recognizing and overcoming difficult site conditions for afforestation of bottomland hardwoods

ECOLOGICAL RESTORATION 223 SEPTEMBER 2004 183

In the last decade about 370000 acres(150000 ha) of economically marginal

farmland in the Lower Mississippi Allu-vial Valley (LMAV) have been restored tobottomland hardwood forests (Stanturfand others 1998 King and Keeland 1999Schoenholtz and others 2001) Plantingof this considerable acreage is due to sev-eral federal programs such as the Wet-lands Reserve Program (WRP) that assistlandowners by financing afforestation(Figure 1)

Unfortunately these operationalplantings have not performed as well assmaller plantings or research plots(Stanturf and others 2001a) For examplea recent survey of WRP plantings in west-central Mississippi revealed that morethan 90 percent of the sites failed to meetthe criteria of 100 woody stems per acre(247 stems per ha) three years after plant-ing or direct seeding While planting 1-0bareroot seedlings of oak was more suc-cessful than direct-seeding acorns only 23percent of the land planted with seedlingsmet the criteria (CJ Schweitzer unpub-lished data) Planting and direct seedingoak (Quercus spp) on public land in thesame area has been more successfulMeanwhile Allen (1990) found 70 per-cent of the planted bottomland hardwoodstands on the national wildlife refuges heevaluated had more than 200 trees peracre (494 stems per ha)

We believe that the recurring prob-lems in operational plantings on privatelands are due in part to the failure ofplanters to recognize adverse site condi-tions and their failure to use appropriatemethods for overcoming site limitationsOur objectives in this paper are to syn-thesize research and experience intoguidelines for recognizing adverse siteconditions due to hydroperiod soil com-peting vegetation and herbivory Wedescribe techniques for overcoming theseconditions and suggest promising re-search areas

Recognizing AdverseHydroperiod ConditionsThe former agricultural sites available forafforestation are often very low and wetEven wetland trees suited for these sitesrequire aerated soil conditions during thegrowing season and their seedlings cannottolerate overtopping by floodwaters oncethey have leafed out

Two conditions of excess water in afloodplain are adverse to tree seedlings 1)prolonged periods of saturated soil includ-ing persistent standing or flowing waterand 2) high water levels that coverseedlings during the growing season Seed-ling tolerance to ldquofloodingrdquo or ldquowater-loggingrdquo generally refers to the ability to

Recognizing andOvercoming Difficult SiteConditions for Afforestationof Bottomland Hardwoodsby JA Stanturf WH Conner ES Gardiner CJ Schweitzer and AW Ezell

PRACTICE AND PERSPECTIVE

Restoring bottomland

hardwoods requires

attention to site con-

ditions matching tree

species to the site

and controlling weeds

and herbivores in order

to achieve success

Ecological Restoration Vol 22 No 3 2004 ISSN 1522-4740 E-ISSN 1543-4079copy2004 by the Board of Regents of the University of Wisconsin System

withstand saturated soil conditions inwhich oxygen is consumed in respirationand soil voids become filled with carbondioxide and accumulated metabolic prod-ucts of anaerobic microbes Some treespecies have developed adaptive traits(McKevlin and others 1998) and can accli-mate to waterlogging but at a cost to theplant Only a few species can withstandwaterlogging for extended periods of timeduring the growing season (Table 1) Evenbaldcypress (Taxodium distichum) and watertupelo (Nyssa aquatica) species that growin deepwater swamps require non-floodedconditions for their seeds to germinate

Immersion of seedlings after theyhave leafed out in the spring by overbankor backwater flooding can cause mortal-ity of even the so-called flood-tolerantspecies During the spring and early sum-mer of 1973 Kennedy and Krinard (1974)observed the effects of a major flood onthe Mississippi River on planted and nat-ural stands One growing season after theflood natural and plantation stands thatwere one year or older at the time of theflood and inundated for two months orless were not severely affected eventhough the seedlings were less than 2 feettall These stands included sweet pecan

(Carya illinoensis) water oak (Quercusnigra) Nuttall oak (Q nuttallii) and cher-rybark oak (Q falcata var pagodaefolia)On the other hand 1-year-old sweetgum(Liquidambar styraciflua) which wasflooded for three months was killed Newplantings of cottonwood (Populus del-toides) cuttings were destroyed but theseedlings seemed to do better

Diagnosing hydroperiod limitations ofpotential afforestation sites may not beeasy however Widespread and numerousflood control structures and the sub-sequent regulation of river flows havechanged the seasonality and extent of floodevents (Dynesius and Nilsson 1994 Ligonand others 1995 Shankman 1999) Posi-tion in the landscape (for example ridgeflat or slough) and soil characteristics areindicative but not diagnostic (Baker andBroadfoot 1979) since land leveling anddrainage due to farming practices canchange site conditions We recommendobtaining at least a five-year history offlooding on the site before choosing thespecies to plant Landowners and farmmanagers can provide good informationabout such things as how often in the lastfive crop years planting was delayed or acrop was lost due to high water

Overcoming Growing-Season FloodingMatching species to site is absolutely crit-ical Species may be planted under less fre-quently flooded conditions than shownfor their tolerance class (Table 2) but notthe reverse For example baldcypress canbe planted and survive on ridges but cher-rybark oak should never be planted insloughs If management objectives allowflexibility in the choice of species simplyplant the species adapted to the worstprobable flooding conditions Alterna-tively one can plant a mixture of speciesadapted to a range of flooding regimes andexpect significant mortality because somespecies will fail to meet the site conditionsHowever when species composition andstocking are critical (for example if finan-cial returns from either timber or carbonsequestration payments is an objective)then other strategies such as controllingflooding delayed planting and planting instanding water are available to overcomegrowing-season flooding

Controlling Flooding Controlling flooding during the treesrsquoestablishment phase may be possible insome situations Restorationists may beable to keep sites managed as greentreereservoirs constructed wetlands or thosedownstream from water control struc-tures from flooding while seedlings ofmoderately tolerant species becomeestablished If flooding is caused bybeaver dams intensive trapping and con-trol may be required for a few years in asmall area while seedlings become tallenough to withstand flooding (C Sloanpers comm)

Delayed PlantingBareroot seedlings should be dormantwhen planted which means December toMarch in the LMAV Some sites are fre-quently under water during this optimumplanting window thereby hamperingrestoration efforts Avoiding flooded con-ditions may be possible although it isoften unreliable and expensive In back-water areas for example waiting to plant

184 ECOLOGICAL RESTORATION 223 SEPTEMBER 2004

Figure 1 Workers carrying bags full of dormant tree saplings mostly oaks fan out to plant a

former agricultural field in the Lower Mississippi River Alluvial Valley to bottomland hardwood

forest Photo by Emile Gardiner

until spring floodwaters recede would bedesirable but planting bareroot stock inJune is risky (Conner and others 1993Allen and others 2001) The US ArmyCorps of Engineers has successfullyplanted container stock later in the sum-mer (J Kiser pers comm) and other

researchers have shown container stock tobe effective but expensive (Williams andCraft 1998 Howell and Harrington 2002Williams and Stroupe 2002) Cost esti-mates vary considerably King and Kee-land (1999) surveyed contractors andagencies in the region and determined

that planting container seedlings costs$100 to $450 per acre compared withaverage contractor costs for barerootseedlings of $32 to $250 per acre Averagecost per seedling in 2003-2004 was $020to $030 for a range of hardwoods com-pared with very large container seedlings(5-6 ft tall) costing $6 each

Planting in Standing WaterPlanting tree seedlings into standingwater stresses a seedling more than plant-ing in terrestrial environments In addi-tion to the normal ldquoshockrdquo of outplantingnursery grown seedlings planted in waterwill shed their existing root system anddevelop one better adapted for life instanding water Producing a new root sys-tem places a large energetic drain on theseedling at a time when it is especiallyvulnerable to other stresses

Heavily root pruning tree seedlings isone method of planting in standing waterthat has been tested extensively in thesouthern United States This practicesimply involves inserting the seedling intothe soil or sediment without digging ahole (Conner 1988 1993 Conner andFlynn 1989 Reed and McLeod 1994Hesse and others 1996 Brantley andConner 1997) The method has beentested in habitats from standing backwa-ter to flowing streams in coastal andinland areas and from Louisiana andSouth Carolina (Figure 2)

Conner and his colleagues (1999)tested bareroot seedlings of baldcypresswater tupelo and green ash (Fraxinus penn-sylvanica) pruned to three different severi-ties moderately severely and withoutroots In the least severe treatment (moder-ately pruned) they pruned lateral and tap-roots to a 9-inch (23-cm) spread Severelypruned seedlings had all of the lateral rootsremoved and the taproot pruned to 9 inches(23 cm) Moderately and severely prunedseedlings were planted by grasping theseedling at the root collar and inserting it 8inches (20 cm) deep into soft sedimentCuttings without roots were prepared byremoving all of the root system below theroot collar and dipping the cut end into acommercially available rooting hormoneThey were planted like the other stock


Table 1 Waterlogging-tolerance ratings for common tree species endemic to majorand minor river bottoms of the southern United States (adapted from Hook 1984)

Most Tolerantbuttonbush (Cephalanthus occidentalis) water tupelo (Nyssa aquatica)Carolina ash (Fraxinus caroliniana) water-elm (Planera aquatica)pumpkin ash (Fraxinus profunda) black willow (Salix nigra)swamp-privet (Forestiera acuminata) baldcypress (Taxodium distichum)swamp tupelo (Nyssa sylvatica var biflora) pondcypress (Taxodium distichum var nutans)

Highly Tolerantwater hickory (Carya aquatica) overcup oak (Quercus lyrata)waterlocust (Gleditsia aquatica)

Moderately Tolerantboxelder (Acer negundo) sweetgum (Liquidambar styraciflua)red maple (Acer rubrum) sweetbay (Magnolia virginiana)silver maple (Acer saccharinum) water oak (Quercus nigra)river birch (Betula nigra) pin oak (Quercus palustris)hawthorn (Crataegus spp) willow oak (Quercus phellos)common persimmon (Diospyros virginiana) Nuttall oak (Quercus nuttallii)green ash (Fraxinus pennsylvanica) sycamore (Platanus occidentalis)honeylocust (Gleditsia triacanthos) eastern cottonwood (Populus deltoides)deciduous holly (Ilex decidua) cedar elm (Ulmus crassifolia)

Weakly TolerantAmerican hornbeam (Carpinus caroliniana) blackgum (Nyssa sylvatica)pecan (Carya illinoensis) cherrybark oak (Quercus falcata var pagodaefolia)shellbark hickory (Carya laciniosa) laurel oak (Quercus hemisphaerica)sugarberry (Celtis laevigata) swamp chestnut oak (Quercus michauxii)hackberry (Celtis occidentalis) Shumard oak (Quercus shumardii)American holly (Ilex opaca) live oak (Quercus virginiana)black walnut (Juglans nigra) winged elm (Ulmus alata)red mulberry (Morus rubra)

Least Tolerantpawpaw (Asimina triloba) black cherry (Prunus serotina)flowering dogwood (Cornus florida) white oak (Quercus alba)American beech (Fagus grandifolia) sassafras (Sassafras albidum)yellow-poplar (Liriodendron tulipifera) slippery elm (Ulmus rubra)eastern hophornbeam (Ostrya virginiana)

Table 2 Waterlogging tolerance classes in terms of flooding duration and season

Tolerance ClassMost Highly Moderately Weakly Least

Duration 100 percent 50-75 percent 50 percent 10 percent1 2 percent1

Winter Yes Yes Yes Yes YesSpring Yes Yes Yes Yes SeldomSummer Yes 1-3 months Early only Seldom No1Refers to growing-season flooding

The researchers found that survivalof baldcypress and water tupelo seedlingswas excellent in both the severe andmoderately severe pruning treatmentsBoth these species are well suited to wetenvironments (Table 1) and prunedseedlings are quickly and easily planted instanding water Moderately root-prunedseedlings of baldcypress and water tupeloplanted in water 1 foot to 2 feet (30-60cm) deep survived as well as seedlingsplanted in shallower water Total removalof the root system was detrimental to bothcypress and tupelo although there wassome survival (33 percent) of baldcypresscuttings after three years No amount ofroot pruning was appropriate for green ashseedlings after three years of almost con-tinuous flooding

Recognizing Adverse Soil ConditionsGrowth of bottomland hardwoods dependson the physical condition of the soil mois-ture availability during the growing seasonnutrient availability and aeration (Bakerand Broadfoot 1979) Bottomland oaks

the most frequently planted bottomlandhardwood species (King and Keeland 1999Schoenholtz and others 2001) grow beston moist well-drained sites with good fer-tility and medium-textured soils Howeverheavy clay soils typify most areas availablefor afforestation and on these soils oaksurvival is often lower and growth less sub-stantial (Stanturf and others 1998)Seedlings planted in clay soils frequentlyface moisture stress during late-summerperiods of low rainfall

Traffic pans (compacted layers formedunder pressure of repeated passes by equip-ment on the surface) are a fairly commonoccurrence in the LMAV especially insoils with high silt content They usuallyform just below the average depth of agri-cultural cultivation about 6 to 8 inches(15-20 cm) Traffic pans impede tree rootpenetration in soil thereby reducing theseedlingrsquos access to the soilrsquos resources

Soil chemistry is another concernwhen planting bottomland hardwoodsMost oaks grow best in soils with a pHrange from 60 to 70 Unfortunatelyrecent alluvial deposits may have a pHapproaching 80 These soils can be aproblem because some oak species espe-

cially Nuttall cherrybark and water oaks(Kennedy 1993) experience low vigorand increased mortality largely due to alack of iron at this pH level

Each of these soil conditions requirethe restorationist to take correctiveactions that should include 1) matchingthe species to the site conditions 2) sitepreparation and 3) soil amelioration

Overcoming SoilLimitationsMatching Species to SiteHardwood foresters use a publication byBaker and Broadfoot (1979) to matchspecies to site for establishing timber plan-tations This approach involves estimatingproductivity from site characteristicsWhile there is no rule for correlating pro-ductivity measures with afforestationpotential for wildlife or other purposesseveral researchers including the leadauthor of this article have suggested thata site be at least minimally acceptable fora tree species (Stanturf and others 1998Groninger and others 2000) According toBaker and Broadfoot (1979) that wouldmean a site is capable of achieving at least54 percent to 63 percent of the maximumproductivity level for that species

As mentioned above several oakspecies do not survive or grow well on highpH soils Shumard oak (Q shumardii)however has been planted successfully onhigh pH soils where other oaks are unsuit-able (Kennedy and Krinard 1985) Inthree separate plantings Shumard oak sur-vived and grew well on soils with pH from78 to 80 Other hardwoods such as greenash and sycamore are more tolerant thanoaks of slightly alkaline conditions (Bakerand Broadfoot 1979)

Site PreparationSite preparation prior to planting formeragricultural land requires disking at leasttwice with a heavy disk in late summer orearly fall Disking should be to a depth ofat least 8 inches (20 cm) preferably to 15inches (38 cm) Deeper plowing or rip-ping is recommended for sites with trafficpans In heavy clay soils ripping should be


Figure 2 Planting techniques for areas with standing water and soft sediments include planting

severely pruned bald cypress (Taxodium distichium) seedlings In this practice the lateral roots

of the bald cypress seedlings are removed and the tap root is cut to 9 inches long The seedlings

are then inserted into the soil or sediment without digging a hole Photo by William H Conner

with a straight shank because winged rip-pers leave subsurface voids that are accen-tuated by shrinking in summer causingroot desiccation and seedling mortalityRipping is commonly prescribed for cot-tonwood on all soils to ease the plantingof cuttings

Site preparation for cottonwoodplantings on former agricultural land ismore intensive than that generally prac-ticed for other hardwoods (Stanturf andPortwood 1999 Stanturf and others2001b) Ideally site preparation beginsimmediately following soybean harvest Ifsoybeans are harvested with a combineplant residues are chopped and shreddedthe fine debris poses no problems Thefirst step in site preparation is double disk-ing (disking in two passes each perpen-dicular to the other) Ripping with astraight shank breaks up the subsoil If atraffic pan has developed subsoiling witha winged ripper will break up the panmore completely than will a straightshank However subsoiling with a wingedripper must be done a year before plantingto allow voids to fill and cracks to closeLiquid nitrogen fertilizer is added in thesame pass to the planting slit made by theripping shank Specialized equipmentplaces the fertilizer 18 to 20 inches (46-51cm) deep in the slit On Sharkey (AericEpiaquerts) and other expanding claysoils it is essential for the slit to undergoseveral wetting and drying cycles (fromprecipitation) in order for fine particles tomove into and fill the slit Otherwise soildrying in the spring and summer willcause the soil to crack along the plantingslit exposing tree roots to desiccation

Soil AmeliorationIn other areas of the South bottomlandsoils may not be as fertile as it is in theLMAV (Francis 1985) and availablephosphorus may limit seedling growth(Stanturf and Schoenholtz 1998) In theLMAV agricultural soils have lowerorganic matter content and may bedepleted of nitrogen (Gardiner and others2001) For this reason high nitrogendemanding species such as cottonwoodreceive nitrogen fertilizer at time of plant-ing Although fertilization may not be

justified economically in terms of in-creased wood production early heightgrowth may reduce risk from flooding andherbivory Broadcast fertilization at timeof planting may stimulate weed competi-tion prompting extensive weed control forseveral years after planting Few guide-lines are available for fertilizing hardwoodplantings other than cottonwood

Recognizing CompetingVegetationEven when species have been properlymatched to site and soil conditions theymust compete with weeds Three condi-tions of competing vegetation can be rec-ognized 1) the ldquonormalrdquo weed complex onthe sitemdasha legacy of past land use and sur-rounding seed sources 2) ldquoproblemrdquoweeds particularly woody vines and 3)ldquoinvasiverdquo non-native species Generallypressure from herbaceous competition willbe severe in old agricultural fields Manyweed species are present there in root-stocks or buried seed that may not be visi-ble immediately after crops are harvested

ldquoProblemrdquo weeds must be recognizedand controlled prior to establishing hard-woods because there are no operationalcontrol options once hardwoods areplanted Kudzu (Pueraria montana) Jap-anese honeysuckle (Lonicera japonica)pepper-vine (Ampelopsis arborea) andtrumpet creeper (Campsis radicans) areserious problems as are bahiagrass anddallisgrass (Paspallum spp) Broom sedge(Andropogon virginicus) is not as serious asthe others listed here but requires controlprior to establishment

Non-native invasive species on bot-tomland hardwood sites include Japaneseclimbing fern (Lygodium japonicum)cogongrass (Imperata cylindrica) Chinesetallow (Sapium sebiferum) Japanese andChinese privet (Ligustrum japonicum Lsinense) Japanese honeysuckle andChinese wisteria (Wisteria sinensis) Mostof these species can be controlled withherbicides (Table 3) but all these non-native species are difficult or impossible tocontrol after planting without harmingthe tree species

Overcoming CompetingVegetationThere are two basic methods for control-ling competing vegetationmdashcultivationand herbicides

CultivationCommon practice in bottomland hard-wood afforestation programs has been toplant without any site preparation imme-diately after the agricultural crop has beenharvested or simply to disk once on fal-lowed sites (Stanturf and others 1998)Cottonwood is a special case where dou-ble disking and herbicides are used(Stanturf and others 2001b) Kennedy(1981a 1981b) compared mowing ordisking to no competition control andfound that mowing was as ineffective asno control Disking on the other handcan significantly improve the survival andgrowth of bottomland hardwood seedlings(Houston and Bucknor 1989 Kennedy1981a 1981b) although access on wetsites can limit use of cultivation as a weedcontrol technique

HerbicidesIn old fields with a ldquonormalrdquo weed com-plex herbicides consistently improve thesurvival of oak by as much as 25 percentand sweetgum from 10 percent to 15 per-cent For cottonwood herbicides canimprove survival by 25 percent comparedto mechanical control and by as much as80 percent compared to no control(Stanturf and others 2001b)

In fields with problem species suchas woody vines it is common to seeseedling mortality of 60 percent or moreeven when herbaceous competition hasbeen controlled If problem species ornon-native invasive plants are presenteffective competition control prior toplanting will likely determine the successor failure of a restoration effort

Chemical Site PreparationMany herbicides labeled for broadcastapplication can be used for bottomlandhardwood site preparation (Anon 1999)


(Table 3) Arsenal ACreg (imazapyr)which is labeled for hardwood manage-ment is extremely lethal to a broad spectrum of woody and herbaceousspecies It is an effective chemical for sitepreparation but because of its soil activ-ity sufficient time must elapse betweenapplication and planting seedlings Otherherbicides containing picloram are la-beled for hardwood management (Allenand others 2001) but are seldom used(Tordonreg 101 and Tordonreg 101RTordonreg K Tordonreg RTU AccessregPathwayreg) In some limited cases whereinjection of undesirable hardwoods isrequired Chopperreg (EC formulation ofimazapyr) may be useful For most pur-poses broadcast application is recom-mended for site preparation rather thanspot or banded application because plant-

ing spots are better identified under actualplanting conditions Another reason tofavor broadcast application is the vigorousregrowth of the main competitors Clear-ing small areas around a newly plantedseedling likely would be ineffective

Weed Control After PlantingControlling groundlayer weeds is possibleafter tree planting but care must be takento use the proper herbicide for the givensituation Oustreg (sulfometuron methyl)controls many broadleaves and somegrasses but does not harm woody species(Ezell and Catchot 1997 Groninger andBabassana 2002) It can be applied afterplanting but before seedlings break dor-mancy Atrazinereg 4L and Princepreg 4L(atrazine and simazine) are other pre-

emergent herbicides that are effective onbroadleaves

Goalreg 2XL (oxyfluorfen) has shownexcellent control of broadleaves in testswith some grass control but no damage tohardwoods (Ezell 1999a) It is currentlylabeled for use on cottonwood andhybrid poplar Scepterreg 70DG (imaza-quin) provides excellent broadleaf con-trol in tests with no damage to cropspecies such as oak sweetgum or cotton-wood It is currently labeled for cotton-wood and the label could be expanded toother hardwoods

Milestonereg (azafenidon) is a pre-emergent herbicide for broadleaf controlIt has shown to be very effective in tankmixes with Oustreg with no damage to thecrop oaks (Ezell 1999b) Milestone is cur-rently labeled for use in citrus orchards


Table 3 Chemical control options for woody vines and non-native invasive species Effective control of these problem speciesrequires application prior to planting hardwood seedlings (Source Miller 1997)

Problem Species Herbicide Comment

kudzu (Pueria montana) Tordonreg K Follow-up applications are often required Escortreg may be the better choice Escortreg because it has less risk of damaging the planted seedling

Japanese honeysuckle Escortreg (metsulfuron methyl) Needs higher rate of Accordreg than for site preparation(Lonicera japonica) Accordreg (glyphosate)

pepper-vine Vanquishreg or Extremely aggressive in abandoned fields will cause significant mortality Must (Ampelopsis arborea) Banvelreg (dicambra) be ldquonon-crop areardquo or pasture not labeled for forestry use

trumpet creeper Accordreg (glyphosate)(Campsis radicans)

redvine (Brunnichea cirrhosa)

cogongrass Arsenalreg (imazapyr) or High rates of Arsenalreg or tank mix ArsenalregAccordreg prior to planting(Imperata cylindrica) ArsenalregAccordreg (glyphosate)

Chinese tallow Arsenalreg (imazapyr) or Apply pre-plant with spot treatments (directed spray) of Accordreg after planting(Sapium sebiferum) Accordreg (glyphosate) prior

to planting

Japanese privet (Ligustrum japonicum)

Chinese privet Arsenalreg (imazapyr high rates) Accordreg (glyphosate) at high rates only will suppress privet(L sinense) or Chopperreg (imazapyr)

for control

Chinese wisteria Garlonreg 4 (triclopyr) Pre-plant application controls from early to mid summer control decreases later(Wisteria sinensis) in the growing season and follow-up treatment by directed spray of Garlonreg 4

may be required

Japanese climbing fern Cannot be controlled by any available herbicide(Lygodium japonicum)

Endurancereg (prodiamine) provides goodherbaceous control in pre-emergent appli-cation It has been tested on cottonwoodand could prove to be a good product forhardwoods

Many herbicides are effective againstgrasses and all are applied post-emergentFusiladereg DX (fluazifop-butyl) controlsmany grasses with the best resultsobtained by making two applicationseach at half the total recommended appli-cation rate Vantagereg and Poastreg are bothsethoxydium with broad-spectrum con-trol of grasses but no effect on broad-leaves These two herbicides work bestwhen crop oil is used as a surfactantalthough this can burn hardwood foliageSelectreg (clethodim) is another effectivegrass herbicide that can be used eitherwith crop oil or a non-ionic surfactant

Herbicides can be applied as a broad-cast spray from a backpack sprayer forsmall areas or with a mechanized rig usinga farm tractor or an all-terrain vehicleBanded spraying (spraying in between thetree plantings) may be effective if theweed complex is known beforehand and isnot very vigorous

Woody Control After PlantingIdeally adequate site preparation will pre-clude the need to control woody speciesafter planting If some control of woodyspecies is needed after planting directedspray of foliar-active herbicides is the pre-ferred method Useful products includeAccordreg (glyphosate) and Garlonreg 3A(triclopyr) Spray drift must be minimizedand contact with crop species avoided(Miller 1993)

Recognizing AdverseEffects of Herbivory Herbivory can dramatically affect the sur-vival and growth of bottomland hardwoodseedlings The major herbivores are beaver(Castor canadensis) nutria (Myocastor coy-pus) and in some localized situationswhite-tail deer (Odocoileus virginianus)Small mammals mostly rodents (forexample hispid cotton rat [Sigmodonhispidus]) and rabbits (Sylvilagus spp) are

often responsible for failures of directlyseeded plantings

Overcoming HerbivoryThere are three basic measures thatforesters and restorationists in the LMAVuse to overcome the effects of herbivorousanimals They are fencing tree sheltersand reducing the amount of plant cover

FencingFencing has been used to increase the sur-vival of natural and planted seedlings byexcluding large herbivores such as deerfrom regeneration areas Cattle-wire fence(8-ft-tall) has proven most effective atexcluding deer in the northeasternUnited States (Marquis and Brenneman1981) Woven wire fence and debrisfences have been used in the southernUnited States to protect commercial cot-tonwood plantations from deer and hogs(McKnight 1970) In 2001 we fenced anexperimental area with 8-ft-high tensilesteel deer fence at an installed cost of$395 per linear foot which included thecost of installing 2-ft-tall poultry wire atthe base of the deer fence to exclude rab-bits For comparison a 10-acre (4-ha) sitewould cost $1471 or more dependingupon layout Although electric fencinghas proven effective in northern hard-woods (Marquis and Brenneman 1981)flooding makes this impractical in mostbottomlands

Tree SheltersThe benefits of tree sheltersmdashdecreasedherbivory stimulated seedling growth andincreased seedling survivalmdashhave beendocumented for northern climates mostlyin cutover natural stands (Frearson andWeiss 1987 Lantagne and others 1990Ponder 1995 Gillespie and others 1996)On bottomland sites subject to heavybrowsing tree shelters may be the onlymeans of successful afforestation (Conner1988 1993 Reed and McLeod 1994)Shelters may increase the competitivenessof slower-growing species such as oaks(Schweitzer and others 1999) but heightgains often are due to temporary shifts in

biomass accumulation and are not main-tained once seedlings grow above the shel-ter (Gardiner and others 2002)

Several types of tree protectiondevices have been tested over the past twodecades In the mid-1980s Conner andToliver (1987) experimented with Vexarreg

plastic mesh tubes and found that they didnot protect baldcypress seedlings fromnutria Plastic tree shelters from 2 ft (60-cm) to 5 ft (150-cm) tall have beentested in various experiments in southernbottomland and wetland sites (Conner1988 1993 Reed and McLeod 1994Schweitzer and others 1999 Conner andothers 2000) Double-wall plastic shelters(commercially available as Tubexreg orTreeProreg) protect seedlings from herbi-vores and create a microenvironmentwith increased carbon dioxide humidityand temperature (Figure 3)

Tree shelters are not a guaranteeagainst mortality from animal herbivoryhowever In most wet areas 1-ft (30-cm)tall tree shelters are generally sufficient toprevent clipping by rabbits or nutria buttaller shelters are necessary to preventexcessive browsing by deer Beaver havebeen observed grazing on seedlings whenfloodwater exceeds the height of the shel-ter (Reed and McLeod 1994) Installingshelters taller than the depth of expected


Beaver have beenobserved grazing onseedlings when flood-water exceeds the heightof the shelter Installingshelters taller than thedepth of expected floodlevels is the only way to prevent this type of herbivory

flood levels is the only way to prevent thistype of herbivory

The cost-effectiveness of tree sheltersfor large restoration areas is uncertainMaterial cost for tubular shelters is about$175 each for 2-ft-tall and $417 for 4-ft-tall shelters plus the costs of installationand removal Shelters are easily knockeddown and swept away by floodwatersBecause of their high cost and uncertaineffectiveness tree shelters probablyshould be limited to small areas of verysevere herbivory in conjunction withcontrol of weed cover (for small mam-mals) and herbivore suppression (beaverand nutria)

Reducing CoverSmall mammals are abundant on afforest-ation sites (Willis and others 1996) andoften are suspected of eating or cachingdirect-seeded acorns They also will clipseedling tops or girdle stems and clip roots(Savage and others 1996) Oak seedlingscan resprout and will usually overcomeanimal browsing (Lasher and Hill 1977Schweitzer and others 1997) provided itis not continuous Baldcypress also re-sprouts readily although green ash re-sprouts less readily and water tupelo not

at all (Conner and others 1999) On mostsites control of herbaceous vegetationremoves cover for small mammals andreduces their effect on seedlings Treeshelters also provide some protectionalthough some small mammals may tun-nel under the bottom of the shelter (PMadsen pers comm)

SummaryProper diagnosis of site conditions andselection of appropriate species are criticalto achieving adequate survival and growthof restored bottomland hardwoods in theLMAV Guidelines (Baker and Broadfoot1979 Hook 1984 Tables 1 and 2) areavailable for matching species to a givensite and hydroperiod according to theirtolerance to waterlogging and their growthpotential We think these guidelinesshould be combined with information onother adverse site conditions such as com-petition and herbivory in order to developprescriptions that are cost effective

Gardiner and his associates (2002)provide a comprehensive review of re-search needs for establishing bottomlandhardwoods under all site conditions Wewould like to suggest the followingresearch needs which we believe are spe-

cific to establishing bottomland hard-woods under adverse site conditionsThese needs include 1) better guidelinesfor properly fertilizing these former agri-cultural sites 2) experiments with tallerseedling and 3) determine better manage-ment and control of troublesome weeds

Fertilizing Agricultural SitesSeedling vigor can be enhanced by fertil-ization but guidelines are needed fornitrogen on former agricultural fields andphosphorous on less fertile sites (Francis1985 Stanturf and Schoenholtz 1998Gardiner and others 2002) Continuouscropping depletes soil organic matter andassociated nutrients particularly nitrogenNitrogen is routinely added in cotton-wood plantings as liquid fertilizer in theplanting slit (Stanturf and others 2001b)to boost early height growth Mostresearch with other species has shownthem to be less responsive to fertilizersalthough evaluations have been for long-term effect on biomass production It alsocould be cost effective to maximize short-term growth in height by placing aseedling beyond the range of deer orflooding and increase its ability to com-pete with weeds Species characteristicsand site conditions will determine whento add the correct fertilizer in the optimalamounts Generally fertilization accom-panied by weed control produces the bestresults

Planting Taller SeedlingsSimply planting taller seedlings may over-come many limitations associated withflooding and herbivory Seedlings tallerthan floodwaters should withstand evensummer flooding as long as high watertemperature does not reduce dissolvedoxygen to lethal levels (Kennedy andKrinard 1974) Tall seedlings with upperleaves beyond the reach of deer mayreduce some effects of herbivory

Some work suggests that planting tallstock is feasible McKnight (1970) forexample described a technique for plant-ing rooted cuttings of cottonwood in areassubject to deep overflows Whips with a16-20 ft (5-6 m) long sprout and 2 feet (60


Figure 3 Double-walled tree shelters are used in bottomland hardwood restorations to pro-

tect young trees from herbivory by deer rabbits and nutria and to create a microenvironment

that accelerates tree growth Photo by Wayne Inabinette

cm) of belowground material were plantedin a hole 40 inches (1 m) deep McKnightreported that the initial growth was poorbut survival and subsequent growth werehigh Stanturf (1995) and Stanturf andKennedy (1996) studied the results ofplanting 2-0 bareroot cherrybark oakseedlings on a cutover site in South Caro-lina Two planting depths (1 ft and 2 ft [30cm and 60 cm]) and top pruning werecompared After 11 years the researchersfound that the cherrybark oaks averaged236 feet (72 m) in height They alsonoted that there were no significant dif-ferences in height or dbh among treat-ments and survival exceeded 50 percent

The main disadvantage of tallerseedlings is the difficulty of planting themwhich leads to higher costs In other stud-ies (Stanturf 1995 Stanturf and Kennedy1996) oak seedlings up to 25 feet (08 m)tall were planted This process requiredthe use of a gasoline-powered postholedigger that required two people to operateand a separate two-person crew to placethe seedling and backfill If fewer tallseedlings can be planted and replanting offailed seedlings avoided the cost differen-tial may not be so great Tree shelters maystimulate height growth and provide addi-tional physical protection from herbivory

Better Control of Competing PlantsBottomland hardwood plantings underWRP seldom benefit from effective con-trol of competing vegetation (Stanturfand others 1998 Stanturf and others2001a) The standard practice is ldquodiskplant and walk awayrdquo without controllingwoody vine competition Some agenciesand a few landowners are averse to usingherbicides and cost-sharing programssuch as the Conservation Reserve Pro-gram seldom reimburse establishmentexpenses incurred after the first year Yetwoody vines and non-native invasiveplants are only controlled effectively byherbicide applied prior to planting andoften require spot application for twoyears after planting (Gardiner and others2002) Herbaceous species such as giantragweed (Ambrosia trifida) are also fiercecompetitors for site resources and must be

controlled when establishing cottonwood(Stanturf and others 2001b) The eco-nomic benefit of competition control forother hardwoods has not been docu-mented but small trials show promisingresults in terms of survival and earlyheight growth

Mechanical weed control is usedeffectively on some better drained sitesbut may be impossible on adverse sitesbecause saturated soil hinders operationsat critical times Fabric mats may providean alternative to herbicides in these situ-ations To control weeds effectively fabricmats must be applied early remain intactand be large enough to provide theseedling with protection (Haywood1999) Limited results show the promiseof fabric mats improving survival andgrowth (Adams 1997 Schweitzer andothers 1999) Fabric mats on flooded sitesmay be ineffective however as floodwatermay lift and float the mats away The costof mats may prohibit use in afforestinglarge areas Although the cost of materialsis moderate (less than $050 each withstaples depending upon material and sizeof mat) installation costs are highbecause each mat must be anchored(McDonald and Helgerson 1990 Hay-wood 1999) Nevertheless the potentialof mulch mats and tree shelters deservefurther testing especially under condi-tions of interacting stressors such as her-bivory and flooding

Tree shelters weed control and fer-tilizers may stimulate height growth ofnormal-sized seedlings sufficiently toovercome effects of growing season flood-ing Additional research is needed partic-ularly side-by-side comparisons toidentify cost-effective combinations ofstock type vegetation control fertiliza-tion and protective devices

REFERENCESAdams JC 1997 Mulching improves early

growth of four oak species in plantationestablishment Southern Journal of AppliedForestry 2144-46

Allen JA 1990 Establishment and growth ofbottomland oak plantations on the YazooNational Wildlife Refuge Complex South-ern Journal of Applied Forestry 14206-210

Allen JA BD Keeland JA Stanturf AFClewell and HE Kennedy Jr 2001 A guide

to bottomland hardwood restoration USGeological Survey Biological ResourcesDivision Information and TechnologyReport USGSBRDITR-2000-0011 U SDepartment of Agriculture Forest ServiceSouthern Research Station GeneralTechnical Report SRS-40

Anon 1999 1999 Crop protection reference15th ed New York CampP Press

Baker JB and WM Broadfoot 1979 Siteevaluation for commercially importantsouthern hardwoods US Department ofAgriculture Forest Service Southern ForestExperiment Station General TechnicalReport SO-26 New Orleans Louisiana

Brantley EF and WH Conner 1997 Growthof root-pruned seedlings in a thermallyimpacted area in South Carolina TreePlanterrsquos Notes 48(3-4)76-80

Conner WH 1988 Natural and artificialregeneration of baldcypress (Taxodium dis-tichum [L] Rich) in the Barataria Basins ofLouisiana PhD dissertation BatonRouge LA Louisiana State University

__ 1993 Artificial regeneration of baldcypressin three South Carolina forested wetlandareas after Hurricane Hugo Pages 185-188in JC Brissette (ed) Proceedings of theseventh biennial southern silviculturalresearch conference US Department ofAgriculture Forest Service SouthernForest Experiment Station New OrleansLouisiana General Technical ReportGTR-SO-93

Conner WH and K Flynn 1989 Growthand survival of baldcypress (Taxodium dis-tichum [L] Rich) planted across a floodinggradient in a Louisiana bottomland forestWetlands 9207-217

Conner WH LW Inabinette and ELFunderburke 2000 The use of tree sheltersin restoring forest species to floodplaindelta 5-year results Ecological Engineering15(Suppl 1)S47-S56

Conner WH KW McLeod LWInabinette VH Parrish and MR Reed1999 Successful planting of tree seedlingsin wet areas Pages 201-204 in JDHaywood (ed) Proceedings of the tenthbiennial southern silvicultural researchconference US Department of Agri-culture Forest Service Southern ResearchStation Asheville North CarolinaGeneral Technical Report SRS-30

Conner WH and JR Toliver 1987 Vexarseedling protectors did not reduce nutriadamage to planted baldcypress seedlingsTree Planterrsquos Notes 3826-29

Conner WH JR Toliver and GR Askew1993 Artificial regeneration of baldcypressin a Louisiana crayfish pond SouthernJournal Applied Forestry 1754-57

DrsquoAntonio C and LA Meyerson 2002Exotic plant species as problems and solu-tions in ecological restoration A synthesisRestoration Ecology 10703-713


Dynesius M and C Nilsson 1994 Frag-mentation and flow regulation of river sys-tems in the northern third of the worldScience 266754-762

Ezell AW 1999a Crop tolerance of five hard-wood species to pre-emergent and post-emergent applications of oxyfluorfenProceedings of the Weed Science Society ofAmerica 39183

__ 1999b Pre-emergent applications ofazafenidon for crop tolerance testing inloblolly pine cherrybark oak Nuttall oakand willow oak Proceedings of the WeedScience Society of America 39182

Ezell AW and AL Catchot Jr 1997 Pre-emergent and post-emergent applications ofsulfometuron methyl for competition con-trol in oak plantations Proceedings of theSouthern Weed Science Society 50109-111

Francis JK 1985 Bottomland hardwood fer-tilizationmdashthe Stoneville experiencePages 346-350 in E Shoulders (ed)Proceedings of the third biennial southernsilvicultural research conference USDepartment of Agriculture Forest ServiceSouthern Forest and Experiment StationNew Orleans Louisiana General Tech-nical Report SO-54

Frearson K and ND Weiss 1987 Improvedgrowth rates within tree shelters QuarterlyJournal of Forestry 81184-187

Gardiner ES DR Russell M Oliver andLC Dorris Jr 2002 Bottomland hard-wood afforestation state of the art Pages75-86 in M Holland ME Warren Jr andJA Stanturf editors Proceedings of theconference on sustainability of wetlandsand water resources US Department ofAgriculture Forest Service SouthernResearch Station Asheville NC GeneralTechnical Report SRS-50

Gardiner ES CJ Schweitzer and JAStanturf 2001 Photosynthesis of Nuttalloak (Quercus nuttallii Palm) seedlingsinterplanted beneath an Eastern cotton-wood (Populus deltoides Bartr ex Marsh)nurse crop Forest Ecology and Management149283-294

Gillespie AR R Rathfon and RK Meyers1996 Rehabilitating a young northern redoak planting under tree shelters NorthernJournal of Applied Forestry 1324-29

Groninger JW WM Aust M Miwa andJA Stanturf 2000 Growth predictions fortree species planted on marginal soybeanlands in the Lower Mississippi ValleyJournal of Soil and Water Conservation5591-95

Groninger JW and DA Babassana 2002Accelerating planted green ash establish-ment on an abandoned soybean field Pages270-272 in KW Outcalt (ed) Pro-ceedings of the eleventh biennial southernsilvicultural research conference USDepartment of Agriculture Forest ServiceSouthern Research Station Asheville

North Carolina General Technical ReportSRS-48

Haywood JD 1999 Durability of selectedmulches their ability to control weeds andinfluence growth of loblolly pine seedlingsNew Forests 18 263-276

Hesse ID WH Conner and JW Day Jr1996 Herbivory impacts on the regenera-tion of forested wetlands Pages 23-28 inKM Flynn (ed) Proceedings of the south-ern forested wetland management andecology conference Clemson UniversityClemson South Carolina

Hook DD 1984 Waterlogging tolerance oflowland tree species of the south SouthernJournal of Applied Forestry 8136-149

Houston AE and ER Bucknor 1989Cultural treatments for the establishment ofbottomland hardwoods in west TennesseePages 285-288 in JH Miller (compiler)Proceedings of the fifth biennial southernsilvicultural research conference USDepartment of Agriculture Forest ServiceSouthern Forest Experiment Station NewOrleans Louisiana General TechnicalReport SO-74

Howell KD and TB Harrington 2002Container size and fertilization affect nurs-ery cost and fifth-year field performance ofcherrybark oak Pages 328-331 in KWOutcalt (ed) Proceedings of the eleventhsouthern silvicultural research conferenceUS Department of Agriculture ForestService Southern Research StationAsheville North Carolina GeneralTechnical Report SRS-48

Kennedy HE Jr 1981a Bottomland hard-woods research on site preparation planta-tion establishment and cultural treatmentsat the Southern Hardwoods LaboratoryPages 75-78 in JP Barnett (ed) Pro-ceedings of the first biennial southern silvicultural research conference USDepartment of Agriculture Forest ServiceSouthern Forest Experiment Station NewOrleans Louisiana General TechnicalReport SO-34

__ 1981b Foliar nutrient concentration andhardwood growth influenced by culturaltreatments Plant and Soil 63307-316

__ 1993 Artificial regeneration of bottom-land oaks Pages 241-249 in DL Loftis andCE McGee (eds) Oak regenerationSerious problems practical recommenda-tions US Department of AgricultureForest Service Southeastern Forest Experi-ment Station Asheville North CarolinaGeneral Technical Report SE-84

Kennedy HE Jr and RM Krinard 19741973 Mississippi River floodrsquos impact onhardwood forests and plantations USDepartment of Agriculture Forest ServiceSouthern Forest Experiment Station NewOrleans Louisiana Research Note SO-177

__ 1985 Shumard oaks successfully plantedon high pH soils US Department of

Agriculture Forest Service Southern ForestExperiment Station New Orleans Louisi-ana Research Note SO-321

King SL and BD Keeland 1999 Evaluationof reforestation in the Lower MississippiRiver Alluvial Valley Restoration Ecology7348-359

Lantagne DO CW Ramm and DI Dick-mann 1990 Tree shelters increase heightsof planted oaks in a Michigan clearcutNorthern Journal Applied Forestry 724-26

Lasher DN and EP Hill 1977 An evalua-tion of polypropylene mesh tubing as a deerbrowse deterrent for southern hardwoodseedlings Pages 239-245 in Proceedings ofthe annual conference Southeastern Asso-ciation of Fish and Wildlife Agencies

Ligon FK WE Dietrich and WJ Trush1995 Downstream ecological effects ofdams BioScience 45183-192

Marquis DA and R Brenneman 1981 Theimpact of deer on forest vegetation inPennsylvania US Department of Agri-culture Forest Service NortheasternExperiment Station Broomall Penn-sylvania General Technical Report NE-65

McDonald PM and OT Helgerson 1990Mulches aid in regenerating California andOregon forests Past present and futureUS Department of Agriculture ForestService Pacific Southwest Research Sta-tion Berkeley California General Tech-nical Report PSW-113

McKevlin MR DD Hook and AARozelle 1998 Adaptations of plants toflooding and soil waterlogging Pages 173-203 in MG Messina and WH Conner(eds) Southern forested wetlands Ecologyand management Boca Raton FloridaLewis PublishersCRC Press

McKnight JS 1970 Planting cottonwoodcuttings for timber production in theSouth US Department of AgricultureForest Service Southern Forest Experi-ment Station New Orleans LouisianaResearch Paper SO-60

Miller JH 1993 Oak plantation establishmentusing mechanical burning and herbicidetreatments Pages 264-289 in DL Loftisand CE McGee (eds) Oak regenerationSerious problems practical recommenda-tions US Department of AgricultureForest Service Southeastern Forest Experi-ment Station Asheville North CarolinaGeneral Technical Report SE-84

__ 1997 Exotic invasive plants in southeast-ern forests Paper presented at Exotic Pestsof Eastern Forests Conference April 8-101997 Nashville Tennessee

Ponder F 1995 Shoot and root growth ofnorthern red oak planted in forest openingsand protected by treeshelters NorthernJournal Applied Forestry 1236-42

Reed MR and KW McLeod 1994 Plantingunconsolidated sediments with flood-toler-ant species Pages 137-146 in FJ Webb Jr


(ed) Proceedings of the twenty-firstannual conference on wetlands restorationand creation Tampa Florida HillsboroughCommunity College

Savage L J Anthony and R Buchholz 1996Rodent damage to direct seeded willow oakin Louisiana Pages 340-349 in AGEversole (ed) Proceedings of the 15thannual conference of the SoutheasternAssociation of Fish and Wildlife Agencies

Schneider RL NE Martin and RR Sharitz1989 Impacts of dam operations on hydrol-ogy and associated floodplain forests ofsoutheastern rivers Pages 1113-1122 inRR Sharitz and JW Gibbons (eds)Freshwater wetlands and wildlife USDOEOffice of Scientific and Technical Informa-tion Oak Ridge Tennessee CONF-8603101 Symposium Series No 61

Schoenholtz SH JP James RK KaminskiBD Leopold and AW Ezell 2001Afforestation of bottomland hardwoods inthe Lower Mississippi Alluvial ValleyStatus and trends Wetlands 21602-613

Schweitzer CJ ES Gardiner JA Stanturfand AW Ezell 1999 Methods to improveestablishment and growth of bottomlandhardwood artificial regeneration Pages209-214 in JW Stringer and DL Loftis(eds) Proceedings of the 12th centralhardwood forest conference US Depart-ment of Agriculture Forest ServiceSouthern Research Station AshevilleNC General Technical Report SRS-24

Schweitzer CJ JA Stanturf JP ShepardTM Wilkins CJ Portwood and LCDorris Jr 1997 Large-scale comparison ofreforestation techniques commonly used inthe Lower Mississippi Alluvial Valley Firstyear results Pages 313-320 in SG PallardyRA Cecich HG Garrett and PSJohnson (eds) Proceedings of the 11thcentral hardwood forest conference USDepartment of Agriculture Forest ServiceNorth Central Forest Experiment StationSt Paul Minnesota General TechnicalReport NC-188

Shankman D 1999 The loss of free-flowingstreams in the Gulf Coastal Plain BulletinAlabama Museum of Natural History 201-10

Stanturf JA 1995 Survival and growth ofplanted and direct-seeded cherrybark oak inthe Santee River floodplain after 11 years

Pages 489-494 in JR Miller (ed)Proceedings of the 8th biennial silvicul-tural research conference US Departmentof Agriculture Forest Service SouthernResearch Station Asheville North Car-olina General Technical Report SRS-1

Stanturf JA ES Gardiner PB Hamel MSDevall TD Leininger and ML Warren Jr2000 Restoring bottomland hardwoodecosystems in the Lower Mississippi Allu-vial Valley Journal of Forestry 9810-16

Stanturf JA and HE Kennedy Jr 1996Survival and growth of planted and direct-seeded cherrybark oak in South CarolinaSouthern Journal of Applied Forestry20194-196

Stanturf JA and CJ Portwood 1999 Econ-omics of afforestation with cottonwoodPages 66-72 in JD Haywood (ed)Proceedings of the tenth biennial southernsilvicultural research conference USDepartment of Agriculture Forest ServiceSouthern Research Station AshevilleNorth Carolina General Technical ReportSRS-30

Stanturf JA and SH Schoenholtz 1998Soils and landforms Pages 123-147 in MGMessina and WH Conner (eds) South-ern forested wetlands Ecology and man-agement Boca Raton Florida LewisPublishersCRC Press

Stanturf JA SH Schoenholtz CJSchweitzer and JP Shepard 2001a Indi-cators of restoration success Myths in bot-tomland hardwood forests RestorationEcology 9189-200

Stanturf JA CJ Schweitzer and ESGardiner 1998 Afforestation of marginalagricultural land in the Lower MississippiRiver Alluvial Valley USA Silva Fennica32281-297

Stanturf JA C van Oosten M Coleman DNetzer and CJ Portwood 2001b Silvi-culture and ecology of poplar plantationsPages 153-206 in D Dickmann J Isebrandsand J Richardson (eds) Poplar culture inNorth America Ottawa Canada NationalResearch Council Press

Williams HM and MN Craft 1998 First-year survival and growth of bareroot con-tainer and direct-seeded Nuttall oakplanted on flood-prone agricultural fieldsPages 300-303 in T Waldrop (ed) Pro-

ceedings of the ninth biennial southern sil-vicultural research conference USDepartment of Agriculture Forest ServiceSouthern Research Station AshevilleNorth Carolina General Technical ReportSRS-20

Williams H and M Stroupe 2002 First-yearsurvival and growth of bareroot and con-tainer water oak and willow oak seedlingsgrown at different levels of mineral nutri-tion Pages 338-341 in KW Outcalt (ed)Proceedings of the eleventh biennial south-ern silvicultural research conference USDepartment of Agriculture Forest ServiceSouthern Research Station AshevilleNorth Carolina General Technical ReportSRS-48

Willis K C Woodson and P Hamel 1996Small mammal populations on agriculturalland in the Mississippi Alluvial Valleyrestored to forest first year results Paper pre-sented at the annual meeting of the WildlifeSociety Mississippi Chapter October 171996 Roosevelt State Park Mississippi

JA Stanturf is a project leader with the USDA ForestService 320 Green Street Athens GA 30606706559-4316 Fax 706559-4317 jstanturffsfedus

WH Conner is a professor at Clemson UniversityBaruch Institute of Coastal Ecology and ForestScience PO Box 596 Georgetown SC 29442843546-6323 Fax 843546-6296 wconnerclemsonedu

ES Gardiner is a research forester with the USDAForest Service Box 227 Stoneville MS 38776662686-3184 Fax 662686-3195 egardinerfsfedus

CJ Schweitzer is a research forester with the USDAForest Service PO Box 1387 Normal AL 35762256372-4230 Fax 256858-8275 cschweitzerfsfedus

AW Ezell is a professor at Mississippi State Uni-versity Forestry and Wildlife Research Center POBox 9680 Mississippi State MS 35762 662325-1688 Fax 662325-8726 aezellcfrmsstateedu


withstand saturated soil conditions inwhich oxygen is consumed in respirationand soil voids become filled with carbondioxide and accumulated metabolic prod-ucts of anaerobic microbes Some treespecies have developed adaptive traits(McKevlin and others 1998) and can accli-mate to waterlogging but at a cost to theplant Only a few species can withstandwaterlogging for extended periods of timeduring the growing season (Table 1) Evenbaldcypress (Taxodium distichum) and watertupelo (Nyssa aquatica) species that growin deepwater swamps require non-floodedconditions for their seeds to germinate

Immersion of seedlings after theyhave leafed out in the spring by overbankor backwater flooding can cause mortal-ity of even the so-called flood-tolerantspecies During the spring and early sum-mer of 1973 Kennedy and Krinard (1974)observed the effects of a major flood onthe Mississippi River on planted and nat-ural stands One growing season after theflood natural and plantation stands thatwere one year or older at the time of theflood and inundated for two months orless were not severely affected eventhough the seedlings were less than 2 feettall These stands included sweet pecan

(Carya illinoensis) water oak (Quercusnigra) Nuttall oak (Q nuttallii) and cher-rybark oak (Q falcata var pagodaefolia)On the other hand 1-year-old sweetgum(Liquidambar styraciflua) which wasflooded for three months was killed Newplantings of cottonwood (Populus del-toides) cuttings were destroyed but theseedlings seemed to do better

Diagnosing hydroperiod limitations ofpotential afforestation sites may not beeasy however Widespread and numerousflood control structures and the sub-sequent regulation of river flows havechanged the seasonality and extent of floodevents (Dynesius and Nilsson 1994 Ligonand others 1995 Shankman 1999) Posi-tion in the landscape (for example ridgeflat or slough) and soil characteristics areindicative but not diagnostic (Baker andBroadfoot 1979) since land leveling anddrainage due to farming practices canchange site conditions We recommendobtaining at least a five-year history offlooding on the site before choosing thespecies to plant Landowners and farmmanagers can provide good informationabout such things as how often in the lastfive crop years planting was delayed or acrop was lost due to high water

Overcoming Growing-Season FloodingMatching species to site is absolutely crit-ical Species may be planted under less fre-quently flooded conditions than shownfor their tolerance class (Table 2) but notthe reverse For example baldcypress canbe planted and survive on ridges but cher-rybark oak should never be planted insloughs If management objectives allowflexibility in the choice of species simplyplant the species adapted to the worstprobable flooding conditions Alterna-tively one can plant a mixture of speciesadapted to a range of flooding regimes andexpect significant mortality because somespecies will fail to meet the site conditionsHowever when species composition andstocking are critical (for example if finan-cial returns from either timber or carbonsequestration payments is an objective)then other strategies such as controllingflooding delayed planting and planting instanding water are available to overcomegrowing-season flooding

Controlling Flooding Controlling flooding during the treesrsquoestablishment phase may be possible insome situations Restorationists may beable to keep sites managed as greentreereservoirs constructed wetlands or thosedownstream from water control struc-tures from flooding while seedlings ofmoderately tolerant species becomeestablished If flooding is caused bybeaver dams intensive trapping and con-trol may be required for a few years in asmall area while seedlings become tallenough to withstand flooding (C Sloanpers comm)

Delayed PlantingBareroot seedlings should be dormantwhen planted which means December toMarch in the LMAV Some sites are fre-quently under water during this optimumplanting window thereby hamperingrestoration efforts Avoiding flooded con-ditions may be possible although it isoften unreliable and expensive In back-water areas for example waiting to plant


Figure 1 Workers carrying bags full of dormant tree saplings mostly oaks fan out to plant a

former agricultural field in the Lower Mississippi River Alluvial Valley to bottomland hardwood

forest Photo by Emile Gardiner






























































to planting



for control


may be required

















































































































































































to planting



for control


may be required




















































































































recognizing and overcoming difficult site conditions for afforestation of bottomland hardwoods

Documents