness in removing weeds. However, in terms of crop selection, we did observe different effects on bean plants, which tended to be flattened when subjected to the most aggressive settings. This tendency was also observed by Raffaelli et al. (2002).

There were no differences in bean yields between different configurations or between treated and untreated areas. The same findings were observed by Raffaelli et al. (2002) and Vangessel M.J. et al. (1995) in their respective works, despite the fact that they used the machinery at much earlier phenological stages of crop development.

We can therefore conclude that the flex tine harrow can be a useful tool for controlling weeds in beans crop, although more trials are needed to see the results for more weed infested areas and areas in which the flex tine harrow was used as the main method of weed control with multiple (3 or 4) passes, in order to observe the effects of further use on both weeds and crop performance.

The mechanical control of weeds in bean plantations with a flex tine harrow affects the straightness of the plants, making them bend in the direction in which the machine passes. Even so, there is no sign that the effects of this treatment have any significant effect on crop yield.

References Raffaelli, M.; Barberi, P.; Peruzzi, A.; Ginanni, M. (2002) Options for mechanical weed control in string bean – effects

on weeds. 5th EWRS Workshop on Physical Weed Control p.176-179 Vangessel, M.J.; Wiles, L.J.; Schweizer E.E.; Westra, P. (1995) Weed Control Efficacy and Pinto Bean ( Phaseolus

vulgaris) Tolerance to Early Season Mechanical Weeding Weed Technology, 9-3: 531-534.


Use of the flex-tine harrow for grain production in an Amaranth crop.

Taberner A1., Zamora N.2, Llenes JM1 and Roque A.1 1Servicio Sanidad Vegetal – Malherbología. 2Univeridad de Lleida. Avda. Alcalde Rovira Roure

191; 25198 Lleida. [email protected]

The genus Amaranthus consists of about 60 species, most of which are considered weeds in summer crops. However, a few species are grown for consumption, either for their young leaves - as in the case of A. tricolor L. A. mantegazzianus L. (Troiani, 2003), A. dubius, A. lividus, and A. cruentus (Stallknecht and Schulz-Schaeffer, 1993) - or for grain - Amaranthus cruentus L., A. hypochondriacus L. (Troiani, 2003) and A. caudatus L. (Stallknecht and Schulz-Schaeffer, 1993). In the Project "Amaranth: Food of the Future", developed within the Sixth European Framework, the immediate objective is to provide tools for a broad and sustained exploitation of amaranth. To do this, from an agronomic point of view, it will be necessary to identify the best practices for its cultivation, one of the most important of which will be weed control. No herbicides have been authorised for this work, so it must be done either by hand or mechanically. On the other hand, having reached a height of at least 50 cm, Amaranth is a crop that competes well against weeds. It is therefore essential to keep plantations clear of weeds in the early stages of development, but this is not so important at later stages of development.

This poster summarizes results showing the efficiency and selectivity of using a flexible-tine harrow to clear A. cruentus cv. Don Guiem, A. hypochondriacus var Nutrisol morfotype Azteca and A. Mantegazzianus cv. Don Juan weed varieties from a demonstration plot.

The demonstration took place at Almacellas (Lleida) on 28/07/2008: 24 days after sowing. It involved amaranth plants at the four to six leaf phenological stage. The size of each plot was 2 x 20 m and there were two 2 x 10 m. replications. There was no rainfall either before or after passing the harrow and the sprinkler irrigation system remained inactive throughout the week in which the trial was carried out. The soil textural class was loam. We used an Einbock machine, set with an angle of 90 degrees between the flex tine and the soil. The tractor speed was 4 km/h. The main weeds present on the plots were: Setaria verticillata, Capsella bursa-pastoris, Veronica spp. Anagalis arvensis, Echinochloa crus-galli.

We made a visual assessment on 06/08/2008. Using the machine at this phenological stage produced crop deformations without any notable effects on crop yield. We therefore concluded that mechanical weeding could be recommended in earlier stages of growth. These results show that both efficiency and selectivity were good and that the flex-tine harrow could be a suitable tool for carrying out this type of control. Combining a stale seed bed, using a flex-tine harrow during the early stages of development and applying inter-row weeding control practices seems to offer a reliable of keeping this crop free from weeds.


Thermal weed control


Thermal weed control – a review of current techniques

J. Ascard Swedish Board of Agriculture, Box 12, SE-230 53 Alnarp, Sweden.

Email: [email protected]

Thermal weed control methods have been developed as alternatives to chemical and mechanical weed control. These include flaming, infrared radiation, hot water, steam, electrical energy, microwave radiation, ultraviolet radiation, laser, and freezing temperatures. Mainly flame weeding has been used commercially, and to some extent infrared radiation, steam, hot water and electrocution. Solarization is also used for weed control but is reviewed in another paper in this workshop. Flaming and several other thermal methods kill plants mainly by rupturing cells, which leads to tissue desiccation. Flaming provides rapid weed control on a variety of weed species, but has relatively low selectivity and does not give residual weed control. The heat treatment dose, in terms of energy input per unit area, must be adjusted to take into consideration weed species and growth stage. The main use of flame weeding in Europe today is in organic production when mechanical methods are less effective. Flaming is used as an integral part of a weed management strategy, typically as a pre-emergence treatment in carrots and other slow-germinating vegetable crops. Flaming is also used after crop emergence in e.g. maize, onions and some other heat-tolerant crops and for potato haulm desiccation. Band-steaming of weeds are also used in organic vegetable crops. To some extent, flaming, hot water and steam are used for weed control in urban areas.

Several thermal methods use much fossil energy and release combustion by-products, but life cycle analyses have also shown environmental benefits in terms of impacts on crop, soil and water compared to mechanical and chemical methods. Thermal weed control options are attractive because they provide rapid weed control and do not leave chemical residues in the crop, soil or water. However, several thermal methods have high equipment and fuel costs and slow treatment speeds. A further technology development is essential for the greater adoption of many of these thermal methods.

References Ascard J, Hatcher PE, Melander B and Upadhyaya MK 2007. Thermal weed control. In Upadhyaya MK and Blackshaw

RE (eds). Non-chemical weed management. Principles, concepts and technology. CAB International, Oxfordshire. pp.155-175.

Ascard J, Van der Weide RY, Cloutier D & Leblanc ML 2007. Thermal weed control with a focus on flame weeding. Canadian Weed Science Society 61th Annual Meeting, November 27-29, Québec, Canada. (a refereed paper will be published)


Weed control with steam and solarization for field-grown flowers and strawberry (Fragaria ananassa L.)

Celeste A. Gilbert, Steven A. Fennimore*, Krishna Subbarao, Rachael Goodhue, J. Ben Weber and Jayesh B. Samtani.

University of California, Davis. USA

Abstract Soil solarization performance in coastal areas of California, USA is often inconsistent due to

fog and cool summer temperatures that do not allow soil to reach high temperatures required to reliably kill soil pests. Coastal California is also the principal strawberry fruit and cut flower production region, and solarization has displaced virtually no methyl bromide (MB) use in these crops. The objective of this research was to develop an economically-feasible combined solarization and steam heat, soil disinfestation system for field-grown cut flowers and strawberry. Field studies were conducted on the central coast of California during 2007 and 2008. For both strawberry and cut flower production, first year results show that steam treatments with and without solarization controlled weeds equal or better than MB and strawberry yields were comparable to MB.

Introduction Soil disinfestation with steam has a long and proven record for controlling soil borne weeds and

pathogens in greenhouse and nursery settings. However, high energy costs and lack of appropriate steam applicators has limited the use of steam in commercial fields. Injecting steam into finished planting beds with a mobile steam generator under clear polyethylene mulch, as used for solarization, would permit steam to supplement solarization as needed. The objective of this research was to develop an economically feasible combined solarization and steam heat, soil disinfestation system for field-grown cut flowers and strawberry.

Materials and Methods For both cut flower and strawberry production systems, raised beds were prepared as finished

seed beds with starter fertilizer and drip irrigation tape installed following standard production practices. Plot beds were 1.3 to 1.8 m wide by 6.1 m long.

Prior to treatment application, pathogen (Verticillium sp.) and weed seed samples were installed in the planting beds. Weed species included Stellaria media (L.) Vill. (common chickweed), Polygonum arenastrum Boreau. (common knotweed), Portulaca oleracea L. (common purslane), Malva parviflora L. (little mallow) and Cyperus esculentus L. (yellow nutsedge). Viability of the weed propagules was determined using tetrazolium tests and enclosed in permeable nylon mesh bags. Weed samples were installed at 5 and 15 cm depth from the soil surface (total of 4 samples per replication). Two of the weed samples were installed towards the edge of bed and remainder two towards bed centre. After completion of treatment application, seeds and pathogens were removed from the bags and their viability determined.

The design was a randomized complete block replicated six times. Treatments included 67% methyl bromide plus 33% chloropicrin (MBPic, 67:33) at 392 kg ha-1 applied by chemigation through the drip irrigation system, control (no MBPic, solarization or steam), solarization alone, steam alone, and steam plus solarization. The beds were irrigated to bring the soil moisture to sufficient levels for proper heat conduction and solarization, and then clear tarp was installed. For steam plus solarization treatments, beds were solarized for two weeks prior to steam application and for two weeks after application (total of four weeks). For the steam treatments, steam was injected for sufficient time to raise the temperature of the beds to 70°C for 30 min. After 30 min., steam disinfestation was discontinued. In 2007, all steam treatments were performed using a steam blanket Swartzinc. In the strawberry trial solarization was conducted Aug. 24 to Sept. 21, 2007 and steam


was applied Sept, 6 to 11, 2007. In the flower trial solarization was conducted Aug. 28 to Sept. 27, 2007 and steam was applied Sept 12 and 13, 2007.

During solarization and steam disinfestations process, temperatures were monitored continuously using temperature probes (Hobo, Pocasset, MA, USA) installed in beds at 5, 15, 30 and 45 cm depths. The amount of fuel, time and labour needed for treatment application were recorded.

Strawberry transplants and Zantedeschia aethiopica Spreng. (calla lily) bulbs for cut flower were planted at the end of the 4 week treatment period. Periodically through the growing season, weed density counts in each plot were recorded. Weeds were harvested from the field, and fresh biomass recorded. Hand weeding times were monitored periodically throughout the growing season. Strawberry fruit were harvested April to Sept. 2008, and flower bulb harvest is scheduled for July 2009. Data was subjected to analysis of variance at P = 0.05 and LSD’s were used for mean separation.

Results and discussion Solarization alone did not control weeds; hold down weeding times or kill weed propagules as

consistently as MBPic (Tables 1, 2, 3, 4). However steam with or without solarization resulted in weed control similar to MBPic. Temperature highs of 42 to 44°C in the solarization treatments did not control weeds as thoroughly as MBPic (Table 5). However, peak temperatures of >78°C in the steam treatments did kill weeds as well as MBPic. Strawberry yields in steam treatments were comparable to MBPic (Table 6).

Table 1. Weed control in strawberry beds at Salinas, CA USA.

Treatment ------------Weed biomass------------

Hand weed time

18 Oct. 07 24 Jan.08 22 Apr.08 22 Apr.08 --------------------grams/ 4.0 m-2-------------------- min./ 4.0 m-2 1. MBPic 0.4 c 337.9 b 1097.9 1.5 c 2. Control 639.2 a 1115.4 a 734.2 4.1 a 3. Solarization 375.0 b 280.0 b 1256.3 2.9 b 4. Steam alone 86.7 c 119.6 b 1487.5 1.2 c 5. Steam + solarization 0.0 c 229.6 b 540.8 1.3 c Treatment P 0.0001 0.0001 0.6040 0.0001

Table 2. Weed densities and hand weeding times in flower beds at Prunedale, CA USA.

Treatment ---------Weed densities------ Hand weed time 4 Feb. 08 21 Mar. 08 4 Feb. 08 21 Mar. 08 ----------- number (7.4 m-2) ------- ----------- time (min. 7.4 m-2) ----1. MBPic 16.8 ab 25.7 b 2.2 abc 2.9 bc 2. Control 24.0 a 92.0 a 3.4 a 6.7 a 3. Solarization 18.8 a 29.0 b 2.9 ab 3.6 b 4. Steam alone 8.8 bc 14.0 b 1.8 bc 2.0 c 5. Steam + solarization 6.0 c 13.8 b 1.5 c 1.8 c Treatment P 0.0018 0.0004 0.0168 0.0001


Table 3. Weed propagule survival in strawberry beds at Salinas, CA USA. Treatment P. oleracea M. parviflora S. media C. esculentus 1 ------------------------- viability (%) ----------------------------- 1. MBPic 9.0 b 37.3 b 0.0 c 8.3 c 2. Control 85.7 a 70.7 a 60.5 a 37.0 b 3. Solarization 82.0 a 39.5 b 49.2 b 55.0 a 4. Steam alone 0.0 b 3.8 c 0.0 c 11.7 c 5. Steam + solarization 2.8 b 6.2 c 0.0 c 0.0 c Treatment P <0.0001 <0.0001 <0.0001 <0.0001

1 Viability of tuber samples buried 5 cm deep in the bed centre Table 4. Weed propagule survival in flower beds at Prunedale, CA USA.

Treatment P. oleracea M. parviflora S. media ---------------------- viability (%) ---------------------------------- 1. MBPic 0.2 c 29.0 b 0.0 c 2. Control 82.6 a 69.8 a 49.4 a 3. Solarization 58.8 a 31.8 b 34.6 b 4. Steam alone 7.3 c 35.0 b 8.3 c 5. Steam + solarization 11.2 c 11.5 c 7.7 c Treatment P <0.0001 <0.0001 <0.0001

Table 5. Treatment effects on 5 cm soil temperatures over a 4 week period.

Treatment High temp. a Average temp. Time > 40 °C strawberry flower strawberry flower strawberry Flower ----------------------- temperature (°C) ------------ ------------ hours --------Control 38.4 36.0 21.4 18.5 0.2 0.1 Solarization 42.4 44.6 28.3 27.3 33.0 27.0 Steam a 99.7 98.7 35.9 34.1 14.0 28.0 Steam + solarization 85.8 78.3 31.5 27.7 50.0 8.0

aSteam temperatures taken over 24hr. Table 6. Fruit yields for strawberry in 2007-2008 growing season at Salinas, CA USA.

Treatment Yield (grams/plant) 1. MBPic 348.1 ab 2. Control 264.9 b 3. Solarization 291.9 b 4. Steam 353.2 ab 5. Steam + solarization 395.6 a Treatment P 0.07


Growth stage impacts tolerance to broadcast flaming in agronomic crops

S.Z. Knezevic, A. Datta and S.M. Ulloa Department of Agronomy and Horticulture, University of Nebraska, Northeast Research and

Extension Center, Concord, NE, 68728-2828, USA. Email: [email protected]

Abstract Flaming can be an additional tool for weed control in organic cropping systems. However,

tolerance of major crops must be determined in order to optimize proper use of flame. Therefore, objective of this study was to collect baseline information on crop tolerance to broadcast flaming as influenced by the growth stage at the time of flaming and propane dose. Field experiments were conducted at two locations during summer of 2007 utilizing six doses of propane and six crops including: field corn (Zea mays), sorghum (Sorghum bicolor), soybean (Glycine max), sunflower (Helianthus annuus), alfalfa (Medicago sativa) and red clover (Trifolium pratense). The propane doses included: 0, 12.1, 30.9, 49.7, 68.5 and 87.22 kg ha-1 at a constant speed of 6.5 km h-1. Crops response to propane doses were described utilizing log-logistic models based on visual estimates of crop injury and dry matter reduction. Overall response to flame varied depending on the crops, growth stage and propane dose. Alfalfa and red clover were the most susceptible to flaming regardless of the growth stage. Corn and sorghum were more tolerant than the broadleaf crops. Soybean tolerated more heat at the cotyledon stage compared to the reproductive growth stage whereas sunflower showed more tolerance at later stages of growth. Of all crops tested, broadcast flaming has the most potential for use in field corn and sorghum.

Introduction Weeds are one of the major pests in both conventional and organic crop production systems and

are responsible for significant crop yield reduction (Stopes & Millington, 1991). Especially in organic farming systems, weed control is recognized as the foremost production-related problem (Kloen & Daniels, 2000). Hand weeding and cultivation are the most popular physical weed control methods practiced by organic producers, but they are laborious, cost prohibitive, increase the chance of soil erosion, and promote the emergence of flushes of weeds (Wszelaki et al., 2007). In addition, very few chemicals are approved for organic farming, which are costly and non-selective in action; thus, alternative methods of weed control are necessary (Wszelaki et al., 2007).

There is a renewed interest in flame weeding, even as an alternative to chemical weed control (Ascard, 1994). Flame weeding is the most commonly used thermal method of weed control and could be a potential additional tool in the toolbox of integrated weed management (Ascard, 1995; Knezevic & Ulloa, 2007; Domingues et al., 2008; Teixeira et al., 2008). This technique utilizes propane as a source of combustion which generates temperatures of up to 1900oC (Ascard, 1995). Flaming weeds is basically a transfer of heat from the flame to the plant tissues resulting in boiling intracellular water molecules which ultimately expand and generate a pressure sufficient enough for cell membrane rupture and leakage (Rifai et al., 1996; Rahkonen & Jokela, 2003). It also results in the denaturation of cell proteins when the temperature reaches above 50oC (Parish, 1990) and eventually the plants die and/or their competitive ability is drastically reduced (Vanhala et al., 2004). The main advantages of flame weeding are the lack of chemical residues in the crop, soil or water, no chance of herbicide carry-over to the next season, wide spectrum of weeds control with no possibility of developing resistance to flaming and compatibility with no-tillage production techniques (Ascard, 1995, 1998; Mojzis, 2002; Cisneros & Zandstra, 2008).

In order to optimize the use of flaming as a weed control tool, the biologically effective dose (ED) of propane for tolerance of major crops must be determined (Knezevic & Ulloa, 2007). Depending on the desired level of tolerable crop injury level, a propane dose could be selected to either control the weed, or reduce its growth thereby offsetting its competitive ability against the


crop. Moreover, growth stage of the plant at the time of flaming determines plant sensitivity to heat (Ascard, 1994). Therefore, the objective of this study was to describe dose-response curves and determine crop tolerance to broadcast flaming as influenced by the growth stage and propane dose in two grassy type crops [field corn (Zea mays) and sorghum (Sorghum bicolor)] and four broadleaves [soybean (Glycine max), sunflower (Helianthus annuus), alfalfa (Medicago sativa) and red clover (Trifolium pratense)].

Materials and Methods Field experiments were conducted at the Haskell Agricultural Laboratory near Concord in

northeast Nebraska (42.37°N, 96.68°W) on Alcester series silty clay loam soil. The study was arranged in a split-plot design with the growth stage as the main plot and the sub-plots of flaming doses. The growth stages of each crop were based on number of true leaves (-V). The treatments consisted of six doses of propane (including untreated control) applied at different crop growth stages. The experiment was replicated 3 times, and repeated at two locations within Haskell Ag Lab during the 2007 growing season.

The experimental area for the first location was cultivated on June 16 and crop seeds were planted on June 29. Crops from both grass and broadleaf families were sown to an individual plot size of 7.6 m × 2.1 m in parallel lines using a manual push-planter, as a single row for each crop in 40 cm row spacing. Each replication had a row of each crop and the flaming treatments were applied across the rows utilizing a custom built flamer mounted on an ATV, which was driven across the crop rows. The flamer used propane as a source for combustion and there were four burners (LT 2 × 8) mounted 30 cm apart (Flame Engineering, 2007). Burners were positioned 20 cm above the soil surface and angled back at 30˚. Flaming treatments were applied using a constant speed of about 6.5 km h-1. Propane pressures included were 0, 69, 207, 345, 483 and 620 kPa. Combining pressure and speed, the doses of propane applied were 0, 12, 31, 50, 69 and 87 kg ha-1. The weather conditions were: wind speed of 11 km h-1, air and soil temperatures of 22°C and relative humidity of 46%. The study was repeated by following the same procedures and planting the exact set of crops on July 6 at the second location. Field was irrigated as needed to obtain proper emergence of crops and volunteer weeds were controlled by hand weeding.

Crop injury was estimated visually at 14 days after treatment (DAT) utilizing a scale of 0 to 100%, with 0 representing no crop injury and 100 representing crop death. At 14 DAT, one linear meter of each crop was also hand harvested for dry matter (DM). Fresh samples were dried at 50°C for 2 weeks and DM weighed and recorded. Crop DM was expressed as a percentage of untreated plants. Crop injury and DM data were analyzed utilizing the four-parameter log-logistic model as suggested by Seefeldt et al. (1995):

Y = C + {D − C / 1 + exp[B(log X − log E)]} [1] where Y is the response (e.g., visual quality or DM), C is the lower limit, D is the upper limit, B

is the slope of the line, X is the propane dose and E is the dose resulting a 50% response (also known as ED50). Curve fitting was done by non-linear regression using the least squares method. All statistical analysis and graphs were performed with R program (R Development Core Team, 2006) utilizing the drc statistical addition package (Knezevic et al., 2007). There was no treatment-by-location interaction thus visuals and DM data were combined over two sites, and a curve was fit for each crop and growth stage. The response of alfalfa and red clover to propane flaming were compared on the basis of DM whereas the tolerance level of field corn, sorghum, soybean and sunflower were based on the visual injuries. Data for the visual injury response variable were not shown; only table was presented with different ED values. The values of ED5 (effective dose that provides 5% injury), ED10 (10% injury) and ED20 (20% injury) were determined from the curves and used as measures of the level of crop damage by flaming treatments.


Results and Discussion Overall response to broadcast flaming varied among crops, their growth stages and propane

dose. In general, grassy type crops were more tolerant to broadcast flaming than broadleaves. Alfalfa and red clover were the most susceptible to broadcast flaming regardless of the growth

stage (Figures 1 and 2); while field corn was the most tolerant, requiring higher dose of propane. About 15 and 20 kg of propane ha-1 were needed to obtain 20% DM reduction (ED20) in alfalfa and red clover, respectively, compared to a much higher propane dose of 57 kg ha-1 in field corn (Tables 1 and 2). Furthermore, field corn and sorghum were more tolerant to flaming than soybean and sunflower (Table 2). About 57 and 46 kg ha-1 of propane were needed to obtain 20% injury in V2 field corn and sorghum, respectively. In contrast, soybean and sunflower flamed at V3 and VE stage required about half as much propane, 21 and 25 kg ha-1, respectively, for the same level of injury.

Figure 1. DM reduction in alfalfa (% of untreated) as influenced by propane dose at four different growth stages at 14 DAT. The regression lines are plotted using equation 1, and the parameter values are recorded in Table 1.

Figure 2. DM reduction in red clover (% of untreated) as influenced by propane dose at four different growth stages at 14 DAT. The regression lines are plotted using equation 1, and the parameter values are recorded in Table 1.

The level of crop damage and ED values were also influenced by the plant size in all crops except in field corn (Tables 1 and 2). For example, flaming corn with about 60 kg ha-1 of propane resulted in 20% injury at both V2 and V7 corn (Table 2). In contrary, there was a higher level of tolerance in alfalfa, red clover and soybean when flamed at earlier stages compared to any other growth stage. For example, when flaming was done at cotyledon stage in alfalfa, about 15 kg ha-1 of propane was needed to cause 20% DM reduction compared to 6 kg ha-1 for V3 and V8 stages (Table 1). Similar trend occurred in sorghum.

Sunflower flamed at crop emergence (VE stage) had 20% injury with 25 kg ha-1 of propane compared to a twice as higher dose of 50 kg ha-1 at V9 stage, suggesting that taller sunflower had more tolerance to flaming. However, such response might have been simply the result of physical positioning of the growing point in taller sunflower (V9) above the flames, as the flaming torches were placed at 20 cm above soil surface. This allowed growing point of taller sunflower to survive flaming operation, while the growing point in smaller plants was exposed to the flames, thus killed. In contrast, there was a reduction in ED values for sorghum and soybean with the increase in their size, indicating that taller sorghum and soybean were less tolerant to flaming, regardless of the fact

Propane consumption (kg/ha)

Re

l. d

ry w

eig

ht (

%)

0

20

40

60

80

100

0 5 20 100

CotyledonV2V4Late vegetative

Propane dose (kg ha-1)

Dry

matt

er

(%)


Re

l. d

ry w

eig

ht (

%)

0

20

40

60

80

100

0 5 20 100



Dry

matt

er

(%)


Re

l. d

ry w

eig

ht (

%)

0

20

40

60

80

100

0 5 20 100



Dry

matt

er

(%)


Re

l. d

ry w

eig

ht (

%)

0

20

40

60

80

100

0 5 20 100



Dry

matt

er

(%)


that their growing point was not directly exposed to the flame. It is possible that the leaf injuries in taller soybean resulted in much water loss, and plants were not able to recover. All of the above hypotheses need to be tested. Table 1. Dose of propane (kg ha-1) needed to obtain 5%, 10% and 20% DM reduction at 14 DAT in alfalfa and red clover (as indicated by the respective ED values (±SE)), as a function of crop growth stage.

Effective dose of propane

---------------------------kg ha-1----------------------

Crops Growth stage Slope (SE)

aI50 (SE) ED5 (SE) ED10 (SE) ED20 (SE) Cotyledon 2.4 (0.3) 27 (2) 8 (2) 11 (2) 15 (2)

V3 1.8 (0.3) 13 (2) 3 (1) 4 (1) 6 (2) V8 1.9 (0.4) 12 (1) 2 (1) 4 (1) 6 (1)

Late vegetative 1.4 (0.2) 14 (2) 2 (1) 3 (1) 5 (2)

Alfalfa

Cotyledon 2.1 (0.5) 39 (5) 10 (3) 14 (4) 20 (4)

V2 2.1 (0.7) 15 (3) 4 (2) 5 (3) 8 (3) V4 1.4 (0.6) 10 (3) 1 (1) 2 (2) 4 (2)

Red clover

Late vegetative 2.1 (0.6) 19 (3) 5 (2) 7 (3) 10 (3) a The dose resulting a 50% DM reduction. Table 2. Dose of propane that resulted in 5%, 10% and 20% crop injury, as indicated by the respective ED values (±SE), based on visual ratings at 14 DAT and a function of crop growth stage

Effective dose of propane (kg ha-1) Crops Growth stage

ED5 (SE) ED10 (SE) ED20 (SE)

V2 35 (11) 44(10) 57 (7) V7 31 (15) 43(14) 63 (9)

Field corn

V2 22 (12) 32(12) 46 (9) V9 9 (5) 14(6) 23 (6)

Sorghum

Cotyledon (VC) 26 (9) 34(9) 44 (7) V3 14 (2) 17 (2) 21 (2)

Flowering (R1) 12 (6) 16(6) 22 (6)

Soybean

Emergence (VE) 15 (4) 19 (4) 25 (4) Sunflower V9 17 (10) 29 (11) 49 (9)

Crop susceptibility to broadcast flaming in this study varied among the species and their growth

stages. These results support previous finding of others, who suggested that plant response to flaming is largely size and species dependent (Ascard, 1994; Wszelaki et al., 2007).

Alfalfa and red clover showed the least tolerance to flaming irrespective of their growth stage. However, those two crops might show higher level of tolerance in their second or third year of growth as they are perennial species. Corn and sorghum showed higher level of tolerance at their earlier growth stages likely because their growing point was below soil surface at the time of


flaming. Soybean was more tolerant to flaming at the cotyledon stage (VC), likely due to the fact that relatively thick cotyledons may have had enough food reserves to overcome loss of some surface area from flaming. Such hypothesis needs to be tested.

Of all crops tested, broadcast flaming has the most potential for use in field corn. Temporary injury of as much as 20% in field corn was obtained with about 60 kg ha-1 of propane. Our previous studies reported that a minimum propane dose of 60 kg ha-1 was needed to control most annual broadleaf weeds and many grasses (Knezevic & Ulloa, 2007; Domingues et al., 2008). Further studies are needed to determine the level of yield reduction in corn due to various injury levels from flaming.

Plant response to flaming for each crop evaluated in this study was unique. It is interesting to note that the plants from the same family and with similar morphological characteristics exhibited different responses to flaming (e.g., corn and sorghum). These results also suggested that more research is needed to evaluate various flaming procedures (e.g., positioning of the flame). For example, flaming inter-row space, or positioning flames below the crop canopy (e.g., away from growing point) might be much safer in broadleaf crops. Studies are needed to test such hypothesis.

Results of this study suggested that there is a potential for utilizing broadcast flaming as one of the tools for weed control, especially in the grass type crops (e.g., corn). Flaming could be repeated as needed during the growing season, or could be integrated with other non-chemical weed management strategies, as suggested by Domingues et al. (2008), in both organic and conventional crop production.

References Ascard J (1994) Dose-response models for flame weeding in relation to plant size and density. Weed Research 34, 377-

385. Ascard J (1995) Effects of flame weeding on weed species at different developmental stages. Weed Research 35, 397-

411. Ascard J (1998) Comparison of flaming and infrared radiation techniques for thermal weed control. Weed Research 38,

69-76. Cisneros JJ and Zandstra BH (2008) Flame weeding effects on several weed species. Weed Technology 22, 290-295. Domingues AC, Ulloa SM, Datta A and Knezevic SZ (2008) Weed response to broadcast flaming. Review of

Undergraduate Research in Agricultural and Life Sciences Vol. 3, Issue 1, Article 2. http://digitalcommons.unl.edu/rurals/vol3/iss1/2.

Flame Engineering (2007) Agricultural flaming guide. http://www.flameengineering.com/AgriculturalFlamingGuide.html. Accessed September 2, 2008. Flame Engineering Inc., LaCrosse, KS 67548.

Kloen H and Daniels L (2000) Onderzoeksagenda Biologische Landbouw & Voeding 2000-2004. Biologica ⁄ Wageningen-UR, Platform Biologica, Utrecht, the Netherlands.

Knezevic SZ, Streibig JC and Ritz C (2007) Utilizing R software package for dose-response studies: the concept and data analysis. Weed Technology 21, 840-848.

Knezevic SZ and Ulloa SM (2007) Flaming: potential new tool for weed control in organically grown agronomic crops. Journal of Agricultural Sciences 52, 95-104.

Mojzis M (2002) Energetic requirements of flame weed control. Research in Agricultural Engineering 48, 94-97. Parish S (1990) A review of non-chemical weed control techniques. Biological Agriculture and Horticulture 7, 117-

137. R Development Core Team (2006) R: A language and environment for statistical computing. R Foundation for

Statistical Computing, Vienna, Austria. URL http://www.R-project.org. Rahkonen J and Jokela H (2003) Infrared radiometry for measuring plant leaf temperature during thermal weed control

treatment. Biosystems Engineering 86, 257-266. Rifai MN, Lacko-Bartosova M and Puskarova V (1996) Weed control for organic vegetable farming. Rostlinna Vyroba

42, 463-466. Seefeldt SS, Jensen JE and Fuerst EP (1995) Log-logistic analysis of herbicide dose-response relationships. Weed

Technology 9, 218-227. Stopes C and Millington S (1991) Weed control in organic farming systems. Proceedings of the Brighton Crop

Protection Conference - Weeds, Brighton, UK, pp. 185-192. Teixeira HZ, Ulloa SM, Datta A and Knezevic SZ (2008) Corn (Zea mays) and soybean (Glycine max) tolerance to

broadcast flaming. Review of Undergraduate Research in Agricultural and Life Sciences Vol. 3, Issue 1, Article 1. http://digitalcommons.unl.edu/rurals/vol3/iss1/1.


Vanhala P, Kurstjens D, Ascard J, Bertram A, Cloutier DC, Mead A, Raffaelli M and Rasmussen J (2004) Guidelines for physical weed control research: flame weeding, weed harrowing and intra-row cultivation. 6th EWRS Workshop on Physical and Cultural Weed Control, 208 Lillehammer, Norway, 8-10 March 2004.

Wszelaki AL, Doohan DJ and Alexandrou A (2007) Weed control and crop quality in cabbage [Brassica oleracea (capitata group)] and tomato (Lycopersicon lycopersicum) using a propane flamer. Crop Protection 26, 134-144.


Response of corn (Zea mays L.) types to broadcast flaming

S.Z. Knezevic, C.M. da Costa, S.M. Ulloa and A. Datta Department of Agronomy and Horticulture, University of Nebraska, Northeast Research and


Abstract Flaming has a potential to be included into an integrated weed management programs in both

conventional and organic production systems. Therefore, the objective of this study was to collect baseline information on the tolerance of different corn types to broadcast propane flaming as influenced by growth stage at the time of flaming and propane dose. Field experiments were conducted during summer of 2008 utilizing three types of corn [field corn (Zea mays L.), popcorn (Zea mays L. var everta) and sweet corn (Zea mays L. var rugosa)] and six doses of propane applied at four growth stages of crop [V1 (1-leaf), V4 (4-leaf), V6 (6-leaf) and V8 (8-leaf)]. The propane doses were 0, 12.1, 30.9, 49.7, 68.5 and 87.22 kg ha-1. Flaming treatments were applied utilizing an ATV mounted flamer moving at a constant speed of 6.5 km h-1. Species response to propane doses were described by log-logistic models based on both visual estimates of crop injury and biomass. Visual injury ratings were conducted at 1, 7 and 14 days after treatment (DAT) with biomass data collected at 14 DAT. Overall response to flame varied among corn types, growth stages and propane dose. The most tolerant species to broadcast flaming appeared to be sweet corn at its later stages of growth (V8). Popcorn was the most sensitive to broadcast flaming regardless of the growth stage.

Introduction The interest for organic crop production in US is on the increase due to strong demand for

organic food from consumers and attractive income potential for organic farmers (Johnson, 2004). An example of such increase is the higher demand for organically grown field corn (Zea mays L.). There was a 300% increase in corn hectares from 1995 to 2005, as the total organic corn production grew from 13212 ha to 52881 ha in those 10 years (AgMRC, 2007). Similar trend occurred for popcorn (Zea mays L. var everta) and sweet corn (Zea mays L. var rugosa). Weeds are one of the major problems in both conventional and organic crop production systems and are responsible for significant crop yield reduction (Stopes & Millington, 1991). Especially in organic farming systems, weed control is recognized as the foremost production-related problem (Kloen & Daniels, 2000). Simply replacing herbicides by other control measures is challenging (Kruidhof et al., 2008). Mechanical cultivation is not very desirable option because of damage to soil structure, potential for soil erosion, increased risk of frost damage to crops and a dependency on weather conditions. Therefore, hand-weeding is often utilized by organic producers, which requires availability of sufficient labor and is costly (Kruidhof et al., 2008). Hence, systems-oriented approaches to weed management that make better use of alternative weed management tactics need to be developed (Liebman & Davis, 2000; Kruidhof et al., 2008).

Propane flaming could be a potential alternative tool in the toolbox of integrated weed management system (Knezevic & Ulloa, 2007; Domingues et al., 2008; Teixeira et al., 2008). Flaming weeds is basically a transfer of heat from the flame to the plant tissues resulting in boiling water molecules inside the cell. The expending water generates a pressure that ultimately ruptures the cell wall resulting in cell leakage (Rifai et al., 1996). It also results in the coagulation of cell proteins when the temperature reaches above 50oC (Parish, 1990) and eventually the plants die and/or their competitive ability is drastically reduced (Vanhala et al., 2004). Unlike other physical weed control methods, thermal control is not causing soil disturbance that can stimulate further flushes of weeds, promotes soil erosion, or harming the crop root system, and it is much less costly than hand-weeding (Ascard, 1990; Nemming, 1994; Wszelaki et al., 2007).


Knezevic & Ulloa (2007) observed in their studies that field corn and sorghum (Sorghum bicolor L.) when flamed at earlier stages were less susceptible than soybean (Glycine max L.). Teixeira et al. (2008) also reported that broadcast flaming had a good potential to be used in field corn when flaming was done by the V5 (5-leaf) stage. Several other flaming studies in vegetable crops had also demonstrated that this method could be utilized as an alternative to control weeds (Ascard, 1995).

Crop susceptibility to propane flaming varies with species and growth stages (Knezevic & Ulloa, 2007). The response of different types of corn with respect to their different growth stages to flaming is not well documented. Therefore, the objective of this study was to collect baseline information on the tolerance of different corn types to broadcast flaming as influenced by growth stage at the time of flaming and propane dose.

Materials and Methods Field experiments were conducted at the Haskell Agricultural Laboratory near Concord, NE

(42.37°N, 96.68°W) to determine the response of three corn types to broadcast flaming applied at 4 different growth stages for each corn type. The experimental design was a split-plot with 24 treatments (5 propane doses + untreated control × 4 growth stages) where the main plot was corn growth stage and the sub-plot was a dose of propane. The experiment was replicated 3 times, and repeated two times during the 2008 growing season.

The experimental area was cultivated on June 16 and July 1, for the first and second site, respectively. Plots (2.1 m wide × 3.8 m long) were planted to field corn, popcorn and sweet corn on June 17 and July 2, utilizing manual push-planters as a single row for each species in 40 cm row spacing. Weeds were controlled by hand weeding two times during the experiment. Flaming was done on July 9, July 16, July 22 and July 31, which corresponded to the growth stages of V2 (2-leaf), V5, V7 and V9 for field corn, and V1, V4, V6 and V8 for popcorn and sweet corn for the first site. The study at the second site was flamed on July 22, July 28, July 31 and August 8, which corresponded to same growth stages as at the first site. Flaming treatments were applied utilizing a custom built flamer mounted on an ATV, which was driven across the crop rows. The flamer was calibrated to deliver appropriate dose of propane as a source for combustion. Calibration procedure was based on combining propane pressure and operating speed (Knezevic et al., 2007a). Flaming treatments were applied at a constant speed of 6.5 km h-1, and propane pressure was changed in order to deliver propane doses that included: 0, 12, 31, 50, 69 and 87 kg ha-1. There were four burners (LT 2 × 8) mounted 30 cm apart (Flame Engineering, 2007). Burners were positioned 20 cm above the soil surface and angled back at 30°.

Crop injury was rated visually at 1, 7 and 14 days after treatment (DAT) using a scale of 0 (no crop injury based on untreated plots) to 100 (plant death). Plant biomass samples were also collected at 14 DAT. One linear meter of each corn type was clipped from each plot. Samples were dried at 50°C for 2 weeks and dry biomass weighed. Plant biomass was expressed as a percentage of untreated plants using a scale from 0 to 100. Crop grain yield was not collected.

Visual ratings and dry biomass data were analyzed utilizing the four-parameter log-logistic model as suggested by Seefeldt et al. (1995):

Y = C + {D − C / 1 + exp[B(log X − log E)]} [1] where Y is the response (e.g., dry biomass), C is the lower limit, D is the upper limit, B is the slope of the line, X is the propane dose and E is the dose giving a 50% response (also known as ED50). Curve fitting was done by non-linear regression using the least squares method. All statistical analysis and graphs were performed with R program (R Development Core Team, 2006) utilizing the drc statistical addition package (Knezevic et al., 2007b). There was no treatment-by-site interaction thus the dry biomass data were combined over two sites, and a curve was fit for each corn type and growth stage. The values of ED5 (effective dose that provides 5% injury), ED10 (10% injury) and ED20 (20% injury) were determined from the curves and used as measures of the level of


crop injury by flaming treatments. Visual rating data was similar to the biomass data, thus it is not reported.

Results and Discussion Overall response to broadcast flaming varied among corn types, their growth stages and

propane dose. In general, popcorn corn was the least tolerant while sweet corn was the most tolerant to broadcast flaming (Figures 1 and 2). Figure 1. Biomass loss (%) as a function of propane dose in three corn types and four growth stages at 14 DAT. Each data represents a mean of 2 sites. The regression lines are plotted using equation 1, and the parameter values are recorded in Table 1.

Popcorn was the least tolerant to broadcast flaming compared to field corn and sweet corn irrespective of the growth stage when flaming was conducted (Fig. 1). More biomass was reduced in popcorn per unit increase in propane dose compared to field corn and sweet corn at all stages of flaming (Table 1). For example, when flaming was done at V4 stage in popcorn, about 2 kg ha-1 of propane was needed to cause 20% biomass reduction compared to 7 kg ha-1 for V4 sweet corn and 10 kg ha-1 for V5 field corn at 14 DAT (Table 1). Similar trend was observed for other stages of popcorn.

There was no difference in field corn tolerance to broadcast flaming across the tested growth stages (Fig. 2). It appeared that the V5 stage showed more tolerance to broadcast flaming than other growth stages based on ED values (Table 1). For example, a slightly higher propane dose of 9.9 kg ha-1 was needed to obtain 20% biomass reduction in V5 field corn compared to 7.5, 8.0 and 7.5 kg

F irst flam in g stage

P ropane dose (kg/ha)

Bio

ma

ss

lo

ss

(%)

0

20

40

60

80

100

0 5 2 0 100

Field corn - V 2Popc orn-V1Sweet corn - V 1

Second flamin g stage

Propane dose (kg/ha)

Bio

ma

ss

lo

ss

(%)

0

20

40

60

80

100

0 5 20 100

Field corn - V 5P opcorn - V4S weet corn - V4

T hird flamin g stage


Bio

ma

ss

lo

ss

(%)

0

20

40

60

80

100

0 5 2 0 100

Field corn - V 7Popc orn - V6Sweet corn - V 6

Fo urth flamin g stag e


Bio

ma

ss

lo

ss

(%)

0

20

40

60

80

100

0 5 20 100


Propane dose (Kg ha-1) Propane dose (Kg ha-1)


Bio

mas

s lo

ss (

%)

Bio

mas

s lo

ss (

%)

Bio

mas

s lo

ss (

%)

Bio

mas

s lo

ss (

%)

F irst flam in g stage


Bio

ma

ss

lo

ss

(%)

0

20

40

60

80

100

0 5 2 0 100

Field corn - V 2Popc orn-V1Sweet corn - V 1

Second flamin g stage


Bio

ma

ss

lo

ss

(%)

0

20

40

60

80

100

0 5 20 100


T hird flamin g stage


Bio

ma

ss

lo

ss

(%)

0

20

40

60

80

100

0 5 2 0 100

Field corn - V 7Popc orn - V6Sweet corn - V 6

Fo urth flamin g stag e


Bio

ma

ss

lo

ss

(%)

0

20

40

60

80

100

0 5 20 100




Bio

mas

s lo

ss (

%)

Bio

mas

s lo

ss (

%)

Bio

mas

s lo

ss (

%)

Bio

mas

s lo

ss (

%)


ha-1 of propane for V2, V7 and V9 field corn, respectively (Table 1). However, those propane doses are not statistically different from each other (based on the size of their standard errors) suggesting that there was no difference in the tolerance level among growth stages in field corn. Similar trend occurred in popcorn when the ED values were compared among growth stages.

However, sweet corn tolerance to broadcast flaming increased with the increase in crop size, as it was evident with the steady increase in ED values with the later growth stages (Table 1). Based on ED values, sweet corn was the most tolerant to broadcast flaming at V8 stage compared to V1, V4 and V6 stages (Fig. 2 and Table 1). For example, the propane dose needed for 20% biomass reduction was 28 kg ha-1 at V8 stage compared to significantly lower 6.9 kg ha-1, 7.2 kg ha-1 and 12.4 kg ha-1 doses for V1, V4 and V6 stages (Table 1). This increased tolerance of sweet corn at later growth stages might be due to the fact that corn generally starts to accelerate its growth around the V6-V7 stage (growing point reaches soil surface). This growth acceleration in sweet corn also increases the storage capacity and concentration of sugars in cells and stem tissues, which requires more energy to boil water in the cell (Taiz & Zeiger, 2002). Response of sweet corn to broadcast flaming from our study is similar to the previous reports of Ascard (1994), who reported that plant size at flaming time had great influence on plant sensitivity, with small plants being more sensitive than large ones.

Similar trend was observed in all three corn types when visual rating data were compared among corn types and their growth stages (data not shown). Table 1. Regression parameters (equation 1) for each corn type and dose of propane (kg ha-1) needed to obtain 5%, 10% and 20% biomass reduction (±SE) in four different growth stages at 14 DAT

cI50 (SE) ED5 (SE) ED10 (SE) ED20 (SE)Species Growth stages

aB bC

--------------------------- kg ha-1--------------------------

Field corn V2 1.6 15.7 17.3 (7.2) 2.9 (2.6) 4.6 (2.8) 7.5 (2.7)

Popcorn V1 0.9 0.2 12.6 (11.3) 0.6 (1.7) 1.3 (2.6) 3.0 (3.1)

Sweet corn V1 1.8 15.9 15.1 (4.5) 2.9 (2.6) 4.4 (2.8) 6.9 (2.7)

Field corn V5 2.1 37.6 19.4 (8.3) 4.7 (4.9) 6.7 (5.3) 9.9 (5.3)

Popcorn V4 0.9 -0.6 9.5 (10.7) 0.3 (1.6) 0.7 (2.7) 1.9 (4)

Sweet corn V4 2.1 35.2 13.9 (5) 3.4 (4.6) 4.9 (4.8) 7.2 (4.3)

Field corn V7 1.4 27.4 21 (20.5) 2.7 (4) 4.6 (4.4) 8.0 (4)

Popcorn V6 1.6 17.9 18.5 (10) 3 (3.4) 4.7 (3.7) 7.8 (3.4)

Sweet corn V6 2.4 34.9 21.7 (7.2) 6.6 (4.3) 8.9 (4.3) 12.4 (4.3)

Field corn V9 1.2 14.2 26 (40) 1.8 (3.4) 3.6 (4.2) 7.5 (5.4)

Popcorn V8 1.0 2.1 19 (32) 0.9 (2.6) 2.0 (3.6) 4.6 (3.9)

Sweet corn V8 1.8 22.6 60 (58) 11.9 (7.7) 18 (2) 28 (8)

a The slope of the line. b The lower limit. c The dose resulting a 50% biomass reduction.


Figure 2. Biomass (% of untreated) as influenced by propane dose in each corn type with four different growth stages at 14 DAT. Each data represents a mean of 2 sites. The regression lines are plotted using equation 1, and the parameter values are recorded in Table 1.

Results of this study demonstrated that popcorn was more sensitive to broadcasts flaming than field corn and sweet corn. Also, there was no difference in growth stage tolerance to flaming in field corn and popcorn, while sweet corn tolerance increased with later growth stages. We believe that propane flaming has a potential for use in organic agriculture, particularly with grass family crops like corn, or could be integrated with other non-chemical weed management strategies.

No practical recommendations can be given from this study at this time, as the obvious concern is that corn injury levels higher than 10% or even 20%, likely will not be acceptable by the organic

Corn

Propane rate (kg/ha)

Dry

wei

ght (

%)

0

20

40

60

80

100

0 5 20 100

V2V5V7V9

Pop Corn


Dry

wei

ght (

%)

0

20

40

60

80

100

0 5 20 100

V1V4V6V8

Sw eet Corn


Dry

wei

ght (

%)

0

20

40

60

80

100

0 5 20 100

V1V4V6V8

Field corn

Popcorn

Sweet corn




Bio

mas

s (%

) B

iom

ass

(%)

Bio

mas

s (%

)


producers. Many producers will be asking this simple question: “Is the 10% crop injury going to cause 10% yield reduction”. Therefore, additional studies are needed to test the relationship between the injury level by flaming, and corresponding crop yields and yield components in each corn type.

References AgMRC: Agricultural Marketing Resource Center (2007) Organic Corn Profile.

http://www.agmrc.org/agmrc/commodity/grainsoilseeds/corn/organiccornprofile.htm. Accessed September 2, 2008.

Ascard J (1990) Thermal weed control with flaming in onions. Proceedings of the 3rd International Conference on Non-chemical Weed Control, Linz, Austria, pp. 175-188.

Ascard J (1994) Dose-response models for flame weeding in relation to plant size and density. Weed Research 34, 377-385.

Ascard J (1995) Effects of flame weeding on weed species at different developmental stages. Weed Research 35, 397-411.

Domingues AC, Ulloa SM, Datta A and Knezevic SZ (2008) Weed response to broadcast flaming. Review of Undergraduate Research in Agricultural and Life Sciences Vol. 3, Issue 1, Article 2. http://digitalcommons.unl.edu/rurals/vol3/iss1/2.


Johnson WC (2004) Weed control with organic production. Proceedings of the Southeast Regional Fruit and Vegetable Conference, Savannah, Georgia, pp. 13-14.



Knezevic SZ, Dana L, Scott JE and Ulloa SM (2007a) Building a research flamer. Proceedings of the North Central Weed Science Society, 62, 32.

Knezevic SZ, Streibig JC and Ritz C (2007b) Utilizing R software package for dose-response studies: the concept and data analysis. Weed Technology 21, 840-848.

Kruidhof HM, Bastiaans L and Kropff MJ (2008) Ecological weed management by cover cropping: effects on weed growth in autumn and weed establishment in spring. Weed Research 48, 492-502.

Liebman M and Davis AS (2000) Integration of soil, crop and weed management in low-external-input farming systems. Weed Research 40, 27-47.

Nemming A (1994) Costs of flame cultivation. Acta Horticulturae 372, 205-212. Parish S (1990) A review of non-chemical weed control techniques. Biological Agriculture and Horticulture 7, 117-

137. R Development Core Team (2006) R: A language and environment for statistical computing. R Foundation for

Statistical Computing, Vienna, Austria. URL http://www.R-project.org. Rifai MN, Lacko-Bartosova M and Puskarova V (1996) Weed control for organic vegetable farming. Rostlinna Vyroba

42, 463-466. Seefeldt SS, Jensen JE and Fuerst EP (1995) Log-logistic analysis of herbicide dose-response relationships. Weed


Protection Conference-Weeds, Brighton, UK, pp. 185-192. Taiz L and Zeiger E (2002) Plant Physiology. Massachussetts: Sinauer Associates, 3rd edition, Chapter 3: Water and

Plant Cells, pp. 33-46. Teixeira HZ, Ulloa SM, Datta A and Knezevic SZ (2008) Corn (Zea mays) and soybean (Glycine max) tolerance to





Tolerance of selected weed species to broadcast flaming at different growth stages

S.Z. Knezevic, A. Datta and S.M. Ulloa Department of Agronomy and Horticulture, University of Nebraska, Northeast Research and


Abstract Organic producers rank weeds as the most important pests that limit crop production. In order

to optimize the use of propane flaming as a weed control tool, the objective of this study was to collect baseline data on weed tolerance to broadcast flaming performed at different growth stages. Field studies were conducted in 2007 utilizing six doses of propane and five major annual weed species in northeast Nebraska, including two grasses [barnyardgrass (Echinochloa crus-galli), green foxtail (Setaria viridis)] and three broadleaves [field bindweed (Convolvulus arvensis), kochia (Kochia scoparia), Venice mallow (Hibiscus trionum)]. The propane doses included were 0, 12.1, 30.9, 49.7, 68.5 and 87.22 kg ha-1. Flaming treatments were applied utilizing an ATV mounted flamer moving at a constant speed of 6.5 km h-1. Species response to propane doses were described by log-logistic models based on relative dry matter for each weed species. Overall response to flame varied among the species, growth stages and propane dose. Broadleaf weeds were more susceptible to flames than grasses regardless of the growth stage. Propane doses of 50-70 kg ha-1 provided a 90% control of most broadleaf species at early growth stages. Although, 70-90 kg ha-1 provided 80% control of grasses, none of the propane doses provided 90% control. Flaming has a potential to be used effectively in organic agriculture.

Introduction For both conventional and organic crop production systems, weeds are considered as one of the

major pests limiting crop yield (Stopes & Millington, 1991). Especially in organic farming, weed control is recognized as the foremost production-related problem (Kloen & Daniels, 2000). Hand weeding and cultivation are the most popular physical methods for weed control utilized by organic growers. Very few organic herbicides are approved for organic farming but they are costly and non selective, thus can injure the crops. Moreover, labor costs associated with hand weeding are high and repeated cultivation destroys soil quality and increases the chance of soil erosion; thus, alternative methods of weed control are necessary (Wszelaki et al., 2007). In addition, ground and surface water contamination and pesticide residues in drinking water and food have sparked public awareness of and restrictions on herbicide use (Mojzis & Rifai, 1995). For these reasons weed scientists are considering alternative and integrated weed management practices to reduce herbicide inputs and impacts (Rifai et al., 2000). The use of propane for flame weeding could be one of the alternative control methods for weed control in organically grown field crops (Knezevic & Ulloa, 2007), as they leave no residual effects on soil, water and food quality (Ascard, 1998).

The efficacy of flame weeding is attributed both to a direct effect of the heat on the cell membranes and to an indirect effect during subsequent desiccation (Vester, 1990). During the flaming process, the heat from the flame is transferred to the plant tissues (Lague et al., 2000) and direct heat injury results in denaturation and aggregation of membrane proteins if the temperature reaches above 50°C (Parish, 1990). Furthermore, exposing plant tissue to a temperature of just above 100°C for a split second (e.g., 0.1 second) can cause water boiling and cell membrane rupture (Morelle, 1993; Pelletier et al., 1995), resulting in loss of water and plant death (Rifai et al., 1996). Plants may survive flaming, either by avoidance or by heat tolerance. The extent to which heat from the flames penetrates plants depends on the flaming technique, and environmental factors including the leaf surface moisture (Vester, 1988; Parish, 1990).


Flame weeding is less costly than hand-weeding and can be used when the soil is too moist or stony for mechanical weeding (Ascard, 1990; Nemming, 1994). In contrast to cultivation, flaming does not disrupt the soil surface, thus reducing the risk of soil erosion, and it does not bring buried weed seeds to the surface, thus minimizing the chance of new flushes of weeds (Wszelaki et al., 2007). Flaming can also provide added benefits, such as insect and/or disease control (Lague et al., 1997). Ascard (1994) reported that plant size had greater influence upon sensitivity than did plant density, with small weeds being more sensitive to flaming than large weeds. Therefore, the objective of this study was to describe dose-response curves for propane when flaming selected weed species at different growth stages. The weed species included were two grasses [barnyardgrass (Echinochloa crus-galli), green foxtail (Setaria viridis)] and three broadleaf species [Field bindweed (Ipomoea hederacea), kochia (Kochia scoparia), Venice mallow (Hibiscus trionum)].

Materials and Methods Field experiments were conducted at the Haskell Agricultural Laboratory near Concord in

northeast Nebraska (42.37°N, 96.68°W) on Kennebec series silty clay loam soil. The experimental design was a split-plot with 18 treatments (5 doses of propane + untreated control × 3 growth stages) where the main plot was the growth stage and the sub-plot was a flaming dose. The experiment was replicated 3 times, and repeated two times during the 2007 growing season.

The experimental area for the first site was cultivated on June 16 and weed seeds were planted on June 29. Plots (7.6 m × 2.1 m) were sown to grass and broadleaf species in parallel lines using push-planter, as a single row for each species in 40 cm row spacing. The growth stages of each weed species were based on number of leaves (-L). Each replication had a row of each weed species and the flaming treatments were applied across the rows utilizing a custom built flamer mounted on an ATV, which was driven across the weed rows. The flamer used propane as a source for combustion and there were four burners (LT 2 × 8) mounted 30 cm apart providing a 120 cm wide flaming swath (Flame Engineering, 2007). Burners were positioned 20 cm above the soil surface and angled back at 30˚. Flaming treatments were applied using a constant speed of about 6.5 km h-1. Propane pressures included were 0, 69, 207, 345, 483 and 620 kPa. Combining pressure and speed, the doses of propane applied were 0, 12, 31, 50, 69 and 87 kg ha-1. The weather conditions were: wind speed of 11 km h-1 (direction NNW), air and soil temperatures of 22°C and relative humidity of 46%. The study was repeated by following the same procedures and planting the exact set of weed species on July 6 at the second site. Field was irrigated as needed to obtain proper emergence of species and volunteer weeds were controlled by hand weeding.

At 14 days after treatment (DAT), one linear meter of each weed species was clipped from each plot to collect dry matter (DM) data. Fresh samples were dried at 50°C for 2 weeks and DM weighed. Weed DM was expressed as a percentage of untreated plants.

DM data were analyzed utilizing the four-parameter log-logistic model as suggested by Seefeldt et al. (1995):

Y = C + {D − C / 1 + exp[B(log X − log E)]} [1] where Y is the response (e.g., DM), C is the lower limit, D is the upper limit, B is the slope of the line, X is the propane dose and E is the dose resulting a 50% response (also known as ED50). Curve fitting was done by non-linear regression using the least squares method. All statistical analysis and graphs were performed with R program (R Development Core Team, 2006) utilizing the drc statistical addition package (Knezevic et al., 2007). There was no treatment-by-site interaction thus DM data were combined over two sites, and a curve was fit for each weed species and growth stage. The values of ED80 (effective dose that provides 80% control) and ED90 (90% control) were determined from the curves and used as measures of the level of weed control by flaming treatments.


Results and Discussion In general, response to broadcast flaming varied among weed species, their growth stages and

propane dose. Overall, broadleaf weeds were much more susceptible to broadcast flaming than grasses, requiring lower doses of propane to obtain the same level of control regardless of the growth stage (Figures 1-5).

Figure 1. DM reduction in field bindweed (% of untreated) as influenced by propane dose in three different growth stages at 14 DAT. The regression lines are plotted using equation 1, and the parameter values are recorded in Table 1.

Figure 2. DM reduction in kochia (% of untreated) as influenced by propane dose in three different growth stages at 14 DAT. The regression lines are plotted using equation 1, and the parameter values are recorded in Table 1.

About 40-75 kg ha-1 of propane was needed to obtain 90% control of field bindweed, kochia

and Venice mallow for their growth stages ranging from 3-L to10-L (Table 1). In contrast, 90% control of barnyardgrass and green foxtail was achieved with a much higher dose of propane, ranging from 125-152 kg ha-1 for their growth stages from 4-L to 7-L. Similarly, Wszelaki et al. (2007) reported that broadleaf weeds were more susceptible to flaming than grass species. Domingues et al. (2008) also suggested that barnyardgrass and green foxtail were more tolerant to flaming than velvetleaf (Abutilon theophrasti) and morningglory (Ipomoea hederacea). Such difference is likely a result of the physical position of the growing point at the time of flaming (Knezevic & Ulloa, 2007). Growing point in broadleaf species was above the ground, thus exposed to the flame and heat. In grass species, growing point was below soil surface, thus protected from the flame.

Annual broadleaf and grass species also differed in the way plant responded to the flame and heat. Leaves of annual broadleaf species turned brown and died within few days after flaming, resulting in no re-growth. Leaves of grass species turned brown shortly after flaming, leaving an appearance of a dead plant; however, within a week period most plants begin to recover with the growth of new leaves, followed by the full plant recovery within few weeks after flaming. These observations are similar of those by Ascard (1995) who reported that grass species flamed at early growth stages exhibited initial plant stunting followed by plant recovery after few weeks. Similar responses were also reported for field crops belonging to grass versus broadleaf families. For instance, Teixeira et al. (2008) demonstrated that soybean (Glycine max) flamed at 3-L stage was more injured by flaming than field corn (Zea mays) at 5-L stage, which re-grew a week later.

It is also important to note that the plant size had a major influence on the propane dose requirement for 80% or 90% DM reduction, regardless of the weed species. For example, there was


Dry

we

ight

(%

)

0

20

40

60

80

100

0 5 20 100

8-L10-L40-L


Dry

mat

ter

(%)


Dry

we

ight

(%

)

0

20

40

60

80

100

0 5 20 100

8-L10-L40-L


Dry

mat

ter

(%)


Dry

we

igh

t (%

)

0

20

40

60

80

100

0 5 20 100

6-L10-LFlowering


Dry

matt

er

(%)


Dry

we

igh

t (%

)

0

20

40

60

80

100

0 5 20 100

6-L10-LFlowering


Dry

matt

er

(%)


much less propane (43 kg ha-1) needed to obtain 90% DM reduction of smaller field bindweed (10-L) compared to 105 kg ha-1 for much larger plant (40-L) (Table 1). Similar trend was observed with other weeds. These findings are similar to those of Ascard (1994) who reported that propane doses above 40 kg ha-1 were needed to control smaller plants (1-2 true leaves), whereas plants with 2-4 leaves needed about 70 kg ha-1. The tolerance of different plant parts to flaming can also be influenced by other factors, including the presence of protective layers of hair and/or wax, lignification and conditions of water status (Ascard, 1995).

Figure 3. DM reduction in Venice mallow (% of untreated) as influenced by propane dose in three different growth stages at 14 DAT. The regression lines are plotted using equation 1, and the parameter values are recorded in Table 1.

We believe that flaming has an excellent potential for annual weed control, especially for broadleaf species. More research is needed with flaming other major weeds, including burner heights in relation to weed and crop growth stages, burner angle, intra-row and inter-row flaming. Information from such research would allow inclusion of flaming into an integrated weed management for both conventional and organic production systems. Flaming might also be repeated as needed during the growing season, or integrated with other chemical or non-chemical weed management strategies, as suggested by Domingues et al. (2008).


Dry

we

igh

t (%

)

0

20

40

60

80

100

0 5 20 100

3-L5-LFlowering


Dry

matt

er (%

)


Dry

we

igh

t (%

)

0

20

40

60

80

100

0 5 20 100

3-L5-LFlowering


Dry

matt

er (%

)


Figure 4. DM reduction in barnyardgrass (% of untreated) as influenced by propane dose in three different growth stages at 14 DAT. The regression lines are plotted using equation 1, and the parameter values are recorded in Table 1.

Figure 5. DM reduction in green foxtail (% of untreated) as influenced by propane dose in three different growth stages at 14 DAT. The regression lines are plotted using equation 1, and the parameter values are recorded in Table 1.

Table1. Dose of propane needed to obtain 80% and 90% weed control, as indicated by the respective ED values (±SE), based on DM reduction at 14 DAT as function of weed growth stage


--------------------------- kg ha-1------------------------

Weed species Growth stage

aB (SE)

bI50 (SE) ED80 (SE) ED90 (SE) 8-L 1.6 (0.2) 15 (1) 36 (4) 60 (7) 10-L 2.1 (0.3) 15 (1) 29 (3) 43 (4)

Field bindweed

40-L 1.2 (0.2) 18 (2) 54 (14) 104 (42) 6-L 3.4 (0.9) 34 (3) 51 (7) 65 (13) 10-L 2.2 (0.7) 15 (2) 28 (6) 40 (12)

Kochia

Flowering 1.5 (0.4) 25 (4) 61 (13) 102 (32) 3-L 1.3 (0.2) 13 (2) 40 (6) 75 (18) 5-L 1.9 (0.3) 19 (2) 40 (5) 61 (10)

Venice mallow

Flowering 1.0 (0.2) 24 (4) 93 (19) 204 (68) 4-L 1.3 (0.2) 29 (3) 82 (13) 152 (38) 7-L 1.2 (0.2) 24 (3) 73 (12) 141 (35)

Barnyardgrass

Flowering 1.4 (0.4) 64 (6) 175 (51) 313 (137) 5-L 1.5 (0.2) 29 (3) 74 (11) 128 (29) 7-L 1.3 (0.2) 22 (3) 66 (11) 125 (32)

Green foxtail

Flowering 1.6 (0.3) 49 (4) 120 (22) 202 (53) a The slope of the line. b The dose resulting a 50% DM reduction.


Dry

we

igh

t (%

)

0

20

40

60

80

100

0 5 20 100

4-L7-LFlowering


Dry

matt

er (

%)


Dry

we

igh

t (%

)

0

20

40

60

80

100

0 5 20 100

4-L7-LFlowering


Dry

matt

er (

%)

Propane consumption (kg/ha) D

ry w

eig

ht (

%)

0

20

40

60

80

100

0 5 20 100

5-L7-LFlowering


Dry

matt

er

(%)

Propane consumption (kg/ha) D

ry w

eig

ht (

%)

0

20

40

60

80

100

0 5 20 100

5-L7-LFlowering


Dry

matt

er

(%)


References Ascard J (1990) Thermal weed control with flaming in onions. Proceedings of the 3rd International Conference on Non-

chemical Weed Control, Linz, Austria, pp. 175-188. Ascard J (1994) Dose-response models for flame weeding in relation to plant size and density. Weed Research 34, 377-

385. Ascard J (1995) Effects of flame weeding on weed species at different developmental stages. Weed Research 35, 397-

411. Ascard J (1998) Comparison of flaming and infrared radiation techniques for thermal weed control. Weed Research 38,

69-76. Domingues AC, Ulloa SM, Datta A and Knezevic SZ (2008) Weed response to broadcast flaming. Review of






Lague C, Gill J, Lehoux N and Peloquin G (1997) Engineering performances of propane flamers used for weed, insect pest, and plant disease control. Applied Engineering in Agriculture 13, 7-16.

Lague C, Gill J and Peloquin G (2000) Thermal control in plant protection. In Physical control methods in plant protection/La lute physique en phytoprotection. Edited by Vincent C, Panneton B and Fleurat-Lessard F, Springer-Verlag, pp. 35-46.

Mojzis M and Rifai MN (1995) Controlling weeds by flame. Proceedings of the Symposium on Ecological Problems of Plant Protection and Contemporary Agriculture, 25-29 September 1995, pp. 97-100.

Morelle B (1993) Le desherbage thermique et ses applications en agriculture et en horticulture, in J.M. Thomas (ed) Proceedings of the Fourth IFOAM International Conference, pp. 109-115.

Nemming A (1994) Costs of flame cultivation. Acta Horticulturae 372, 205-212. Parish S (1990) A review of non-chemical weed control techniques. Biological Agriculture and Horticulture 7, 117-

137. Pelletier Y, McLeod CD and Bernard G (1995) Description of sub-lethal injuries caused to the Colorado potato beetle

by propane flamer treatment. Journal of Economic Entomology 88, 1203-1205. R Development Core Team (2006) R: A language and environment for statistical computing. R Foundation for

Statistical Computing, Vienna, Austria. URL http://www.R-project.org. Rifai MN, Lacko-Bartosova M and Puskarova V (1996) Weed control for organic vegetable farming. Rostlinna Vyroba

42, 463-466. Rifai MN, Lacko-Bartosova M and Brunclik P (2000) Alternative methods for weed control in apple orchards. Pakistan

Journal of Biological Science 3, 933-938. Seefeldt SS, Jensen JE and Fuerst EP (1995) Log-logistic analysis of herbicide dose-response relationships. Weed


Protection Conference - Weeds, Brighton, UK, pp. 185-192. Teixeira HZ, Ulloa SM, Datta A and Knezevic SZ (2008) Corn (Zea mays) and soybean (Glycine max) tolerance to


Vester J (1988) Flame cultivation for weed control, 2 years results. In: Cavalloro R & El Titi A, eds. Proceedings of a Meeting of the EC Experts Group/Stuttgart, 28-31 October 1986. Weed Control in Vegetable Production. Rotterdam/Brookfild: A. A, Balkema, pp. 153-167.

Vester J (1990) Flammebehandling, behandlingsintensiter ogukrudtseffekt (Summary: Flame treatment – intensity and effects on weeds). Nordic Postgraduate Course in Plant Production Science. Eleventh Course: Weeds and Weed Control, Garpenberg, Sweden. Uppsala: Swedish University of Agricultural Sciences, 10, pp. 1-17.



Winter wheat (Triticum aestivum L.) tolerance to broadcast flaming

S.Z. Knezevic, J.F. Neto, S.M. Ulloa and A. Datta Department of Agronomy and Horticulture, University of Nebraska, Northeast Research and


Abstract Flaming has a potential to be utilized as part of an integrated weed management program for

organic wheat (Triticum aestivum L.) production. Therefore, objective of this study was to collect baseline information on winter wheat tolerance to broadcast flaming as influenced by the growth stage of wheat at the time of flaming and dose of propane. Field experiments were conducted during summer of 2008 utilizing six doses of propane applied at three growth stages of winter wheat including: shoot elongation, first node and boot stage. The propane doses were 0, 12.1, 30.9, 49.7, 68.5 and 87.22 kg ha-1 and were applied using a constant speed of 6.5 km h-1. Crop response to propane doses was described by log-logistic models based on visual estimates of crop injury, various yield components and grain yield. Overall response to flaming was influenced by the growth stage of wheat and propane dose. In general, wheat at the shoot elongation stage was the most tolerant to broadcast flaming. Crop yield, number of spikes m-2, number of seeds spike-1 and weight of 100-seed were significantly reduced as the propane dose increased, and with later timing of flaming operation.

Introduction The area and production under organic crops are increasing steadily in the US mainly due to

strong demand for organic food from the consumers (Johnson, 2004). The state of Nebraska was ninth nationally in 2006 in certified organic crop hectares where wheat (Triticum aestivum L.) was the top ranked certified organic crop followed by corn (Zea mays L.) and soybeans (Glycine max L.) (Parsons, 2008). For both conventional and organic crop production systems, weeds are one of the major problems and responsible for significant crop yield reduction (Stopes & Millington, 1991). In particular, organic farmers rank weed control as the number one problem limiting production. However, controlling weeds in organic farming requires the use of many techniques and strategies to achieve economically acceptable weed control and yields (Walz, 1999). Popular physical weed control methods such as hand weeding and cultivation are cost prohibitive, increase the chance of soil erosion, destroy soil quality and promote the emergence of flushes of weeds (Wszelaki et al., 2007). In addition, very few chemicals are approved for organic farming but they are costly and non-selective in action; thus, alternative methods of weed control are necessary (Wszelaki et al., 2007).

Recent developments in conventional cropping systems, including an increasing number of herbicide-resistant weeds, higher costs of herbicides and more concern about pesticides in the environment, have resulted in a renewed interest in flaming as an alternative tool for weed control (Wszelaki et al., 2007). During the flaming process, the heat from the flame is transferred to the plant tissues (Lague et al., 2000) and results in the coagulation of cells protein at the temperature above 50oC (Parish, 1990), and eventually the plants die and/or their competitive ability is drastically reduced (Vanhala et al., 2004). Flaming could be an essential component of a multi-faceted weed control program, which could lessen the reliance on herbicides, hand weeding and/or mechanical cultivation (Wszelaki et al., 2007). Flaming may provide added benefits for insect or disease control (Lague et al., 1997).

Crop susceptibility to propane flaming varies with species and growth stages (Knezevic & Ulloa, 2007; Teixeira et al., 2008; Domingues et al., 2008). The response of major agronomic crop such as wheat with respect to their different growth stages to propane flaming is not well documented. Knezevic & Ulloa (2007) observed in their studies that field corn and sorghum


(Sorghum bicolor L.) when flamed at earlier stages were less susceptible than soybean (Glycine max L.). Teixeira et al. (2008) also reported that broadcast flaming had a good potential to be used in field corn when flaming was done by the V5 (5-leaf) stage. Moreover, the response of the major crops to flaming needs to be determined, with the intention to optimize the use of flaming as a weed control tool. Therefore, the objective of this study was to collect baseline information on winter wheat tolerance to broadcast flaming as influenced by the growth stage of wheat at the time of flaming and dose of propane.

Materials and Methods A field experiment was conducted at the Haskell Agricultural Laboratory near Concord, NE

(42.37°N, 96.68°W) to determine the response of winter wheat to broadcast flaming applied at 3 different growth stages. The experiment was set up as a split-plot design with 18 treatments (5 propane doses + untreated control × 3 growth stages) where the main plot was wheat growth stage and the sub-plot was a flaming dose. The study was replicated three times. The stages of wheat for flaming included shoot elongation-SE (Feeks 4.0), first node-FN (Feeks 6.0) and boot stage-BS (Feeks 10.1). Wheat was planted on October 5, 2007 in a no-tillage system by maintaining 20 cm row spacing and the plot dimensions were 2 m × 12 m. Weeds inside the plots were controlled by hand weeding. Flaming was done on May 7, May 21 and June 2 which corresponded to the SE, FN and BS, respectively.

Treatments were applied with a custom built flamer mounted on an ATV. The flamer used propane as a source for combustion. There were four burners (LT 2 × 8) mounted 30 cm apart (Flame Engineering, 2007). Burners were positioned 20 cm above the soil surface and angled back at 30°. Flaming treatments were applied using a constant speed of 6.5 km h-1. The propane pressures included were: 69, 207, 345, 483 and 620 kPa. Combining pressure and speed, the doses of propane applied were: 0, 12, 31, 50, 69 and 87 kg ha-1. The weather conditions were: wind speed of 7 km h-1

(direction NW), air and soil temperatures of 22°C and relative humidity of 81%. Crop injury was rated visually at 1 and 28 days after treatment (DAT) using a scale of 0 (no

crop injury based on untreated plots) to 100 (plant death). In addition to visual ratings, yield components (spikes m-2, seeds spike-1 and 100-seed weight) and grain yield data were also collected.

Visual estimations and yield were analyzed for each rating date utilizing the four-parameter log-logistic model as suggested by Seefeldt et al. (1995):

Y = C + {D − C / 1 + exp[B(log X − log E)]} [1] where Y is the response (e.g., visual quality), C is the lower limit, D is the upper limit, B is the slope of the line, X is the propane dose and E is the dose resulting a 50% response (also known as ED50). Curve fitting was done by non-linear regression using the least squares method. All statistical analysis and graphs were performed with R program (R Development Core Team, 2006) utilizing the drc statistical addition package (Knezevic et al., 2007). The values of ED5 (effective dose that provides 5% injury), ED10 (10% injury) and ED20 (20% injury) were determined from the curves and used as measures of the level of crop damage by flaming treatments.

Wheat yields under propane treatments for each growth stage were converted to the percentage of the yield obtained in the control plots and from that yield loss was calculated. A regression of the yield loss data against propane dose was then performed using the hyperbolic model as described by Cousens (1985):

Y = I*d/(1 + I*d/A) [2] Where Y is the yield loss, I is the slope, d is the propane dose and A is the asymptote of the hyperbolic line. The regression parameters were obtained using the Non Linear Least Squares (NLS) function of the statistical software R (R Development Core Team, 2006) and graphical presentation was generated using the same software.


Results and Discussion Overall response to broadcast flaming was influenced by the growth stage of wheat and

propane dose. In general, winter wheat flamed at the SE stage showed the most tolerance to broadcast flaming.

Based on visual ratings of crop injuries, the SE stage was the most tolerant period for flaming followed by the FN and BS stages (Fig. 1). Despite the fact that wheat plants flamed at the SE stage exhibited the highest injury level at 1 DAT, they have recovered by 28 DAT much better than the plants flamed at the FN and BS stages. For example, at 1 DAT about 20% injury (the ED20) was achieved with 22, 47 and 34 kg ha-1 compared to 76, 45 and 56 kg ha-1 for SE, FN and BS stages, respectively, at 28 DAT (Table 1). The slope comparisons also indicated that winter wheat flamed at BS had the longest lasting injuries. At 28 DAT, the slopes of the lines for BS, FN and SE were 4.6, 2.6 and 0.9 respectively, suggesting that for every kg ha-1 in propane dose increase, the injury was the highest for BS followed by the FN and SE stages (Table 1).

Based on yield components and relative yield, SE was also the most tolerant stage for flaming (Fig. 2). Of all yield components, spikes m-2 was most affected by flaming followed by number of seeds spike-1 and 100-seed weight. The slopes of the line for spikes m-2 at BS and FN were significantly higher than the slope for SE stage, indicating the least effects of flaming at SE stage (data not shown). The relative yield of wheat was also less affected when flaming was done at the SE stage compared to other growth stages. It is indicated by the lowest value of the slope for the line representing the SE stage, which indicates the lowest rate of yield reduction. For example, for every 1 kg ha-1 of propane dose increase, 0.3%, 0.6% and 0.8% yield was reduced for SE, FN and BS, respectively (data not shown). Figure 1. Winter wheat injury as influenced by propane dose and growth stage at 1 and 28 DAT. The regression lines are plotted using equation 1, and the parameter values are recorded in Table 1.

1 DAT

Propane rate (kg ha−1

)

Cro

p d

amag

e (%

)

0

20

40

60

80

100

0 20 50 100

a) SESb) FNSc) BS

28 DAT


)

Cro

p d

amag

e (%

)

0

20

40

60

80

100

0 20 50 100

Propane dose (kg ha-1) Propane dose (kg ha-1)

Cro

p inj

ury

(%

)

1 DAT28 DAT

Cro

p inj

ury

(%

)

SE

FN

BS


Table 1. Propane doses (kg ha-1) that resulted in 5%, 10% and 20% injury in three growth stages of winter wheat based on visual ratings from 1 and 28 DAT (Fig. 1). Regression parameters were estimated using equation 1


--------------------------- kg ha-1---------------------------

Growth stages

DAT aB (SE)

bI50 (SE) ED5 (SE) ED10 (SE) ED20 (SE)

1 -2.4 (0.4) 40 (2) 12 (3) 16 (3) 22 (3) Shoot elongation

28 -0.9 (0.3) 336 (218) 14 (8) 32 (9) 76 (13)

1 -4.2 (0.8) 65 (3) 32 (3) 39 (3) 47 (3) First node

28 -2.6 (0.4) 77 (3) 25 (4) 33 (4) 45 (3)

1 -2.2 (0.4) 65 (4) 17 (4) 24 (5) 34 (5) Boot stage

28 -4.6 (0.9) 75 (2) 40 (2) 47 (2) 56 (2) a The slope of the line. b The dose resulting a 50% visual damage.

Reduction of grain yield increased with increase in propane dose at each growth stage. The SE stage exhibited the least yield reductions from flaming compared to the FN and BS stages (Fig. 3). The yield loss curves suggested that the highest yield losses of about 78, 56 and 31% were achieved with the propane dose of about 100 kg ha-1 applied at BS, FN and SE stages, respectively (Fig 3). Our previous studies reported that a minimum propane dose of 60 kg ha-1 was needed to control most annual broadleaf weeds and many grasses (Knezevic & Ulloa, 2007; Domingues et al., 2008). Such propane dose can cause about 25%, 32% and 43% yield reduction for SE, FN and BS stages (Fig. 3), respectively, which is likely not going to be acceptable by the organic wheat producers. Table 2. Regression parameters for each growth stage of wheat as influenced by different doses of propane (Fig. 3). Regression parameters were estimated using equation 2

Growth stages I (SE) A (SE) Shoot elongation 2.1 (0.7) 31.3 (4.4) First node 1.3 (0.3) 55.5 (15.5) Boot stage 1.6 (0.3) 78.1(17.8)

I = percentage of yield loss per unit of propane applied. A = percentage of yield loss as propane dose applied approaches infinity.

Although this was only a fist year study, the results clearly demonstrated that the SE stage of winter wheat was more tolerant to broadcast flaming than the other studied growth stages, and BS was the most susceptible stage with highest yield losses.

From the practical standpoint, we believe that the presented injury levels and crop yield reductions would not be acceptable by the organic wheat producers. Such high levels of winter wheat injury also suggested that there is a need to evaluate additional timings for flaming procedures, perhaps relative to the earlier growth stages of the crop. For example, flaming winter wheat at the time of first few leaf stages (V2-V4), or at the time of tillering might be much safer for the crop. Studies are needed to test such hypothesis.


Figure 2. Influence of propane flaming on relative yield and yield components of wheat as affected by growth stages.

0 20 40 60 80 100

0

20

40

60

80

100

Relative yield (%)

Propane rate (kg ha−1)

Rel

ativ

e yi

eld

(%)

a) SESb) FNSc) BS

0 20 40 60 80 100

2.0

2.5

3.0

3.5

4.0

100 seed weight (g)

Propane rate (kg ha−1)10

0 s

eeds

wei

ght

(g)

0 20 40 60 80 100

400

500

600

700

800

Spikes m−2


)

Spi

kes

m−2

0 20 40 60 80 100

10

12

14

16

18

20

22

24

Seeds spike−1


)

See

ds s

pike

−1



Rela

tive

yie

ld (

%)

100-s

eed w

eig

ht

(g)

Seed

s sp

ike

-1

Spik

es m

-2

Spikes m-2 Seeds spike-1

Relative yield (%) 100-seed weight (g)

SE FN BS


Figure 3. Wheat yield loss (%) as a function of propane dose in three growth stages. The regression lines are plotted using equation 2, and the parameter values are recorded in Table 2.

References Cousens R (1985) An empirical model relating crop yield to weed and crop density and a statistical comparison with

other models. The Journal of Agricultural Sciences 105, 513-521. Domingues AC, Ulloa SM, Datta A and Knezevic SZ (2008) Weed response to broadcast flaming. Review of



Johnson WC (2004) Weed control with organic production. Proceedings of the Southeast Regional Fruit and Vegetable Conference, Savannah, Georgia, pp. 13-14.



Lague C, Gill J, Lehoux N and Peloquin G (1997) Engineering performances of propane flamers used for weed, insect pest, and plant disease control. Applied Engineering in Agriculture 13, 7-16.

Lague C, Gill J and Peloquin G (2000) Thermal control in plant protection. In Physical control methods in plant protection/La lute physique en phytoprotection. Edited by Vincent C, Panneton B and Fleurat-Lessard F, Springer-Verlag, pp. 35-46.

Parish S (1990) A review of non-chemical weed control techniques. Biological Agriculture and Horticulture 7, 117-137.

Parsons JL (2008) Nebraska Ag. Rank and Ag. Business Facts. National Agricultural Statistics Service (NASS). Web page: http://www.nass.usda.gov/Statistics_by_State/Nebraska/Publications/Rank_and_Agribusiness/rank2008.pdf. Accessed: December 5, 2008.

R Development Core Team (2006) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org.

Seefeldt SS, Jensen JE and Fuerst EP (1995) Log-logistic analysis of herbicide dose-response relationships. Weed Technology 9, 218-227.

0 20 40 60 80 100

0

10

20

30

40

50

60


)

Yie

ld L

oss

(%)

a) SESb) FNSc) BS


Yie

ld los

s (%

)

SE

FN

BS


Stopes C and Millington S (1991) Weed control in organic farming systems. Proceedings of the Brighton Crop Protection Conference - Weeds, Brighton, UK, pp. 185-192.

Teixeira HZ, Ulloa SM, Datta A and Knezevic SZ (2008) Corn (Zea mays) and soybean (Glycine max) tolerance to broadcast flaming. Review of Undergraduate Research in Agricultural and Life Sciences Vol. 3, Issue 1, Article 1. http://digitalcommons.unl.edu/rurals/vol3/iss1/1.


Walz E (1999) Final Results of the Third Biennial National Organic Farmers’ Survey. Organic Farming Research Foundation, Santa Cruz, CA.



Dose-response of weeds to flaming

Leblanc, M.L.1, Cloutier, D.C.2 and Sivesind, E.3 1 Institut de recherche et de développement en agroenvironnement, Saint-Hyacinthe, QC; 2 Institut

de malherbologie, Beaconsfield, QC; 3 McGill University, QC.

A two year study was conducted to determine the thermo-sensitivity of 4 weed species: Amaranthus retroflexus, Chenopodium album, Echinochloa crusgali and Setaria pumila to propane flaming based on propane doses (speed at 2, 3, 4, 5 and 6 km h-1 ; pressure-rate at 2,7, 4,3 et 5,9 kg h-1 giving propane doses varying from 0,4 to 3 g m-1) and weed growth stages. At least 20 seedlings per growth stage, per species, and per dose were tagged to determine their response to flaming. Weed thermo-sensitivity increased with the dose but decreased as the weed growth stages increased. More than 90 % of the dicotyledonous seedlings at cotyledon, 1 or 2 leaf-stages were thermo-sensitive to doses less than 1 g m-1 whereas at the 6-leaf stage, a propane dose of at least 3 g m-1 was needed to reach a similar level of weed control. The grasses tested were not thermo-sensitive to propane flaming at the doses used in this project. The 1- to 3-leaf-stages had some thermo-sensitivity with a mortality varying between 30 to 50 %. This low response to flaming is caused by a regrowth of the grass seedlings since their growing point is under the soil surface when flamed.


Solarization as a tool for non-chemical weed management

Baruch Rubin RH Smith Institute of Plant Science and Genetics in Agriculture, RH Smith Faculty of Agriculture,

Food and Environment, The Hebrew University of Jerusalem, Rehovot 76100, Israel ([email protected])

Soil solarization or solar heating of the soil (SH) is an effective non-chemical method for weed control that can serve as partial substitution for the banned methyl-bromide. Soil solarization is based on utilizing the solar energy for heating soil mulched with a transparent polyethylene (PE) reaching a level of 40-55 ºC in the upper soil layer. There is a gradient of temperatures from upper to lower soil layer during the appropriate season. The temperature elevation and the weed control process are facilitated by wetting the soil before and/or during mulching with the PE sheet. The main factor involved in the pest control process is the physical mechanism of thermal killing that may act in combination with chemical and biological processes. The duration of soil mulching required for successful effect is usually four to six weeks, depending on the weed, soil characteristics, climatic conditions and the PE properties. The use of organic amendments (biofumigation), such as animal manure, or incorporated cover crop residues combined with soil solarization may further elevate the soil temperature by additional 1 to 3ºC. Biofumigation leads to the generation of toxic volatile compounds such as ammonia, methanethiol, dimethyl sulfide, allylisothiocyanates, phenylisothiocyanates and aldehydes that accumulate under the plastic mulch and consequently enhance the sensitivity of soil organisms to soil solarization. These accumulated compounds further deteriorate the weed seedbank persistence. Weeds differ in their response to SH: some exhibit high tolerance and some are very sensitive. Seeds of perennial weeds are more susceptible to SH than propagules. In addition, winter annual weeds are more sensitive than summer annuals and dormant and non-dormant seeds respond differently to SH. Physical dormancy is a common factor that determines the tolerance of weeds to SH. Compact seeds (small, smooth and globular seeds) tend to penetrate to the deeper soil layer outside the reach of the heat. Thus, soil disturbance after SH may exhume these seeds and thereby reducing the effect of SH. However, large seeds that are deeply buried are able to emerge without soil disturbance. The major constraints that limit the adoption of soil solarization in practice are the relatively long duration of the process and the climatic dependency. The cost, of solarization is relatively low compared with other available alternative. However in some countries it can be a limiting factor depending on to the crop type, the production system (e.g., organic versus conventional farming) and the availability of alternative methods. Hence, due to the high cost associated with the polyethylene mulching the method is particularly suitable for use in cash crops.


Effects of soil steaming on weed seed viability

F. Vidotto1, M. Letey1 and D. Ricauda-Aimonino2 1Dipartimento di Agronomia, Selvicoltura e Gestione del Territorio, University of Torino, Italy

Email: [email protected] 2Dipartimento di Economia e Ingegneria Agraria Forestale e Ambientale, University of Torino, Italy

A study aimed at evaluating the efficacy against weeds of soil steam application have been carried out both in the field and laboratory in Italy in 2008.

Field trials were conducted in two sites using a self-propelled machine equipped with a multi-injector steam distribution system. It consisted of 96 injectors arranged in a 2x2.5 m2 plate. In each injector four holes ensured steam deliver at a single depth of about 18 cm. Steam applications of 6, 8, and 10 minutes duration were compared on plots with area ranging from 5 (site 1) to 25 m2 (site 2). Temperature data at several depths up to 16 cm were collected at 30 sec intervals for about 2 hrs, starting from beginning of steam application, by means of appropriate multi-channel temperature probes, connected to a datalogger. Effects on weed seeds were estimated from emergences recorded both in the field and from soil samples (0-15 cm layer) collected before and 12 hrs after the treatment.

In the laboratory, 15-min steam treatment was performed on a soil volume of about 40x40x35 cm by using either superficial application (A) or sub-soil injection at 12 cm depth (B). The two systems of steam application were compared in terms of temperature dynamics in the soil and effects on germinability of Echinochloa crus-galli, Setaria viridis, Solanum nigrum, Galinsoga ciliata, Chenopodium album e Portulaca oleracea seeds kept buried in bags at 2, 7, 13 and 19 cm during the treatment.

Main weeds found in the field trial were G. ciliata and P. oleracea in both sites. Good levels of heating were already obtained after 6 min of steam supply in the entire considered slab of soil, with a mean temperature of about 60°C measured one hour after the end of the treatment. In comparison to untreated plots, emergences recorded in the field were reduced by 89% to 95% (65% to 70% in terms of biomass) and by 100% in site 1 and 2, respectively, with no correlation with the treatment duration. Emergences from soil samples were reduced by 75%, 84%, and 93% at 6, 8, and 10 min duration, respectively.

In the laboratory trial, steam application with system A resulted in a large non-linear temperature gradient from surface (about 90 °C) to the deepest layer (about 40 °C) that prevented germination in seeds buried up to 7 cm; at 13 cm germinability was reduced by 50%, at 19 cm germinability was generally not affected, excepting seeds of E. crus-galli, which were stimulated. System B reduced germinability of all species by at least 80% in seeds buried at up to 13 cm; from surface to 13 cm-depth, loss of germinability was inversely correlated to the distance from the injector and apparently unrelated to depth of burial. This is likely to be attributable to the homogeneous heating due to the upward steam flow occurring in the bulk of soil close to the injector.

The results of this study pointed out that soil steaming is an effective method to reduce weed seed germination. As observed for E. crus-galli, sub-lethal temperatures, which can occur on the margins of the treated areas, can contribute to break seed dormancy, thus increasing the overall emergences.


Various weed control systems


Autonomous navigation in weed-infested maize fields

Abadia D.1, Ballano S.1, Uson F.1, Paniagua J.1, Seco T.1, Cirujeda A.2 and Zaragoza C.2

1Instituto Tecnológico de Aragón, Gobierno de Aragón , Maria de Luna 8, 50018 Zaragoza, Spain 2Centro de Investigación y Tecnología Agroalimentaria de Aragón. Ave. Montañana, 930. 50059

Zaragoza

The work presented in this paper is in the context of the research project SAAPIN (Autonomous System for Precision and Integrated Agriculture), whose aim is to develop an autonomous system able to navigate through maize field and perform weeding operations.

A small tractor was selected and adapted to perform autonomous navigation, by integrating sensors to get necessary information to navigate (environment: vision, GPS, odometry) and by performing mechanical adaptations (direction and transmission). Due to the need of performing intra-row-weeding the tractor navigates directly over the crop row.

A simulation environment allows to design the best suited algorithm strategy by taking into account the needed sensors and by taking into account the tractor, the weeding implement model and the environment defining the maize field conditions (dimensions and sowing distances) and the sensors model. Several navigation strategies were analyzed. An algorithm to cover the whole extension in a robust way was proposed by taking into account the required sensors.

Afterwards a vision-based perception system has been chosen to identify the maize crop rows with weeds, to let the system navigate autonomously, the final goal is to achieve the relative position to the closer crop row, by determining the offset distance and the relative angle.

The perception system is based on a color camera mounted on the tractor. The algorithm to identify crop rows consists in several phases. After the image has been taken, the distortion is corrected, and a segmentation of the vegetation (crop and weeds) is performed based on the hue component H (HSV space). The image is then processed to minimize the distortion effect of the weeds towards the crop row identification. Therefore a fill operation is performed (to fill up the internal holes) and a dynamic opening operation (erosion and dilation) depending on the distance is applied, with the final goal to minimize the overlaps between crop and weeds and to eliminate the very small weeds.

The contours of the crop plants are obtained based on the Canny operator as they have structural and geometrical information of the detected image components. Based on the contours, the probabilistic version of the Hough transformation is applied, defining minimum length of valid crop segments and taking into account the concatenation among aligned segments. Once the segments are obtained they are projected by means of the homography to the terrain plane, where several rules are applied based on the prior knowledge of the maize distribution. The obtained lines are grouped into a group representative. Based on the axis of the tractor and the projection of a virtual tractor point, the closest crop row is calculated and also the offset distance and the relative angle to the maize crop row.

References Åstrand B., Baerveldt A. J., A vision based row-following system for agricultural field machinery, Mechatronics,

Volume 15, Issue 2, March 2005, pages 251-269 Leemans V., Destain M.-F., Line cluster detection using a variant of the Hough transform for culture row localisation,

Image and Vision Computing 24, 2006, pages 541-550


Vision based crop plant identification for weeding operations

Abadia D.1, Gonzalez S.1, del-Hoyo R.1, Paniagua J.1, Seco T.1, Cirujeda A.2 and Zaragoza C.2

1Instituto Tecnológico de Aragon, Aragon Government, Maria de Luna 8, 50018 Zaragoza Spain 2Centro de Investigación y Tecnología Agroalimentaria de Aragón. Ave. Montañana, 930. 50059

Zaragoza

The context is the research project SAAPIN (Autonomous System for Precision and Integrated Agriculture), whose aim is to develop an autonomous system able to navigate through maize fields and to perform mechanical weeding operations. The scope of the presented work is to propose a novel algorithm to identify maize plants for optimizing weeding operations during the first weeks. This algorithm is based on the stem identification of the crop plants in coexistence with weeds. The discrimination system to identify the maize crop stalks is an intelligent vision-based system. The emphasis of the weeding operation is intra-row, where the competition is more important due to the proximity between crop plants and weeds.

The autonomous system is comprised by a tractor platform, which is going to navigate autonomously and by the mechanical weeding implement, to perform the weeding operations. A small tractor was selected and adapted to perform autonomous navigation by integrating sensors and by performing mechanical adaptations. Due to the need of performing intra-row-weeding the tractor navigates directly over the crop row. The weeding operation will be done by vertical axis rotary brushes. The concept of the weeding system is presented with two main tasks: to discriminate between crop and weed, and to act on the weeds. The discrimination task will be performed by a vision perception system, where the maize plants are identified and security areas without weeding are defined around them. The discrimination system proposed is placed covering the scene providing artificial light to have controlled light conditions. A color camera is used and the resolution and height was determined by taking into account the problem features.

The proposed algorithm is comprised of several phases: first an image is taken, afterwards the vegetation (crop and weed) is segmented based on chromatic indexation. Afterwards morphological operations are applied to modify the shape of the image objects to compensate crop defects and to minimize the existent overlaps between crop and weeds as well as the weed itself. Once very small weeds are discarded, morphological descriptors of each image component are extracted to identify its shape and to discriminate morphologically maize plants in the different growth states. Several classifiers were analyzed to determine which one requires the minimum number of parameters, and which was the most efficient in terms of implementation and classification. The classifier finally chosen was Ripper. Once the classification takes place, the maize stems have to be identified. The maize crop detection is based on a set of geometric rules and morphological operations whose aim is to determine the lines which fit better with leafs which allow to determine the stem position. The stem search is based on the maize crop structure by taking into account that the stem is closed to one of the leaf extremes. Trajectories and safe areas for the rotary brushes are generated based on the position of the stems.

References Åstrand B., Baerveldt A. J., A vision based row-following system for agricultural field machinery, Mechatronics,

Volume 15, Issue 2, March 2005, Pages 251-269 Southall B., Hague T., Marchant J.A., Buxton B.F., An autonomous crop treatment robot. A Kalman filter model for

localization and crop/weed classification. The international journal of robotics research, 2002 Vol.21, No1, pp. 61-74.


Control of docks (Rumex spp.) in organic fodder production – experiments for optimizing the soil tillage effect on docks when renewing highly infested areas

L. O Brandsæter1&2, K. Mangerud3 1Norwegian Institute for Agricultural and Environmental Research

Høgskoleveien 7 and 2Norwegian University of Life Sciences, Department of Plant and Environmental Sciences, 1432 Ås, Norway. 3 Hedmark University College,

N2322 Ridabu, Norway, [email protected]

Control of dock species are a true bottleneck in the development of grassland based organic forage production in Norway. Rumex obtusifolius, Rumex crispus and Rumex longifolius are among the most important perennial weeds in grassland areas throughout the world. These dock- species are undesired in grasslands because they decrease yields and reduce forage feeding value. The experiment in our study is carried out as a full-factorial design, including key-factors, which may influence dock behaviour significantly. The first factor, (i) date of grassland establishment, may be important for preventing /decreasing the flush of seedlings from seeds as well as shoots from root fragments. The purpose of the second factor, (ii) black fallow, is both false seedbed preparation and decreasing food reserves in underground plant parts. The third factor, (iii) is the use of equipment for cutting the taproot either (a) before ploughing by using a tractor propelled rotovator, or (b) cutting the dock taproot in the same operation as ploughing by using a prototype “two layer dock-plough”. The biological background for cutting the taproot before ploughing is that many studies have shown that new shoots only come from the 5 upper cm of the taproot. Furthermore, our hypothesis is that shoots from highly fragmented regenerative parts (the neck) of the taproot placed deep will not reach the soil surface before their reserves are depleted. Experiments were carried out at 3 and 4 locations in 2007 and 2008, respectively. Weed development were assessed as number of emerging seedlings as well as number of sprouting plants from root fragments, both in the year when the treatments were carried out and the following year. The results are yet not completely analyzed, but preliminary results indicate that plants from seeds frequently are more numerous than plants from roots. At least at some locations and years both the use of rotovator and the “dock plough”, has reduced the number of plants from root fragments with approx. 50%. However, our experiments have shown that “dock plough” prototype has to be improved, especially because it did not cut the taproot near the open furrow, and did not bury the green parts well enough.


Evaluation of different mulches for weed control in processing tomato

A. Cirujeda1, J. Aibar2, A. Anzalone3, M. Gutierrez4, S. Fernández-Cavada4, A. Pardo5, Mª L. Suso5, A. Royo5, L. Martín-Closas6, J. Costa6, A. M. Pelacho6, M.M. Moreno7, A. Moreno7, R.

Meco7, I. Lahoz8, J.I. Macua8 and C. Zaragoza1 1Centro de Investigación y Tecnología Agroalimentaria (Gobierno de Aragón), Avda. Montañana

930; 50059 Zaragoza. Spain ([email protected]) 2Escuela Politécnica Superior de Huesca, Carretera de Cuarte s/n; 22071 Huesca. Spain.

3Dept. Fitotecnia. Universidad Centroccidental “Lisandro Alvarado”. Apdo. Postal 400 Venezuela. 4Dept. Agricultura y Alimentación, Gobierno de Aragón, Avda. de Montañana 930; 50059 Zaragoza

5CIDA. Ctra Mendavia-Logroño NA-134 km88; 26071 Logroño. Spain. 6Dep. d´Hortofruticultura, Botànica i Jardineria, Alcalde Rovira Roure 191; 25198 Lleida. Spain.

7Centro “El Chaparrillo”. SIA Junta de Castilla-La Mancha. C/ Alarcos 21; 13071 Ciudad Real 8Finca Exp. del Gobierno de Navarra (ITGA); Camino Alfaro s/n; 31515 Cadreita. Spain.

A two-year study testing alternatives to the use of black polyethylene mulch (PE) is presented in this work. PE remaining in the field after the harvest is a waste difficult to manage both in conventional and organic agriculture. During the years 2006 and 2007 ten field trials have been carried out on processing tomato at five different Spanish locations. Different biodegradable alternatives have been tested: two biodegradable plastics (Mater-Bi® and Biofilm®), an oxobiodegradable film material (Enviroplast®), two papers (black Mimcord® and brown recycled Saikraft®), an organic mulch with barley straw, PE and two control treatments (unweeded and manual weeding). Drip irrigation was used in all trials and different mulches irrigated individually to avoid that some treatments could have more water than the others. All films were placed on the soil mechanically but especially the brown paper needed a special adjustment to avoid cracks.

Despite the differences in weed composition (Amaranthus retroflexus, A. blitoides, Chenopodium album, Convolvulus arvensis, Sonchus oleraceus as main species) and density, in all locations and both years weed control was good or excellent for all mulches excepting the straw. Tomato yield was very similar for all mulch treatments and both years but slightly higher for PE that provided excellent weed control. The two biodegradable plastics and the black paper have been very productive treatments with a very good control. Despite the unsatisfactory weed control (it was difficult to maintain the straw on the soil in some locations due to wind dispersal) straw mulch yielded high in some locations. The biodegradable plastics started their decomposition when the crop covered sufficiently the soil and only slight differences were observed among materials and locations. The buried part of the materials decomposed first for the papers and caused fractures in the aerial part, when strong wind blew, in some locations. The oxobiodegradable plastic had a very irregular behaviour among locations. The buried part of this material did not degrade in any case. In 2007 mean yield was slightly lower for the brown paper probably due to the lower temperatures of the season. The conclusion is that technically viable alternatives exist to substitute the PE mulch in processing tomato but it is necessary to take into account the economic costs of these materials, which are in some cases 3 to 4-fold the price of PE.


Is yardwaste mulch a weed-free substrate?

O. Daugovish1, B. Faber1 and J. Downer1 1University of California Cooperative Extension Farm Advisors, 669 County Square Drive, Suite

100, Ventura, CA 93303, USA, Email: [email protected]

Experiments conducted at Oxnard, CA, USA compared survival of seed of Malva parviflora (little mallow) and Medicago polymorpha (California burclover), rhizomes of Cynodon dactylon (bermudagrass) and tubers of Cyperus esculentus (yellow nutsedge) in 7.6 m3 static piles of freshly ground mulch and 18 months aged mulch. Heat resistant permeable bags with weed propagules were placed at 0 (surface), 0.15, 0.3 and 1 m depths in the mulch piles and removed at 0.25, 1, 2, 4, 7, 14, 21, 28 and 56 d. The experiment was repeated three times (fall, winter and spring) and the patterns of weed survival in mulch were similar among the seasons. All weeds were killed in freshly ground mulch after 2 d at 1 m and after 7 d at 0.3 m, however, germination and viability (for M. parviflora, due to high dormancy) were variable at 0.15 m and not affected at 0 m. Temperatures greater than 60 C generated at depths greater than 0.3 m in freshly ground mulch were most likely responsible for destruction of weed propagules. Weed germination and viability at all depths and removal times were not affected in aged mulch. Aged mulch, which has previously completed microbiological composting, did not heat up following pile creation (with exception of 1 m depth that did heat up to 50 C). Thus, lethal temperatures achieved during fresh mulch composting are essential for weed propagule destruction. However, the exact temperatures causing complete or partial mortality of studied weeds were not determined, thus, a controlled environment laboratory study was carried out. That laboratory study examined survival of the propagules of the four weeds at 80, 65, 50, 35 and 20 C in water saturated paper towels after 1, 3, 5 and 7-day exposure. Linear relationships between survival of the studied weeds and time and temperature provided good explanation of variability, except for M. polymorpha seed. With exception of burclover burs, no significant effect of time was observed in the studied 1 to 7 day interval and temperature range. The resulting relationships were:

M. polymorpha burs, % germination = 99.5 – 2.27 Time (d) – 1.17 Temperature (C) (R²=0.78) M. polymorpha seed germination, % = 20.58 – 0.24 Temperature (C); (R²=0.33) C. dactylon germination, % = 111.96 – 1.6 Temperature (C); (R²=0.73) C. esculentus germination, % = 133.42 – 1.9 Temperature (C); (R²=0.75) M.parviflora viability, % = 114.5 – 1.6 Temperature (C); (R²=0.76) Considering relative ease with which temperatures can me measured, these equations provide

useful estimates of levels of weed survival expected at least 1 day following exposure to the particular temperature (and time, in case of burclover burs). For example, 72 C for at least one day is needed to reduce viability of M. parviflora to 0%, thus, ensuring that the mulch will be free from this weed.

The lethal temperatures established in the lab experiment corresponded closely with those recorded in the field study with fresh much, however, temperature non-uniformity in constructed piles and, therefore availability of safe sites for weed survival may account for rare occurrence of viable weed propagules at what would be expected lethal temperatures. Most importantly, the studied troublesome weeds are likely to survive on the surface or at depth less than 0.3 m in fresh mulch piles. It is essential, therefore, to mix the mulch and expose the initially surviving weeds to lethal temperatures normally existing at depths greater than 0.3 m. These studies also showed that if re-infested, aged mulch has no mechanisms to suppress weeds and therefore may become a weed carrying substrate.


Management issues related to the use of the herbicide glyphosate and the transformations operated on urban vegetation in Genoa (Northern Italy). First

note.

A. Di Turi and G. Paola Università di Genova, DIPTERIS, Orto Botanico Hanbury. Corso Dogali 1M, I - 16136 Genova

ITA Email: [email protected]

Since many years undesirable species are being controlled in urban areas of Genoa (Northern Italy), through the use of glyphosate: transformations caused on urban vegetation by the intense use of herbicides are being currently analyzed.

Some comparisons were done, in first place, between the floristic populations living in few sample areas and the situation of the same areas about 15 years ago (G. Barberis and A. Di Turi, 2000), when glyphosate was not being used yet. The present study has been revealing the heavy reduction or disappearance of some species and the explosive appearance of others, whose presence was much more limited in the past. Among these we may list Parietaria judaica L. which, in addition to being unattractive, is unfortunately also one of the most intensely allergenic species in the Mediterranean area. The proliferation of this species is now creating an important urban management problem.

The main targets of this study include: reducing and, if possible, replacing chemicals in the green urban management of Genoa, to

inhibit the proliferation of urban weeds; restraining the spreading of Parietaria judaica and other allergenic species, especially in

sensitive sites (hospitals, school gardens, etc.). In order to explore the possibilities of restraining the widespread proliferation of Parietaria

judaica, the research is based on two different and integrating approaches. The first line of the research aims at evaluating methodologies to introduce other species in

those habitats where Parietaria judaica is predominant, especially stone walls, in order to deprive it of some of its ecological spaces.

Amongst the species commonly found on Italian urban walls (Hruska, 1987) and well settled on stone walls in Genoa, we picked the exotic Erigeron karvinskianus DC., and the native Cymbalaria muralis s.l. Gaertn. B. Mey & Scherb. and Viola reichenbachiana Jord. ex Boreau. Data collected so far show that Erigeron karvinskianus seems to be able to successfully limit Parietaria judaica in such an environment: it has turned out to be the easiest to grow and has the best chance of taking root in urban walls (A. Di Turi and G. Paola, 2008).

Other tests are being carried out on smilograss (Piptatherum miliaceum (L.) Coss. s.l.), a grass that forms large clusters which seem to suffocate Parietaria judaica in green areas. Phytosociological relevèes carried out in six sample plots in urban areas have shown a progressive, even if slow, increasing of smilograss to Parietaria disadvantage.

The second line of the research focuses on testing alternative mechanical weed control actions or combining them with the chemical methods. Still the aim is to try to gradually reduce the latter, especially in sensitive sites. At this time experiments for controlling Parietaria are carried out with the help of different kinds of mulching (Pinus sp. pl. needles and holm-oak leaves) in some sample plots. These kinds of actions seem to be giving good results, especially on keeping down Parietaria seedlings. References Barberis G. and Di Turi A. (2000). Studio preliminare sulla flora urbana di Genova. (Unpublished). Di Turi A. and Paola G. (2008). First note about the management of exotic and native weeds on the urban stone walls of

Genoa (Liguria, Northwestern Italy). Bocconea (in press). Hruska K. (1987). Syntaxonomical study of Italian wall vegetation. Vegetatio 73: 13-20.


Possibility of using mustard meal of Sinapis alba and Brassica juncea for weed control

Fredrik Fogelberg JTI - Swedish Institute of Agricultural and Environmental Engineering, POB 7033, SE-750 07

Uppsala, Sweden. E-mail: [email protected]

The use of allelopathy or allelopathic substances for weed control has until today not been considered as a useful tool for large-scale cropping. However, in order to build crop production systems with minimised input of agrochemicals the interest for weed sanitary crops or “natural” weed control substances begins to rise. Experiments in Sweden in the early 1990’s showed that mustard meal had a weed control effect, but additional research has not been carried out until now.

Objectives The objective was to investigate weed control effect and effect of treatment time and doses of

products based on Sinapis alba or Brassica juncea.

Material and methods In 2007 we investigated the weed control effect of various mustard-based products. Mustard

meal of Sinapis alba and Brassica juncea were mainly used, but we did also use grinded residues from mustard oil production. In the main field survey 36 combinations of products, doses and application times were used in plots of 2 x 2 m. Annual weeds were assessed in number and fresh weight after treatment.

In addition to the survey we also studied effects of mustard meal on yield in direct-seeded bulb onions and transplanted broccoli. In these crops we studied doses and application time and the effect on weed number, fresh weight and crop yield. Generally we used a randomised block structure with 4-6 replications of each treatment.

The mustard meal products we used were obtained from a range of sources; the main product was meal from Sinapis alba cultivars grown in Sweden intended as a basis for production of mustard. Another product was meal from Brassica juncea, cv Com Brown, also used for mustard production.

In the survey our intention was to study doses, type of product, application time (before or after weed emergence) and related questions. In the subsequent study in broccoli and onions, we selected a few products, mainly medium ground seeds of Sinapis alba, and used them in four doses (0; 500; 1000 and 2000 kg ha-1). Application of the meal was carried out after crop emergence respectively after transplanting.

Results and discussion Generally, we noticed that mustard meal resulted in a lower fresh weight (40-65% reduction) of

annual weeds, but not in reduced number of weeds. Moreover, the mustard meal is obviously a perishable product, it should be spread quite rapidly in order to have effect and preferable on moist soil, before emergence of weed seedlings.

Meal based on Brassica juncea tended to give a higher weed control effect compared to meal based on Sinapis alba, but the effect was only seen as a reduction in fresh weight, not in weed number. Typically an untreated plot had a total weed fresh weight of 565 g whereas a treated plot resulted in a weed fresh weight of 220 g. The weed flora mainly consisted of Chenopodium album, Capsella bursa-pastoris, Stellaria media, Viola arvensis and Myosotis arvensis.

In bulb onions two application times were studied; 12 and 16 days after seeding. There were no differences in weed number between treatments, but the early treatment had a significantly (p=0.03)


lower weed fresh weight. The control effect was however quite low, 48 % reduction compared to non-treated.

Yield of bulb onions was not recorded due to lack of funding. In transplanted broccoli the yield was not affected by mustard meal in any of the studies. We did not store the broccoli after harvest and thus cannot make any statements on possible effects on storability.

An interesting treatment was to let mustard meal soak in tap-water for a few hours and then spread the water on bare soil. This treatment reduced weed fresh weight of annual weeds by 73-89 % using 2500-10 000 L ha-1.

It is obvious that mustard meal contains compounds that influence weed development, but these compounds, or the applied doses in the experiments, are not as effective as anticipated. However, we used mustard cultivars intended for production of mustard for human consumption where glukosinolates and other growth inhibiting substances generally are low. A recent topic of agricultural science is biofumigation, i.e. use of mustard plants that, when incorporated into the soil creates a fumigation of the soil controlling nematodes and possibly fungi.

An interesting approach in a new experiment on mustard meal would be to use seed of cultivars intended for biofumigation or possibly a ground mixture of whole plants and seeds that is spread as in our reported study.

How can we use the results in practice? As mustard meal can reduce weed weight but not weed number, application of mustard meal

could probably be used at early stages of crop development in order to favour crop plant competition.

A solution of “mustard-water” could be used for weed control on hard-surface areas. It might be possible to use the mustard-water prior to seeding of vegetables or as a post-emergence treatment. Additional experiments should be carried out to study effects on crop plant seedlings.

Acknowledgement The project was funded by the Ekhaga Foundation. The field experiments have been conducted

at the Torslunda Experimental Station and at the Petersborg Farm in Sweden.

Literature Beckie, H. J.; Johnson, E. N.; Blackshaw, R. E. & Gan, Y. 2008. Weed suppression by canola and mustard cultivars.

Weed Technoloy 22, 182-185. Bialy, Z.; Oleszek, W.; Lewis, J.& Fenwick, G.R. 1990. Allelopathic potential of glucosinilates and their degradation

products agains wheat. Pland and Soil 129: 277-281. Boydston, R.; Vaughn, S. & Anderson, T. 2008. Mustard (Sinapis alba) seed meal suppresses weeds in container grown

ornamentals. HortScience 43, 800-803. Bugg, R.L. 1995. Cover crop biology: A mini-review. Univ. California Sustainable Agriculture Research and Education

Program. Jaakola, S. 2005. White mustard mulch is ineffective in weed control. In: Proceedings of the fourth world congress on

allelopathy, 21-26 August 2005, Charles Sturt University, Australien. Johansson, H. 1992. Ogräsbekämpning i grönsaksodling med vitsenapsexpeller. SLU Info/trädgård rapporter. Sveriges

lantbruksuniversitet, Alnarp, Trädgård 371. (in Swedish). Jonasson, T. 1990. Mustard Meal Mulching: A biological method for cabbage root fly control. Nordisk

Jordbrugsforskning 72, 453. Khan, T. D.; Xuan, T. D. & Chung, I. M. 2007. Rice allelopathy and the possibility for weed management. Annals of

Applied Biology, 1-15 Molisch, H. 1937. Der einfluss einer Pflanze auf die andere – Allelopathie. Fisher, Jena, Tyskland. Purvis, C. E.; Jessop, R. S. & Lovett, J. V. 1985. Selective regulation of germination and growth of annual weeds by

crop residues. Weed Research 25, 415-421. Rademacher, B.; Kolb, F. & Börner, H. 1961. Untersuchungen über die gegenseitige Beeinflussung von Kulturpflanzen

und Unkräutern in Wasserkultur. Weed Research 1, 44-58. Turk, M. A. & Tawaha, A. M. 2003. Allelopathic effect of black mustard (Brassica nigra L.) on germination and

growth of wild oat (Avena fatua L.). Crop Protection 22, 673-677. Wolf, R.; Spencer, G.F. & Kwolek, W.F. 1984. Inhibition of velvetleaf (Abutilon theoprasti) germination and growth by

benzyl isothiocyanate, a natural toxicant. Weed Science 32, 612-615.


Fig. 1. A wide range of mustard products, doses and application times were studied in a pilot experiment.


Fig. 2. Mustard meal was tested in transplanted broccoli in order to study weed control effect and impact on yield.

Fig. 3. The mustard meal was spread manually by a simple grass seeder.


Impact of weed-control mulches and disks on fertilizer placement and water use in nursery container production

M. Lanthier 1, S. Peters 1, J. Atland 2, S. Harel 1 1 CropHealth Advising & Research, Kelowna BC, V1W 4A6, Canada Email

[email protected] 2 USDA-ARS, Ohio Agricultural Research and Development Center, Wooster OH 44691, USA

A series of trials were conducted from 2003 to 2008 at commercial nurseries in British Columbia (Canada) and Oregon (United States). Hydrangea rooted liners were manually potted in early spring into standard containers filled with commercial growing media. Treatments were either commercial products or readily available materials, including sawdust (1.25 cm thickness), Biotop (1.25 cm thickness), crumb rubber (1.25 cm thickness), corn gluten (2.0 grams on the surface), moulded plastic lid (one disk per container) and woven coco fiber disk (one disk per container). Data were subjected to analysis of variance and, were significant, means were separated with Tukey’s LSD (p<0.05 level).

In 2003, treatments with significantly fewer weeds than the hand-weeding control were sawdust, biotop, plastic lid, coco-fiber disk and crumb rubber. Corn gluten treatment had no significant impact on weed emergence. Top dry weight was statistically higher for plants grown in containers treated with coco fiber disk, plastic lid or corn gluten. Media nutrient analysis indicated no difference in concentration of heavy metals in containers treated with crumb rubber.

In 2005 and 2006, the sub-treatment was a controlled release fertiliser applied above or under the weed control treatment (Osmocote 19-5-8 at 15 grams per container or 4.5 kg / m3 of growing media). Plants were significantly larger, with higher quality ratings and foliar nitrogen, when grown in containers with controlled release fertiliser placed below the mulch or disk. After correcting for plant size, there were minor differences in water loss between mulched and non-mulched containers.

In 2008, plants were grown with a mulch (woven coco fiber disk or crumb rubber) and a sub-treatment (Osmocote 19-6-12, Nutricote 18-6-8 and Plantacote 14-9-15). Again, plants were significantly larger when grown in containers with controlled release fertiliser placed below the mulch or disk. There was no difference in nutrient content for growing media treated with either coco disk or rubber mulch. There were significant differences in residual nutrient at the end of the growing season between fertiliser prills placed above or under the mulch.

In conclusion, mulches and disks offer effective weed control in nursery container production. It is a cost effective method for indoor production of high-value plants or outdoor production of single-stem slow growing plants in regions receiving extensive rainfall. Our research indicates that fertiliser placement on top of a mulch or disk results in smaller plants compared to placement under the mulch or disk. The difference in plant growth is from reduced nutrient release from the prills and not from binding of nutrients in the mulch or disk.

References Atland J and Lanthier M (2007). Iinfluence of Container Mulches on Irrigation and Nutrient Management. J. Environ.

Hort. 25(4):234-238. Lanthier M, Peters S, Harel S, McInnis M (2006). Non-chemical weed control with mulches and disks for nursery

container production. Poster presentation. Can Weed Sci Soc. Victoria, B.C.


Weed community response to six IWM systems in a four-year crop rotation

A. Légère1, A.G. Thomas1, J.Y. Leeson1, F.C. Stevenson2, F.A. Holm3, B. Gradin4, D Kratchmer5

1Agriculture and Agri-Food Canada, Saskatoon, SK S7N 0X2 Email: [email protected]; 2Research Consultant, Saskatoon, SK; 3Plant Sciences Department, University of Saskatchewan, Saskatoon, SK; 4Pest Management Regulatory Agency, Saskatoon, SK; 5Viterra, Watrous, SK

Integrated Weed Management (IWM) implies knowledge of weeds and agroecosystem functions. IMW uses a variety of tools to keep the competitive balance in favour of the crop to the detriment of weeds. These systems are developed on a multi-season time scale with the realization that weeds will be contained but not eliminated. On the Canadian Prairies, the use of agronomic practices such as choice of crop, seeding date and rate, row width, fertilizer timing and in-crop herbicide rates has contributed to successful weed management and reduced herbicide inputs. Our hypothesis was that herbicide or tillage use could be complemented or replaced by other crop management practices to provide similar levels of weed management in Prairie cropping systems. Our objective was to determine the effects of six IWM systems on changes in weed communities in a 4-yr crop rotation.

The study was conducted at two sites: the University of Saskatchewan Kernen Crop Research Farm at Saskatoon (plot size: 4 x 20 m) and the Viterra Research Farm at Watrous SK (plot size: 7.6 x 15 m). The experimental design was a split plot with crop (wheat-canola-barley-pea) as main factor, and IWM systems as sub-factor. The six systems were as follows: HH/ZT: high herbicide/zero till; MH/ZT: medium herbicide/zero till; LH/ZT: low herbicide/zero till; LH/LT: low herbicide/low till;

MH/MT: medium herbicide/medium till; NH/HT: no herbicide/high till. The six IWM system treatments were achieved by varying: tillage timing (fall, pre-seed, in-crop) and intensity (0-11 times over 4 years); herbicide timing (fall, pre-seed, in-crop) and intensity (0-12 times over 4 years) herbicide rate (0.5, 0.6, 1.0X the recommended rate); seeding date (early, mid, late); and seeding rate (1.0, 1.3, 1.5X the recommended rate).

Weed density was assessed from 20 quadrats (0.5 x 0.5 m) prior to spring weed management treatments (1998-2000), prior to in-crop weed management (1997-2000), and after in-crop weed management (residual counts) (1996-2000). Seedbanks were assessed in the spring of 1997 and 2001, from 20 soil cores (11 cm deep, 4.5 cm diameter); samples were germinated twice in the greenhouse. Weed community change over time is compared across IWM systems with principal response curves using the MH/MT system as a reference point. Redundancy analysis was used to quantify the change in weed seedbanks due to IWM system between the 1997 baseline and 2001.

Responses of weed communities and seedbanks were similar in five out of the six IWM systems. The principal response curve for the first axis indicated a gradual increase of species such as Thlaspi arvense, Chenopodium album, Amaranthus retroflexus and Polygonum convolvulus in the NH/HT system. The principal response curve for the second axis indicated an increase in winter and early spring annuals, and perennials in most systems but particularly in the LH/ZT and MH/ZT systems. Seedbanks of most species increased over the four years in all systems but mainly in the NH/HT system. Five of the six IWM systems resulted in comparable agronomic and economic returns. However, changes in weed communities would suggest that operations and maybe even crop sequence should be revised to improve the overall management of certain weeds and reduce seed return to the seedbank in the NH/HT system and in systems with low herbicide and zero tillage. This study confirms that IWM in various forms can be successfully implemented under the challenging conditions of the Canadian Prairies.


Influence of different soil cultivation systems on weed population in soybeans

G. Malidza1, S. Vrbnicanin2, I. Kurjacki and D. Pavlovic4

1Institute for Field and Vegetable Crops, Novi Sad, Serbia 2Faculty of Agriculture, Belgrade, Serbia Email: [email protected]

3Faculty of Agriculture, Novi Sad, Serbia 4Institute for Plant Protection and Environment, Belgrade, Serbia

A goal of our research was to determine the influence of a 10-year different soil cultivation consisting of plowing, discing, and chiseling and no–till on a weed population in soybeans.

The experiment in soybeans was set up in 1989 on a soil type chernozem at a area Novi Sad (Serbia). A design used was split-plot design with 4 replications and experimental unit of 255 m2. The ground work treatments were: plowing (to 25 cm), chiselling (25 cm), discing (10 cm) and no-till. The crop systems were soybeans (monoculture) and a two-crop system (corn-soybens) within each treatment. During the experiment normal fertilizing practices were applied. In a no-till system soybeans was planted by a direct seeder „Kinze“ and for other treatments seeder „Nodet“ was used. Except for ground work as a part of this trial and herbicides, no additional weed control measures were applied. In both crop systems combination of soil herbicide alachlor + linuron was applied, and as needed foliar applications of nicosulfuron (in corn) and fluazifop-P-butil in soybeans were applied. In 1998 in a randomly chosen areas (8.5 m x 6 m) of each treatment no herbicides or any other weed control measures were applied. In these areas of each treatment we have determined a number of weed species per m2 at 1 and 2 months after planting and just before harvest.

The 10-year different soil cultivation systems, monoculture, crop rotation (soybean-corn), and application of the same herbicides have significantly influenced weed population in soybens. The lowest number of weed species were recorded after plowing (39/m2 in monoculture and 51/m2 in two-crop system at 1 month after planting), while other treatments contained 2-3 times more weeds. The highest number of perennial weed species was in a no-till system, and the lowest numbers where present where plowing was used. Weed species Erigeron canadensis was not registered where plowing was used for many years, while it was present in other treatments. Weed species Lactuca serriola was only present in a no-till system. Sorghum halepense from rizom was present in all of the treatments, and the highest population was present in chiseling and discing treatments. Crop rotation had a strong influence on population density of this weed species. Higher numbers of S. halepense were present in areas with crop rotation (soybean-corn) and lower in monoculture. Additionally, ground preparation type influenced the presence of small-seed weed species Amaranthus retroflexus and Chenopodium album. The numbers of these weed species were higher in treatments with chiseling and discing, and lower where plowing was used and in no-till treatments. In a no-till system weed species in higest numbers present was Echinochloa crus-galli (88/m2 in monoculture and 22/m2 in a two-crop system at 1 month after planting), while in other 3 treatments population was lower (1-2/m2). This difference can be explained by the influence of ground preparation on lowering the seed bank of this weed species and weaker efficacy of soil herbicides in no-till system where high amount of plant residues is present. The differences in weed population also influenced soybean yield. The highest yield was in treatments with plowing (2.3 t/ha) and the lowest in a no-till production (0.5 t/ha).


Weed population dynamics in fields with different management

D. Piliksere State Priekuli Plant Breeding Institute, Latvia

E-mail: [email protected]

The investigation on weed population dynamics in arable fields was carried out at State Priekuli Plant Breeding Institute in Latvia (57˚19΄ N, 25˚20΄ E). Weed taxa (at species or a family level) and weed abundance (plants per m-2) were determined in three 6-field crop rotations in the first decade of June during the period from 2006-2008. Crop rotations were: 1) spring cereals-spring cereals/clover-clover-winter cereals-potato-oilseed radish or rape (organic crop rotation); 2) barley-clover/grass-barley-rye-barley-potato; 3) barley-clover/grass-clover/grass-rye-barley-potato. The latter two crop rotations were established under five different fertilization systems: unfertilized, animal manure (20 t ha-1), NPK (66, 90, 135), animal manure + NPK (66, 90, 135), NPK (132, 180, 270). No herbicides were used throughout the experiments.

The aim of this research was to study out the weed population dynamics in differently managed arable fields. This information should serve as the knowledge base for making recommendations regarding control of weeds in the upland regions of NE Europe.

The amount and diversity of weeds differ among the years 2006-2008. Data from organic crop rotation show more favourable agroecological conditions for weed growth – larger amount of weeds in total and more dominant (≥10 plants m-2) weed taxa – in 2007 compared with 2008. Different observations were from conventional crop rotations. There were only few weeds and no dominant taxa in 2007. This could be explained mainly by crop rotation, where potatoes were the field crop grown in conventional fields in the given year. Comparing the years 2006 and 2008, when barley was the crop in both conventional crop rotations, more weeds and greater weed diversity was registered in the year 2006. In this case, the differences in weed population dynamics could be caused by the complex of factors (crop management, weather conditions, etc.).

In regard to weed diversity, Chenopodium spp., Thlaspi arvense and Viola arvensis were among dominant annual weed taxa in all crop rotations. Galeopsis spp., Vicia spp., Centaurea cianus and Polygonum convolvulus dominated only in conventional crop rotations, but Lamium spp., Veronica spp., Capsella bursa-pastoris, Fumaria officinalis and Polygonum aviculare – in organic crop rotation. Matricaria spp. and Stelaria media dominated in the organic crop rotation and in the conventional crop rotation with a one year clover/grass, but Spergula arvensis – only in the latter of both crop rotations. In total, just two perennial weed species – Equisetum arvense and Sonchus arvensis – were registered among the dominant weeds.

Results of this investigation build the base for further data analysis and conclusions on environmentally friendly (without using the herbicides) weed management measures. Weed population dynamics depends on different agroecological and field management factors. The interaction of these factors also should be analysed to reach the desirable result – recommendations for effective weed management.


Effect of Elymus repens on yield of winter wheat, spring barley and faba bean in an organic crop rotation experiment.

I.A. Rasmussen1, M. Sønderskov1, C. Damgaard2 and K. Kristensen3 1University of Aarhus, Faculty of Agricultural Sciences, Dept. of Integrated Pest Management,

Flakkebjerg, DK-4200 Slagelse, Denmark. Email: [email protected]. 2University of Aarhus, NERI (DMU), Dept. of Terrestrial Ecology, DK-8600 Silkeborg, Denmark,

Email: [email protected]. 3University of Aarhus, Faculty of Agricultural Sciences, Dept. of Genetics and Biotechnology,

Foulum, DK-8830 Tjele, Denmark. Email: [email protected].

The impact of crop rotation, nutrient levels and use of catch crops on effect of E. repens on a sandy soil at Jyndevad on yield of winter wheat (2006), spring barley (2007-2008) and faba bean (2006-2008) was studied in an existing organic crop rotation experiment (Olesen et al., 2000; Rasmussen et al., 2006). Some of the objectives were to determine the yield loss at different levels of infestation of the weed, and to determine whether this relationship was influenced by the treatments.

Small plots of 1 m2 were established in the larger plots of the crop rotation experiment. 2 small plots at high density of the weed, and 2 small plots at medium density were marked early in the season. In addition, 2 small plots with initial low density were kept free of the weed during the season. All small plots were kept free of other perennial weeds during the growing season, but annual weeds were allowed to grow, since this would be the normal situation in organic farming. Just prior to harvest, the central ½ m2 part of the small plots were hand-harvested, the E. repens shoots were counted, biomass was determined for this as well as for crop and annual weeds, and the crop was threshed and the biomass of kernels/seeds determined.

In this paper, the total biomass of the crop is referred to as the “yield”. For winter wheat and spring barley, there was a close relationship between total biomass of the crop and kernel yield, even across years. This relationship was not as good for faba bean. The results were analysed in a non-linear mixed model as a competition model:

)(1

)()(d

i

iym

xb

zcaaY

+×−−

=

Where Y is the total biomass of the crop (g m-2), am is a parameter for the total biomass of the crop with no weeds present as a mean of the years, ay is a parameter for the effect on yield of the year, c is a parameter for the effect of annual weeds, z is the biomass of annual weeds (g m-2), b is a parameter for the effect of E. repens, x is the number of E. repens shoots (# m-2), d denotes whether the function is linear or not, and i indicates the treatment.

For all crops, the treatments had a high impact on the yield. The two treatments that had no manure applied for up to 12 years consistently had the lowest yields. In spring barley, the two treatments with manure and with catch crops consistently had the highest yields. In faba bean, the treatment with manure and without catch crops had the highest yields. As for the effect of E. repens shoots on yield, in spring barley, there was a larger decrease in the system without grass clover. The same tendency was seen for winter wheat. For spring barley and faba bean, within each system (with or without grass clover), the yield in treatments without manure was less influenced by E. repens than in treatments with manure. References Olesen, J.E., Askegaard, M. & Rasmussen, I.A. (2000): Design of an Organic Farming Crop-Rotation

Experiment. Acta Agriculturae Scandinavica, Sect. B, Soil and Plant Science, 50, 13-21. Rasmussen, I.A., Askegaard, M., Olesen, J.E. & Kristensen, K. (2006): Effect on weeds of management in

newly converted organic crop rotations in Denmark. Agriculture, Ecosystems & Environment 113, 184-195.


Identifying weed distribution using soil properties

H. Salehian and S. Soltani Ghaemshahr Islamic Azad University, 163Ghaemshahr, Iran. Email: [email protected]

Abstract Site properties and weed species abundance are known to vary spatially across fields. The

extent to which they covary is not well understood. The objective of this research was to assess how canonical correlation analysis could be used to identify associations among site properties and weed species abundance within an agricultural field. For this reason one hundred eighty pieces of lands from a soybean farm, were chosen (hundred square meters each) at the research station of Bicola in Neka on eastern part of Sari for sampling purposes in 2007.

Eight site properties were considered in relation to six weed species that were identified and counted after all weed control operations were completed. Calculation of canonical correlation for each pair of canonical variables were shown that the first and second pairs of canonical variables, explained the majority of variation in the data and were used to identify associations among site properties and weed species abundance. The study of the first pair of canonical variable is shown which plots with extra organic carbon are included a lot of abusive weed such as Amaranthus bulbosa and Abutilon theophrasti. So, the study of second pair of canonical variable is shown that Cucumis melo and Chrozophora tinctoria most likely were in places which, there were percentage of silt and phosphorus very high.

These results show which canonical correlation analysis is one of the best methods for determination of relationship between environmental specialties and weed species abundance.

Introduction Weed infestations in crops are still a challenge that has to be met in agriculture (Diaz et al.,

2005). Typically, weeds are not evenly distributed across large agricultural fields but are aggregated into patches (Medlin et al., 2001). The spatial variability of weed abundance constitutes the basis for site specific weed management systems. Using these systems, farmers could spray selectively to reduce the amount of herbicide usage thereby diminishing environmental impact as well as economic cost (Earl et al., 1996).

Predicting the occurrence of a given weed species within an agricultural field, based on their measurable factors such as site properties, provides an opportunity to weed effective control (Dille et al.,2002). Both site properties, such as soil texture, fertility, and topography, and weed species abundance vary considerably within fields (Johnson et al., 1995; Johnson et al., 1996; Cardina et al., 1995).

Covariation of site properties and weed abundance within a field, however, is not well understood. For a number of years, researchers have investigated possible links between weed distribution patterns and the surrounding environment and, more specifically, soil properties. Soil properties have been shown to vary significantly within a scale of 150 m or less, and this variation might potentially affect the pattern and density of weed species within a field (Walter et al., 2002).

Across fields, crop type and clay content helped explain variation in weed species presence in 316 fields planted to eight crops over 3 yr in Denmark (Andreasen et al., 1991b). Within a field, Hausler and Nordmeyer (1995) reported that the visual distribution of Polygonum amphibium L. (Water smartweed) was similar to the distribution of high soil phosphorus concentration and clay content and low sand content, whereas Veronica hederifolia L. ( Ivy leaf speedwell) distribution was similar to that of sand content. The effect of soil pH on weed growth has been researched extensively. Buchanan et al., (1975) reported large crabgrass [Digitaria sanguinalis (L.) Scop] as very tolerant to low-pH soils but Jim sonweed (Datura stramonium L.) as intolerant to low –pH soils. Weed species occurrence and relative density can vary with soil nutrient availability. In a


study of 37 weed species, Andreasen et al. (1991b) found the distribution to vary for 18 species with soil organic matter, ll species with exchangeable potassium, and 7 to 8 species with pH, phosphorus, magnesium, or manganese. In the Canadian prairies, weed community groups were separated across a climatic gradient and soil zones (Dale et al., 1992). The predominant association described a relationship between weed species abundance and variation in herbicide activity, as indicated by spatial variation in the atrazine sorption coefficient (Njovac et al., 1997).

Multivariate statistical approaches can elucidate complex interactions among multiple variables. Associations among plant species are often described and classified using multivariate techniques (Breaks, 1987). Subsequent interpretation of that association is enhanced with knowledge of site properties (Post, 1988). For example, Dieleman et al. (2000) used canonical correlation analysis (CCA) to interpret association among several weed species and certain site properties by covariance based coefficients. Different canonical correlations were able to describe between 25 - 80% of weed variance and about 27-55% of variance in site properties.

Edapic factors affecting weed species abundance in North of Iran have not been studied. The objective of this research was to evaluate whether weed species abundance was consistently associated with a variety of site properties within a farmer-managed field? The multivariate technique of CCA was selected to identify possible associations.

Materials and methods Site properties and weed abundance data were collected in a 16-ha field in the Baycola

Research Center located near the town of Naka. A piece of land was chosen in this farm (1.8 ha area) and divided into 180 small pieces (100 m 2 ) (9 rows, and 20 columns) (Fig. 1). Three randomly located soil cores were taken within a 1-m circle each grid using a 6-cm- diam core to a depth of 15 cm of these, 6 properties were used in this study (Table 1).

The field was mold board plowed on November 2006 and soybean sown in 0.6-m-wide rows on June, 2007. Trifluralin. (0.9 kg ha 1− ) and Bentazon (1.3 kg ha 1− ) before sowing and for post emergence, respectively were used to weed control in soybean. Weed seedling surveys were conducted after all weed control practices were completed. At the center of each grid cell, equivalent to a 1-m 2 sampling area, weeds counted (for a total of 180 observations). Figure 1. Pattern of sampling points for site properties (●) and weed species abundance (1-m 2 sample from center of each 10 by 10 m square) in a farmer – managed field in Baycola Research Center.


Table 1. Descriptive statistics- number of samples (n), mean, standard deviation (SD), coefficient of variation (CV) and maximum values – for site properties in a farmer managed field in Baycola Research Center.

Variable 1 Units n Mean SD CV Maximum

Oc g m2−

50 3316 0.346 13.6 4609

P g m2−

50 4.03 11.193 62.4 11.12

K g m2−

50 122.34 156.647 28.8 228.7

pH log[H+] 50 7.9 0.106 1.3 8.12

Silt % 50 28.02 1.597 5.6 32

Sand % 50 37.32 1.236 3.3 41

Clay % 50 34.16 1.462 4.2 37

Ec mmhos 50 0.642 0.189 29.4 1.42

1 Oc, soil organic carbon content; P, phosphate; K, extractable potassium; Ec, electrical conductivity

The statistical analysis consisted of two stages. First, site property and weed species abundance data were described using traditional descriptive statistics and Pearson correlations among site properties or weed species abundance. Finally, the canonical correlation analysis procedure, PROC CANCORR (SAS 2000) was used to analyze the extent of association between a number of site properties (p) and a number of weed species (q). The starting point for the canonical correlation analysis was the correlation matrix R:

Equation (1)

where R11 is a p by p matrix of correlations among the site properties, R 22 is a q by q matrix of

correlations among the weed species, R12 is a p by q matrix of correlations between the site

properties and weed species abundance, and R 21 is a q by p matrix of correlations between weed species abundance and site properties.

Each canonical correlation is made up of two canonical variates: one representing a linear combination of site properties (S= SITE) and the other representing a linear combination of weed species abundance (W=WEED). Using the partitioned correlation matrix (Equation 1), linear combinations for each set of variables were derived.

Si = a 1 x 1 + a 2 x 2 + ...... + apxp i = 1 , 2 , … , p

Wi = b 1 y 1 + b 2 y 2 + ….. + bqyq i = 1 , 2 , … , q

R 11 R 12

R 21 R 22 R=


Si and Wi are the i th canonical variates, a and b are the standardized canonical coefficients, and x and y are the observed site property and weed species abundance values (standardized to zero mean and unit variance) at each sampled point, respectively. Successive pairs of canonical variates are uncorrelated and orthogonal to Preceding variates.

Typically, most of the covariation between the two sets will be concentrated in the first, two or three pairs of canonical variates, reducing the complexity of the multivariate data set for easier interpretation.

Results and discussion This field was characterized as having low relief and closed depressions areas. Range in pH

was 7. 64 to 8.12 and the greatest coefficient of variation (C.V) values were for (P, K) and (Clay%, pH) concentrations (0.73 and 0.57 respectively) (Table 1). In general, pH was negatively correlated with each other property (except percentage of clay).

Six weed species were observed in this survey (Table 2). The species observed after completion of all weed control practices were Convolvulus arvensis, Abutilon theophrasti, Sorghum halepense, Cucumis melo, Amaranthus retroflexus and Chrozophora tinctoria. The frequency distributions of weed counts were highly skewed with a large number of surveyed quadrates having no weeds while a few quadrates had very high densities of weeds. SORHA was the most abundant in soybean field. CUMMO abundance was correlated with AMASS. Canonical correlation coefficients for each pairs of canonical variables are shown in Table 3.

Ranges of six canonical correlations were between 0.12 and 0.49. These six canonical correlations cumulatively described 78% of the variance in site properties and 81% of the variance in weed abundance. Table 2. Descriptive statistics, mean, standard deviation (SD), coefficient of variation (CV), and maximum values for weed species abundance sampled from a l-by-1 m quadrate placed in a farmer-managed field in Baycola Research Center (n=180)

Weed speciesa

Mean

(pl.m 2− ) SD CV

Max

(pl.m 2− ) CONAR ABUTH SORHA CUMMO AMASS CRZTI

6.39 2.58 8.44 2.87 2.45 4.21

11.57 4.08 20.53 6.97 6.52 3.79

1.80 1.58 2.43 2.42 2.65 0.90

95 26 153 48 53 22

a weed species identified using Bayer codes: CONAR, Convolvulus arvensis. ABUTH, Abutilon theophrasti. SORHA, Sorghum halepense. CUMMO, Cucumis melo. AMASS, Amaranthus bulbosa. CRZTI, Chrozophora tinctoria. Table 3. Canonical correlation coefficients for each pair of canonical variates describing site properties (SITE = S) and weed species abundance (WEED=W) and the cumulative variance explained for each canonical variate.

Canonical correlation Cumulative variance S1/W1 S2/W2 S3/W3 S4/W4 S5/W5 S6/W6 WEED SITE 0.49 ns 0.39 ns 0.34 ns 0.26 ns 0.17 ns 0.12 ns 0.81 0.78

ns: not significant by Bartlet test. First pair of canonical variates.

The first site property canonical variate (S1) (standardized to zero mean and unit variance) had positive coefficients for the Electrical conductivity (Ec), pH, Organic carbon (Oc), Phosphorus (P), percent sand, Silt and Clay and had negative coefficients for potassium (K) (Table 4).


The first weed species abundance canonical variate (W1) was a linear combination of positive coefficients for standardized abundance of CRZTI, ABUTH, AMASS and with a negative coefficient for SORHA, CUMMO and CONAR.

Even though canonical correlations were calculated in Table 3 are very large numbers, but with the Bartlet test had no meanings. Therefore it looks like there is not evidence to show there are relationships between site properties and weed abundance.

Table 4. Standardized canonical coefficients for each pair of SITE (S) canonical variates and corresponding WEED (W) canonical variate. Variable S1 S2 S3 S4 S5 S6 Ec pH Oc% P K Sand% Silt% Clay

0.365 0.237 0.736 0.419 - 0.473 0.639 0.068 0.512

0.304 0.561 0.425 0.843 -1.066 1.724 2.070 1.058

-0.323 -1.267 - 0.056 - 0.294 0.208 0.359 0.208 0.552

-0.179 0.330 -0.394 -0.058 0.376 1.338 1.144 1.297

1.020 0.682 0.284 -0.807 0.091 1.713 2.654 2.153

0.328 0.769 -0.358 -0.359 1.590 -1.362 -1.605 -2.171

SORHA CRZTI ABUTH AMASS CUMMO CONAR

W1 -0.039 0.203 0.464 0.884 -0.767 0.248

W2 -0.439 0.516 -0.119 -0.136 0.538 0.414

W3 0.821 0.430 0.184 -0.270 0.100 0.339

W4 0.355 -0.416 -0.704 0.871 0.155 0.110

W5 0.044 -0.778 0.322 0.380 -0.008 0.822

W6 0.201 -0.444 0.464 -0.093 0.825 -0.235

If we put a side matter of correlation which had no meaning in our experiment, the explanation

about the first pair of canonical variables (W1 , S1 ) are very interesting. With survey about

equation, seems S1 is a equation which compares X 5 ( amount of potassium ) with other site

variables . S 1 = 0.365 Ec + 0. 237 pH + . . . – 0.473 K + . . . + 0.512 clay

Then, S1 is shown as a lack of K. On the other hand, W1 has a large positive coefficient for Y 4

(AMASS) and Y 3 (ABUTH). It seems abundances of AMASS and ABUTH are in the plots, they

have less amounts of K (Fig. 2). Correlation between S1 and eight site variables are calculated in Table 5. According to the table

are shown that S1 has negative correlation and more with X7 (Silt%) and positive correlation with the rest of variables (especially with Oc%). So the best way for expression of S1 is that, this canonical variable shows the large amounts of Oc%.

Correlation between W1, and six other variables (different weeds) are shown in the last table. It seems there is a relationship between S1 with AMASS and ABUTH.

In general to express S1 and W1 with respect to correlation, the pieces contain organic carbon, have more weeds such as AMASS and ABUTH (Fig. 2).

These results show that site properties are related to the weed abundance. Influence of site properties on weed abundance could result from effects of these properties on variation in available soil moisture and soil fertility. For example, Andreasen et al. (1991b) stated that organic matter would be linked not directly to weed occurrence, but through its link to soil moisture availability. This pair of canonical variates was interpreted as a "weed presence-soil type" association largely because of the high values of organic carbon with presence AMASS and ABUTH.


In a same way about the pairs of second variables canonical (S2 , W2) on a basis of coefficients and correlation tables (Tables 4,5) also can get the results, the parts of farm where the Silt% and P is very high , have more abundances of weeds such as CUMMO and CRZTI.

In general, a dense weed population that remained after application of weed control was observed in central part of farm (Fig. 3). It was likely, a poor crop stand, and a lot concentration of site properties that contributed to high weed species abundance observed in this point.

Several mechanisms could be responsible for the associations identified using CCA at this field site. The degree to which any of these mechanisms is responsible needs to be tested experimentally in the field. Another test of the validity of the associations is through the development of models to predict the potential for weed species occurrence based on site property information.


Table 5. Correlation between site properties and weed abundance with corresponding canonic variates

Canonical variables Variable S1 S2 S3 S4 S5 S6 Ec pH Oc% P K Sand% Silt% Clay%

0.326 0.015 0.721 0.412 0.384 0.279

- 0.620 0.426

0.097 - 0.235 - 0.007 0.429 0.060 0.256 0.499

- 0.631

0.350 - 0.871 0.151 0.198 0.444 0.417 0.134

- 0.352

- 0.503 0.372

- 0.337 0.009 0.229 0.545

- 0.426 0.250

0.356 - 0.002 0.103

- 0.283 0.042

- 0.219 0.187 0.118

0.274 - 0.139 0.196 0.376 0.626 0.131 0.186

- 0.331 W1 W2 W3 W4 W5 W6

SORHA CRZTI ABUTH AMASS CUMMO CONAR

- 0.051 0.408 0.543 0.679

- 0.035 - 0.137

- 0.486 0.672 0.087 0.426 0.604 0.437

0.808 0.431 0.164

- 0.035 0.028 0.317

0.286 - 0.002 - 0.552 0.538 0.286

- 0.047

- 0.109 - 0.425 0.280 0.006

- 0.159 0.704

0.116 - 0.115 0.535 0.256 0.724

- 0.436

References Andreasen. C., Streibig. J.C. and Haas. H. 1991b. Soil properties affecting the distribution of 37 weed species in Danish

fild. Weed Res. 31: 181-188. Breaks, C.J.F. 1987. The analysis of vegetation – environment relationship by canonical correspondence analysis.

Vegetation. 69: 69-77. Buchanan.G.A., Hoveland. C.S. and Harris. M.C. 1975. Response of weeds to soil pH. Weed Sci. 23: 473-477. Cardina, J., Sparrow. D.H.and Mc Coy. E.L. 1995. Analysis of spatial distribution of common lambs quarters

(Chenopodium album) in no till soybean (Glycine max). Weed Sci. 43: 258-269. Dale.M.R.T., Thomas. A.G. and John E.A. 1992. Environmental factor influencing management practices as correlates

of weed community composition in spring seeded crops. Can.J.Bot. 70: 1931-1939. Diaz. B., Ribeiro. A., Bueno R., Guinea D., Barroso J., Ruiz D. and Quintanilla C.F. 2005. Modelling will-oat density in

terms of soil factors: A machine learning approach. Precision Agriculture. 6: 213-228. Dieleman.J., Mortensen D., Buhler D., Cambardella C. and Moorman T. 2000. Identifying associations among site

properties and weed species abundance I. Multivariate analysis. Weed Sci. 48: 567-575. Dille, J.A., Mortensen D.A. and Yong L.J. 2002. Predicting weed species occurrence based on site properties and

previous years weed presence. Precision Agriculture. 3: 193-207. Earl, R., Wheeler P., Blackmore B. and Godwin R. 1996. Precision farming: The management of variability.

Landwards. 51: 18-23. Hausler. A.,and Nardmeuer H. 1995. Impact of soil properties on weed distribution. Pages 186-189 in S.E.Oleson,ed.

Proceedings, seminar on site specific farming. Taele: Danish Institute of plant and soil science, sp-teport 26. Johnson, G.A., Mortensen D.A., Yong J. and Martin A.R. 1995. The stability of weed seedling population models and

parameters in eastern Nebraska corn (Zea mays) and soybean (Glycine max) fields. Weed Sci. 43: 604-611. Johnson, G.A., Mortensen D.A., and Gotway C.A. 1996. Spatial and temporal analysis of weed seedling populations

using geostatistics. Weed Sci. 44: 704-710. Medlin. C.R., Shaw D.R., Cox M.S., Gerard P.D., Abshire M.J. and Wardlaw III M.C. 2001. Using soil parameters to

predict weed infestations in soybean. Weed Sci. 49: 367-374. NJonak, J.M., Mortensen T.B. and Cambardella C.A. 1997. Atrazine sorption. J.Environ. Qual. 26: 1271-1277.


Post, B.J.1988. Multivariate analysis in weed science. Weed Res. 28: 425-430. [SAS] Statistical analysis systems. 2000. SAS/STAT user's Guid. Cary, NC: Statistical Analysis Systems Institute. Walter, A.M., Christensen S. and Simmel Sgaard S.E. 2002. Spatial correlation between weed species densities and soil

properties. Weed Res. 42: 26-38.


Control of weeds in flooded rice by non-chemical means

Taberner A., Cónsola S., Llenes JM., Roque A. Servei Sanitat Vegetal Malherbologia Rovira Roure 191 25198 Lleida. [email protected]

Promoting weed control in rice growing in the Ebro Delta by non-chemical means is interesting for the implementation of Integrated Weed Management Programmes in order to promote as much respect as possible for the natural environment in an area of great natural and ecological value. Farmers must therefore follow a number of indications in order to prevent polluting the soil and water in this area and not increasing soil salinity, which could have a negative effect on local fauna and flora. Controlling weeds in rice plantations without the use of herbicides is a complicated and tedious task, but given our interest in promoting such control techniques, objectives of the present work were: 1) To evaluate the effectiveness of strategies used to control weeds in rice plantations by non-chemical methods, such as flooding and mudding, combined with rice sowing or transplanting 2) To evaluation the economic viability of these strategies.

The field trial was carried out in a 0.9 ha commercial field located in the municipality of Deltebre. In 2007, the variety used was Tebre and in 2008 it was Gleva both types of Japanese, round grain and small size. Only one repetition was performed for each treatment. In 2007, the following variants were tested: a) Flooding (F) (5 cm sheet of water) + Mudding (M) at F +30 days + Manual transplanting at M+1. b) Flooding + Mudding at F + 15 + Sowing. c) Flooding + Sowing. d) Overflooding (O) (30 cm sheet of water) + Delayed sowing at O + 27 e) Overflooding and Manual transplanting at O + 30. In 2008, the alternatives tested were: a) Flooding (F) (10 cm sheet of water) + Mudding (M) at F + 50 days + Manual transplanting at M + 4. b) Overflooding (O) + Mudding at O + 25 + Sowing at M +5. c) Overflooding + Sowing at O +30. d) Overflooding (O) + Manual transplanting at O+55. In 2008, the herbicide bentazone was applied in all the overflooded thesis in early July, due to the infestation of the rice plantation by high density Alisma plantago-aquatica seedlings.

Overflooding was very effective against Echinochloa sp., with levels of efficacy above 95%, while in the test conducted with a normal column of water, efficacy was somewhat lower. These results were similar for both years. In the thesis with mudding, the efficacy was lower than for the control with Echinochloa sp. Overflooding did not work as well as the control with Scirpus maritimus, because in an overflooded test, the plant density of this species was even higher than in the test with normal water level, even though the plants were weaker. In contrast, in the first year, Scirpus maritimus was present in all the treatments, although in smaller numbers in the overflooded plots. In the case of Alisma plantago-aquatica, the degree of infestation was independent of the control method used in each test, with very low efficiencies being obtained in all cases.

In these trials, we observed that overflooding was a good practice for controlling Echinochloa sp. There was also significant weakening of Scirpus sp., whereas other species, such as Alisma plantago-aquatica, were able to survive in high water columns and indeed proliferate. In this test, it was therefore necessary to use a herbicide to obtain acceptable efficiencies in controlling broadleaf weeds. We found that it was necessary to adopt cultural measures in order to reduce weed populations.

With regard to the economic evaluation, costs were greater without the use of herbicides, both for transplanting and manual weeding. Moreover, the shortage of manpower to carry out these tasks, which are certainly strenuous, was another argument against their use.

References Hidejiro Shibayama (2001) Weeds and weed management in rice production in Japan. Weed Biology and

Management. 1(1): 53-60. Rutto K.L, Son CY (2002) Are herbicides essential for paddy weed-control in East Asia? Pakistan Journal of Biological

Sciences 5(12): 1352-1362


False seedbeds in organic grown winter wheat

A.Verschwele Julius Kühn Institute, Messweg 11/12, 38104 Braunschweig, Germany, Email:

[email protected]

The false or stale seedbed technique is known as an effective tool of integrated weed control especially in winter crops. However, only little data are available on the additional false seedbed effect compared to the only effect of late sowing.

The effect of the false seedbed technique on weed abundance and yield has been tested in organic grown winter wheat at 3 trials (2005-2007). The investigations were focussed on the evaluation of the different effects of the false seedbed and the effects of the different sowing dates. In addition to the sowing system the seed density was also varied (300 and 450 grains m-2).

Early sowing resulted in high weed density and also low grain yield in all years. In 2of 3 years weed biomass was significantly higher (28-33 g m-2 DM) in the early sown wheat. No significant differences were found between the false seed bed technique and the late sowing. Thus, the more intensive soil cultivation had no effect on weed density, weed biomass and crop yield. Increasing seed density resulted significantly in lower weed biomass, but weed density was the same at all 3 sowing systems.

0

10

20

30

40

50

60

70

80

2005 2006 2007

Wee

d de

nsit

y (n

m-2

)

False seedbed

Early sowing

Late sowing

a

b

a

a a

a a

b

b

Fig 1: Weed density in spring (BBCH 21-25 of winter wheat)

However, in a long-term the false seedbed might reduce the weed seed bank, but due to the

more intensive soil tillage there is also a high risk of sealing. Because of a weak correlation between weed density and wheat grain yield, yield was obviously more effected by other factors (e.g. nitrogen supply). Consequently, at least for silty soils we cannot recommend the false seedbed technique.

References Barberi P (2002): Weed management in organic agriculture: are we addressing the right issues? Weed Research 42 (3),

177-193 Melander B; Rasmussen, IA; Barberi P (2005): Integrating physical and cultural methods of weed control - Examples

from European research. Weed Science 53 (3), 369-381. Rasmussen IA (2004): The effect of sowing date, stale seedbed, row width and mechanical weed control on weeds and

yields of organic winter wheat. Weed Research 44 (1), 12-20.


Round table reports


Report of the round table “Relevant and non-relevant parameters when studying cover crop and mulching

effects.”

Eric Gallandt Round Table Chair (University of Maine, USA)

There were approximately 20 participants in the Round Table discussion related to cover crop and mulch effects. The session chair offered some introductory comments, including examples of non-relevant parameters (e.g., the weed community and/or weed biomass within a cover crop that is terminated prior weed seed production). Of particular note was the point that disturbance associated with the establishment, management, or termination of a cover crop may offer a greater stress to weeds than the cover crop itself. This was related to the need for explicit linkages between the suite of practices that a cover crop interval may include and the phenology of the targeted weed species. The need to focus on the mechanisms responsible for perturbations in weed performance or population dynamics was viewed as essential to advance beyond what is possible with many empirical studies. It was noted that such a diverse group would have benefitted from a review of terminology early in the session as cover crop, green manure, intercrop, living mulch, smother crop have different meanings in different countries. Lastly, there was considerable discussion of regionally-specific cover cropping and/or mulching practices, underscoring the tremendous variation across climates and cropping systems represented by the group. Given more time, it would have been a useful exercise to bring these many examples back to some fundamental principles relating the timing of disturbance, phenology of the weed species, and the strengths or weaknesses the the cover crop or mulching practice.


Report of the round table Unifying parameters in mechanical weed control research

Jesper Rasmussen University of Copenhagen, Faculty of Life Sciences, Department of Agriculture and Ecology,

Copenhagen, Denmark

Abstract This report summarises (i) the introduction given at the initiation of the roundtable discussion

about unifying parameters in mechanical weed control and (ii) the following discussion at the EWRS Physical and Cultural Weed Control Group meeting in Zaragoza 2009.

Previous roundtable discussions Research methodology has always played an important role in the EWRS Physical and Cultural

Weed Control Group, and in 2004 the group published a comprehensive guideline paper on research methodology in physical weed control (Vanhala et al., 2004). The paper focuses on flame weeding, weed harrowing and intra-row cultivation and it deals with the adjustment and use of mechanical weeders, recording the factors that may have an impact on weeding performance, methods to assess effectiveness, experimental designs and statistical analysis. The aim was to share experiences, to increase the comparability between experiments from different environments and to improve research quality.

The guideline paper does not present a single research agenda but emphasises that high-quality research should be promoted through proper methodology. The underlying assumption was that high-quality research is first and foremost characterized by the appropriateness of the applied methods and not so much by the research content itself. Research content, however, is of primary importance to scientific progress. In other words, an appropriate methodology does not ipso facto create scientific progress if the subject matter is trivial.

It is far from simple to decide if research content is crucial or trivial; as with beauty it is all in the eye of the beholder. People who attend the Physical and Cultural Weed Control Group meetings represent diverse backgrounds and interests and do not necessarily share a common perception of good research (Rasmussen, 2004).

Emphasising research objectives The three roundtable discussions at the Zaragoza-meeting 2009 were each devoted to a specific

research objective; the estimation of key parameters in physical and cultural weed control research. The aim of this roundtable discussion about mechanical weed control was to evaluate, whether the working group is mature enough to prioritize future research objectives and whether it can agree on a number of parameters that may facilitate collaboration and progress within future research.

From testing null hypotheses to estimation of meaningful parameters The introduction to the roundtable discussion was based on the assumption that mechanical

weed control can be improved through research and that the evolutionary stage of testing null hypotheses more or less has been passed. It is difficult to envisage that there is much to learn from simple comparisons of qualitative treatments which are subjected to analysis of variance (testing null hypotheses). We are now at the stage where emphasis should be given to quantification of important relationships, and in this context, unifying parameters in mathematical models may play an important role. Such parameters can summarize huge amounts of data and quantify the importance of key factors for the success of mechanical weed control. Key parameters should either fit into decision support models or facilitate the accumulation of basic knowledge.


Definition of a parameter A parameter is a constant in the equation of a curve that can be varied to yield a family of similar curves. If asked to imagine the graph of the relationship y = ax2, one typically visualizes a range of values of x, but only one value of a. Parameter a can therefore be considered to be a parameter: less variable than the variable x, but less constant than the constant to the power 2. Modified after http://en.wikipedia.org/wiki/Parameter

Parameters hold different qualities depending on the type of mathematical equation (curve) in which they take part, and priority should be given to so-called meaningful parameters, which express features that may be considered biologically important. Polynomial regression should be avoided because parameters hold little meaningful information. In contrast, many exponential functions include interpretive parameters that are easy to understand. For example, the resistance parameter, b, presented in Table 1, expresses the relative decline in leaf cover or crop density for each pass with a cultivator. The parameter may easily be converted to percentage decline per pass (Rasmussen et al., 2009).

It is an intellectual challenge is to make what is “meaningful” explicit when new parameters are suggested, but discussions about the meaningfulness of new parameters are much more rewarding than discussions of ANOVA tables and treatment means.

Examples To give an idea about how meaningful parameters may look, three parameters from own

research in post-emergence weed harrowing were presented: the parameters of 1) crop resistance, 2) weed control and 3) crop tolerance, which all are parameters in the family of exponential functions. From the crop resistance (b) and the weed control (d) parameters (Table 1), the selectivity curve can be deduced and calculated as the relationship between weed control (WC) and crop soil cover (CSC) (Rasmussen et al., 2008):

1100 1 1 ln 1100

dCSCWC

b

− = ⋅ − − −

Rasmussen et al. (2009) showed how crop recovery can be derived from the crop resistance

parameter and the so-called crop tolerance parameter, which expresses how crop yield respond to increasing cultivation intensity in weed-free environments (not shown here).

Examples were given to show that factors like row spacing and crop species influence the parameter values (Rasmussen et al., 2008, 2009). The quantification of the importance of these and other factors should be subjected to future research. Finally, it was briefly outlined how the three parameters could be integrated into models with predictive power in order to determine the optimum aggressiveness of cultivation in respect to crop yield.


Table 1. Description of two key parameters in mechanical weed control research Name Crop resistance parameter Weed control parameter Definition The ability of the crop to resist

cultivation. Assessment shortly after cultivation before recovery takes place.

The decline in weed density immediately after cultivation.

Mathematical notification

Parameter b in ( )exp0L L b I= ⋅ − ⋅

The resistance parameter (b) expresses the relative decline rate of L relative to I. L is leaf cover or crop density; L0 is leaf cover or crop density in untreated plots. I is the cultivation intensity, which could be number of passes.

Parameter d in ( )( )exp ln 10W W d I= ⋅ − ⋅ +

The weed control parameter (d) expresses the relative decline rate of weed density (W) relative to I. W0 is weed density in untreated plots; I is cultivation intensity, which could be number of passes

Estimation After transformation of the response linear regression is possible: ln( ) ln( )0L L b I= − ⋅

After transformation of the weed density and the intensity linear regression is possible ln( ) ln( ) ln( )0W W d I 1= − ⋅ +

Other If L is leaf cover the percentage of crop soil cover (CSC) is calculated as: ( )( expCSC 100 1 b I= ⋅ − − ⋅

The percentage of weed control (WC) is calculated as

( )( exp )ln(WC 100 1 d I 1= ⋅ − − ⋅ +

Protocol for experiments and statistics

Rasmussen et al. (2008) Rasmussen et al. (2008)

Summarising the roundtable discussions Among the participants there was consensus that the roundtable should be focused on

cultivation with low selectivity, which means post-emergence broadcast cultivation and intra-row cultivation.

Three groups were formed and each group was asked to choose a specific cultivation technique and come up with important parameters, and factors that may influence the parameters. Each parameter should be given a descriptive name. Finally the importance of the proposed parameters should be prioritized and if possible suggestions of experimental designs should be given.

It was decided to use the majority of the allocated time for the roundtable in smaller groups. In plenum, however, it became evident that few were familiar with the concept of “meaningful” parameters, model development and advanced regression analysis. Diverse backgrounds and lack of experience with advanced regression analysis made it difficult to find a common platform for the discussions. In general, the questions raised in the introduction were turned into new questions and there seemed to be more focus on the limitations of the proposed parameter approach than the prospects of the approach. It was obvious that the time was too short for an in-depth discussion of possible benefits of the unifying parameter approach. Reducing the complexity of mechanical weed control into a number of key parameters in mathematical models appeared overwhelming for many participants.

Therefore, all groups more or less created their own agenda for the discussion, and again the roundtable moved into a discussion about different questions related to methodology. The guideline paper from 2004 (Vanhala et al., 2004) deals with a number of the raised questions but not all.


The following questions were raised and discussed in the groups:

• Should we use densities, leaf cover or biomass when crop and weed impacts are assessed? • Mechanical weed control is so complicated that it is often considered an art, how does

science cope with this complexity? • How can we bridge the gab between science and practise? • How much should we go into plant/crop physiology and other basic disciplines to

understand crop and weed responses to cultivation? • Parameters are important – but do we focus on the right ones? • Do we always focus on the right responses? • Do scientists always know what happens in the field when they work with field experiments

and what are the implications of too little knowledge about the field work? • How do different environments (soil type) influence resistance against weeding? • How important is uprooting versus soil covering of weeds? • Cultivation before crop emergence – how should it be modelled? • Why don’t we have – or use – standards for crop and weed assessments?

Conclusion The roundtable was a success if engagement and pointing out of new questions related to

research methodology were the criteria. It failed, if the success criterion was an in-depth discussion of key parameters as outlined in the introduction to the roundtable. In retrospect, it may have been too ambitious to expect the roundtable to agree on a number of important parameters and prioritize future research objectives. This would require much more time than was allocated. Even if it is difficult to conclude from the roundtable, there seems to be an agreement that models with meaningful parameters should be given higher priority in future. It is, however, important that protocols for experimental design and statistical procedures are available if parameter estimation should out-compete experiments that are designed to answer whether different treatments give different results.

References Rasmussen J (2004) Are we making progress in mechanical weed control research? Pages 115-122 in Prodeedings 6th

EWRS Workshop on Physical and Cultural Weed Control. Lillehammer, Norway: European Weed Research Society.

Rasmussen J, Bibby B; Schou AP (2008) Investigating the selectivity of weed harrowing with new methods. Weed Research 48, 523-532.

Rasmussen J, Nielsen HH & Gundersen H 2009 Tolerance and selectivity of cereal species and cultivars to postemergence weed harrowing. Weed Science 57 (in print).

Vanhala P, Kurstjens DAG, Ascard J, Bertram B, Cloutier DC, Mead A, Raffaelli M, and Rasmussen J (2004) Guidelines for physical weed control research: flame weeding, weed harrowing and intra-row cultivation. Pages 208-239 in Prodeedings 6 th EWRS Workshop on Physical and Cultural Weed Control. Lillehammer, Norway: European Weed Research Society.


Report of the round table discussion “Research methodology in thermal weed control”

Johan Ascard1 and Daniel Cloutier2 1Swedish Board of Agriculture, Box 12, SE-230 53 Alnarp, Sweden.

E-mail: [email protected] 2Institut de malherbologie, Beaconsfield, Canada

Prior to the meeting, each participant was asked to read the Guidelines for physical weed control research: Vanhala et al. (2004).

There were 16 participants to the round table on thermal weed control. There were almost as many different research interests as there were participants. The research topics ranged broadly and the discussion eventually settled on a few topics.

In the introduction, it was established that methodology and techniques will differ depending on whether the research is being conducted on non-crop land, in orchards, in vegetable crops or in arable crops. Research methods are also likely to differ if the thermal weed control method(s) is (are) used as an integral part in a weed managament system comparison, or if the objective is to develop tools and/or technologies. There is also a huge difference in technology and methods for above ground weed control versus below ground soil disinfection. Therefore there cannot be a single general methodology and technique in thermal weed control research. Instead the methodology has to be selected according to the kind of research that is carried out. Some general aspects were discussed further.

Liquefied petroleum gas (LPG): During the course of the discussion, the exact content of the various components commonly

referred to as liquefied petroleum gas (LPG) was debated. It was mentioned that it is a mixture of propane and butane but that the proportions of theses gases vary between suppliers and even between seasons in temperate zones. Therefore it was suggested that the exact composition of the LPG used should be clearly determined when experiment(s) are reported to insure proper interpretation of different experiments.

Rate (dose): The term dose-response was mentioned during the discussion which led to discussing the

proper terminology that should be used (rate or dose) when presenting data where LPG was used. There was some debate which ended up with no consensus.

Looking at The Eleventh Edition of the Concise Oxford English Dictionary, it has the following definitions:

dose (noun) - a quantity of a medicine or drug taken at one time. - an amount of ionizing radiation received or absorbed at one time. rate (noun) - a measure, quantity, or frequency - the speed with which something moves or happens.

It seems that dose means a quantity in some fields, including toxicology. Consequently, pesticide quantities might be included which might explain the general use of dose-response in agriculture.


How to express LPG doses Some time was spent discussing how to express doses of LPG gas when reporting experiments.

Currently there is some confusion when trying to convert results expressed as quantity/area (kg/ha) for broadcast treatment with experiments where flaming is done as a banded treatment or only in the crop rows. In in-row treatments, using so called cross flaming, the quantity might be expressed on a distance basis (g/m) instead of amount of fuel per area Consequently, the rate is difficult to convert to an area basis without knowing the flame width or, in the case of cross flaming, having to make some hypotheses such as (area treated intra-row) prior to converting this arbitrary value to a per hectare basis. The concept becomes more confusing since the type of equipments used are also totally different (broadcast vs intra-row flamers) and not relevant to the reciprocal situation(s).

The consensus for this particular section was that it is absolutely important to report and measure as many parameters as possible, including conversion factors, in order for others to be able to relate their own work to what was done within a given experiment. This implies measuring the actual quantity of gas used (not just look at the pressure etc and calculate the quantities used). It involves weighing the LPG tanks before and after using the flamers at the various experimental doses, preferably calibrating prior to conducting the experiment and after the experiment is done to ensure that the actual quantities used are reported.

Is kg a good measure of LPG use? After some of the debates above, a participant suggested that the true quantity to express

flaming dose should be on an energy basis (Joules) instead of mass (kg). This would have the advantage of taking into account the gas composition (since propane/butane ratios have different energy) and would probably be a more accurate variable to express flaming results. However, the conversion difficulties between area (broadcast) vs in-row flamers (on the row) would remain. Effective and lethal doses

We discussed the different ways to express the effective dose or lethal dose, i.e. the amount of fuel or energy needed to achieve a certain level of weed control. For example, in some recent literature you may find estimations of the fuel required to reach 50, 80, 90 or 95 % weed control on weed biomass in terms of ED50, ED80, ED90 and ED95. In some papers, estimates for percent weed control on plant numbers are stated, e.g. LD90. These differences make it difficult to compare fuel rates between experiments. However, we concluded that the relevant and desired weed control levels are different in e.g. arable crops vs vegetable crops, and, therefore, we cannot recommend one standard. However, it may be useful to standardize the way of expressing dose rates in similar types of experiments.

Future activities At the end of the round table, all the participants were unanimous in expressing the need to

update and review the Vanhala et al (2004) Guidelines as far as flaming is concerned and some volunteered to contribute to a revision and add material to new/updated guidelines. Reference Vanhala P, Kurstjens DAG, Ascard J, Bertram B, Cloutier DC, Mead A, Raffaelli M, and Rasmussen J (2004)

Guidelines for physical weed control research: flame weeding, weed harrowing and intra-row cultivation. Pages 208-239 In: Cloutier DC and Ascard J (eds). Proceedings of the 6th EWRS Workshop on Physical and Cultural Weed Control, Lillehammer, Norway, 8-10 March 2004. p. 208-239.

EUROPEAN WEED RESEARCH SOCIETY

Proceedings8th EWRS Workshop on

Physical and Cultural Weed ControlZaragoza, Spain

9-11 March 2009

european weed research society - ewrs · zaragoza, spain 9-11 march 2009. european weed research...

Documents