drescher 2006 the role of grass stems as structural foraging deterrents and their effects on the...

8/3/2019 DRESCHER 2006 the Role of Grass Stems as Structural Foraging Deterrents and Their Effects on the Foraging Beha…

http://slidepdf.com/reader/full/drescher-2006-the-role-of-grass-stems-as-structural-foraging-deterrents-and 1/17

The role of grass stems as structural foragingdeterrents and their effects on the foraging

behaviour of cattle

Michael Dreschera,b,

*, Ignas M.A. Heitk oniga

,Jan G. Raats c, Herbert H.T. Prins a

a Resource Ecology Group, Wageningen University, 6708 PD Wageningen, The Netherlands

b Department of Integrative Biology, University of Guelph, Guelph, Ont. N1G 2W1, Canadac

Faculty of Agricultural and Environmental Sciences,

University of Fort Hare, Alice 5700, South Africa

Accepted 16 January 2006

Abstract

Little quantitative information is available about the role of stems as structural foraging deterrents

for large grazers and the actual mechanisms by which such deterrents affect foraging behaviour. We

measured bite size, bite rate and the rate of forage intake of cattle foraging on artificial micro-swards

of the tropical, broad-leaved guinea grass. These micro-swards varied in total forage mass density, in

forage quality, defined as the proportion of high-quality plant parts (leaves) and of foraging deterrents

(stems), and in the spatial pattern of plant parts. We hypothesized that stems interfere with the process

of grasping of leaves and predicted that decreasing forage quality, by reducing bite size and bite rate,

depresses the slope and the asymptotic maximum of the functional response curve. Further we

hypothesized that increasing cluster size of leaves increases the accessibility of leaves to cattle, thusalleviating the negative effects of decreasing forage quality, and predicted that increasing leaf cluster

size has positive effects on bite size and on the rate of forage intake.

The slope of the functional response curve decreased with decreasing forage quality, mainly

because of depressed bite size. These effects were also found when the amount of leaves in the grass

sward was kept constant while only the amount of stems increased. Thus, the observed effects are not

merely the result of decreased availability of leaves, but at least in part caused by the increasing

www.elsevier.com/locate/applanim

Applied Animal Behaviour Science xxx (2006) xxx–xxx

* Corresponding author at: Ministry of Natural Resources, Ontario Forest Research Institute, 1235 Queen Street

East, Sault Ste. Marie, Ont. P6A 2E5, Canada. Tel.: +1 705 946 7406; fax: +1 705 946 2030.

E-mail address: [email protected] (M. Drescher).

0168-1591/$ – see front matter # 2006 Elsevier B.V. All rights reserved.

doi:10.1016/j.applanim.2006.01.011

APPLAN-2506; No of Pages 17



interference of stems with the grasping of leaves. Leaf cluster size had a positive effect on bite size.

However, this effect did not show in the rate of forage intake, because for small leaf clusters high bite

rates compensated for decreased bite sizes. Instead of the familiar negative relationship of bite rate

with bite size, we found a positive relationship. We speculate that this effect is the result of decreasedchewing times for small bites and increased time needed to grasp these bites, effectively changing the

limitation of bite rate from chewing time to a limitation by the time needed to grasp bites.

# 2006 Elsevier B.V. All rights reserved.

Keywords: Bite rate; Bite size; Cattle; Forage intake rate; Grazer; Sward structure

1. Introduction

The functional response (i.e., the relationship of forage intake rate with forageavailability, Holling, 1959) plays a central role in the interaction of a forager with a forage

resource. Because forage intake has consequences for animal fitness and production (e.g.,

Romney et al., 2000; Smallegange and Brunsting, 2002; Landete-Castillejos et al., 2003),

quantifying the effects of resource structure on the rate of forage intake and the shape of the

functional response curve (e.g., Durant et al., 2003) is critical for furthering our

understanding of foraging ecology.

Theoretical studies have shown that the rate of forage intake in large herbivores is

mainly determined by bite size (Spalinger and Hobbs, 1992; Parsons et al., 1994;

Farnsworth and Illius, 1996). This notion is in line with various empirical studies in mostly

temperate environments (Spalinger et al., 1988; Gross et al., 1993a, 1993b; Ginnet and

Demment, 1995). However, when processing capacity is not limiting, bite rate also has the

potential to affect the rate of forage intake (Spalinger and Hobbs, 1992).

Nevertheless, explaining the variation in the rate of forage intake with bite size is only a

proximate answer. The ultimate question is which characteristics of the forage resource

determine bite size and bite rate and thus, ultimately, the rate of forage intake.

In simple, uniformly leafy swards, grass sward height (m), grass surface density

(g mÀ2), and grass bulk density (g mÀ3) were found to determine bite size of grazers (Black

and Kenney, 1984; Arias et al., 1990; Ungar et al., 1991; Laca et al., 1992a, 1992b, 1994;

Gong et al., 1996; Bakker et al., 1998; WallisDeVries et al., 1998; Wilmshurst et al., 1999).However, mature natural grass swards are a complex mixture of various plant parts (e.g.,

leaves and stems). The structural and chemical properties of plant parts in grasses differ

(Van Soest, 1994), though not as markedly as in woody plants. Some studies have indicated

that grass stems can act as structural foraging deterrents in grasses, much like thorns in

shrubs, and can limit bite size (Arias et al., 1990; Ginnet et al., 1999; Bergman et al., 2000).

However, most of these studies were restricted to investigating the effects of grass stems in

sub-canopy horizons of the grass sward on bite depth. To our knowledge, only the study of

Hongo (1998) dealt with the effects of grass stems in the grass sward canopy itself.

Unfortunately, his experimental design did not lend itself to clearly distinguish between the

effects of a number of sward characteristics. Past studies on the effects of grass stems onbite rate yielded contradictory results because some studies found a negative effect (Ruyle

et al., 1987; Ginnet et al., 1999), while other studies found no effect (Arias et al., 1990;

M. Drescher et al. / Applied Animal Behaviour Science xxx (2006) xxx–xxx2



Hongo, 1998). Neither of these studies addressed the question whether the spatial pattern of

structural foraging deterrents in the sward canopy determine bite size and bite rate, and thus

the rate of forage intake.

The objectives of this paper are to show that grass stems in the grass sward canopy canact as structural foraging deterrents to cattle by depressing both bite size and bite rate, thus

decreasing the rate of forage intake and changing the shape of the functional response

curve. Further, we want to show how this effect varies with the spatial pattern of the plant

parts in the grass sward canopy and with season.

If we define forage quality as the proportion of high-quality forage parts (i.e., grass

leaves) in the grass sward, then a decrease in forage quality means an increase in the

proportion of foraging deterrents (i.e., grass stems) in the grass sward canopy. For a

given total forage mass density, decreasing forage quality means an increase in the

density of foraging deterrents. Given that cattle select for leaves, we hypothesized that

stems interfere with the process of grasping of leaves and that therefore the proportion of

stems in the grass sward has negative effects on forage ingestion. More specifically, we

predicted that decreasing forage quality has the following effects: (i) because foraging

deterrents stand closer together, cattle decrease bite size showing in a reduced slope of

the bite size-forage mass relationship; (ii) because foraging deterrents stand closer

together, cattle have to spend more time to locate and crop interspersed leaves, thus

decreasing bite rate; (iii) mainly because of the effect on bite size, the slope and the

maximum potential rate of forage intake of the functional response curve will be

depressed.

For a given forage quality (i.e., the same proportion of grass leaves and grass stems) wedefine spatial pattern as the size (surface area) of clusters of high-quality forage parts

(grass leaves) in a matrix of foraging deterrents (grass stems). We hypothesized that

increasing cluster size of leaves increases the accessibility of these forage parts and

alleviates the negative effects of decreasing forage quality, as outlined above. In other

words, for a given forage mass and forage quality, with increasing cluster size, we

expected to observe: (i) an increase in bite size, because the cropping of leaves is less

hindered by the surrounding stems; (ii) a change in bite rate that can either be positive, if

bite rate is limited by cropping, or negative, if bite rate is limited by the time between the

cropping of bites; (iii) an increase in the rate of forage intake mainly due to the effect of

cluster size on bite size.

2. Materials and methods

The study was conducted at the Research Farm of the University of Fort Hare, at Alice,

Eastern Cape, South Africa. We performed experiments with cattle ( Bos taurus L.),

foraging on hand-constructed micro-swards of the (sub-)tropical guinea grass (Panicum

maximum Jacq.), an erect, broad-leaved species. The experimental swards differed in total

mass, in the proportion of grass leaves and grass stems, and in the size (surface area) of

clusters of leaves surrounded by grass stems. The experiments were performed during thelate wet season and repeated during the early dry season to additionally investigate

potential effects of season on foraging behaviour.

M. Drescher et al. / Applied Animal Behaviour Science xxx (2006) xxx–xxx 3



2.1. Experimental swards

All experimental work was performed using hand-constructed swards (Black and

Kenney, 1984; Laca et al., 1992b). Each morning between 07:30 and 09:30 h, tillers of guinea grass were clipped in the field. At all times, plant parts were kept in water, sprayed

and, if necessary, covered with plastic to prevent wilting. Grass leaves and grass stems were

separated and then threaded through holes in wooden blocks. In each hole we inserted

either three leaves (subsequently we call this one leaf tiller), or one stem, or holes were left

empty. In each block of 10 cm Â 10 cm there were 16 holes in a square grid at equal

distance. In each block a minimum of one hole and a maximum of eight holes were filled

with leaves or stems, with the remaining holes left empty. Plant parts were clamped to the

wooden block by shifting a plastic plate underneath the block and screwing it tightly

against it. We fastened between 40 and 50 blocks on a large plastic board, to produce one

hand-constructed sward of between 0.4 and 0.5 m2. Leaf tillers and stems were always

trimmed to produce a uniform canopy height of 15 cm. Hand-constructed swards were

offered to the animals in a shed between 11:00 and 15:00 h.

2.2. Animals

We used four Nguni oxen of 1.5 to 2 years, weighing 250–320 kg. Two of the oxen were

oesophageally fistulated more than 4 months earlier for other experiments. All animals

were duly cared for by a veterinarian. When not in experimental trials and between

experimental series, the animals were kept on a rangeland, but had daily contact with theexperimenters. During the period of the experiments, the animals were not fed dietary

supplements.

2.3. Experimental design

The experiments were performed from February to March 2000, during the late wet

season, and from May to June, during the subsequent early dry season. Before each

experimental series, the animals were trained for at least 3 weeks to accustom them to the

experimental set-up. Each day, before beginning with the experimental trial, the animals

were collected from the rangeland and brought to a shed, where they were fasted for 2.5–5 h. Subsequently, one animal was led to the hand-constructed sward and allowed to graze.

The grazing process was video taped and the observation was ended after 50–90% of the

high-quality forage parts were removed.

Swards differed in total mass (i.e., g leaf + g stem), forage quality (i.e., the proportion of

leaves = g leaf/g total Â 100%), and spatial pattern of leaves and stems (Fig. 1). All masses

in this study are given as dry matter. We varied total mass by changing the number of filled

holes (i.e., by changing the bulk density). Because of logistic constraints, we used

incomplete factorial designs. We offered swards of: low (approximately 20 g mÀ2),

medium (approximately 80 g mÀ2), or high (approximately 180 g mÀ2) total mass. Forage

quality was low (approximately 25% leaves), medium (approximately 45% leaves), or high(100% leaves). The spatial pattern of leaf tillers and grass stems was varied by building leaf

clusters of varying surface area: small (25 cm2), medium (100 cm2), large (200 cm2), and




very large (400 cm2). A leaf cluster, in this study, is one leaf tiller or a number of leaf tillers

directly adjacent to one another. The perimeter of a leaf cluster is spatially defined by thestems surrounding it.

In the first experiment on forage quality, leaf tillers and stems were always arranged

uniformly over the grass sward, thus avoiding clustering of leaf tillers or stems. The

combinations of total forage mass and forage quality used for this experiment were:

20 g mÀ2 with 45% and 100% leaves, 80 g mÀ2 with 25%, 45%, and 100% leaves;

180 g mÀ2 with 25% and 45% leaves. The arrangement of leaf tillers and stems in these

treatments are shown in Fig. 1.

In the second experiment on spatial pattern of plant parts, we arranged tillers of leaves in

various degrees of clustering in a matrix of grass stems. Total forage mass (approximately

80 g mÀ2

) and forage quality (approximately 45% leaves) were kept constant over alltreatments. The spatial pattern of plant parts in this experiment are shown in Fig. 1.

2.4. Measurements

We measured the time of active grazing (s) and counted the number of bites using the

video recordings; periods when animals engaged in other activities were excluded from the

measurements. Bite rate (sÀ1) was calculated as active grazing time divided by the number

of bites.

Available pre-grazing forage mass (g) was estimated via randomly selected and

removed blocks. Total ingested forage mass (g) was calculated as pre-grazing forage massminus the residual forage mass after grazing. Bite size (g) was calculated as total ingested

forage mass divided by the number of bites. Average intake rate (g sÀ1) was calculated as


Fig. 1. Spatial arrangements of leaf tillers and stems of guinea grass on hand-constructed swards that were used

for foraging experiments with cattle. At the top left of the figure, an entire hand-constructed sward of 5 Â 10 = 50

blocks is shown. At the top right of the figure, a part of these 50 blocks is shown representing various treatments(‘h’–‘j’), with each block being filled with only leaf tillers (L) or only stems (S), or left empty (/). At the bottom of

the figure, single blocks are shown (‘a’–‘g’), with each block filled with leaf tillers, stems, or leaf tillers and stems.

To construct uniform swards, 50 blocks of only one type (‘a’–‘g’) were used for each sward. To construct swards

with different leaf cluster sizes, two different types of blocks were used: blocks with only leaf tillers and blocks

with only stems were arranged in groups of different sizes (‘h’–‘j’). Hand-constructed swards made of blocks of

type ‘d’ were also used for the smallest leaf cluster size. The resulting leaf cluster sizes were ‘d’ = 25 cm2,

‘h’ = 100 cm2, ‘i’ = 200 cm2, and ‘j’= 400 cm2.



totally ingested forage mass divided by active grazing time. In the remainder of this paper,

forage mass is expressed as mass per area, thus as forage mass density (g mÀ2). Each of the

four animals was exposed to each treatment and animals were treated as replicate

measurements.

2.5. Statistical analyses

In the first experiment, we examined the effects of forage quality and total forage mass

density on the rate of forage intake, bite size, and bite rate using mixed ANCOVAs. We

treated season and forage quality as fixed factors, individual as random factor, and total

forage mass density as covariate. We limited our analyses to main factors and two-way

interactions and excluded non-significant terms. When a significant interaction effect of

forage quality with total forage mass density on forage intake rate or bite size was found,

we performed unplanned pairwise comparisons of slopes between forage qualities to

identify significant differences between slopes with sequential Bonferroni correction

(Rice, 1989). For bite rate, we performed unplanned pairwise comparisons of least squares

means with sequential Bonferroni correction between forage qualities to identify

significant differences between means. In the second experiment, for a given total forage

mass density, we examined the effects of the spatial pattern of plant parts on the rate of

forage intake, bite size, and bite rate using mixed ANOVAs. We treated season and leaf

cluster size as fixed factors, and individual as random factor. We limited our analyses to

main factors and two-way interactions and excluded non-significant factors. We

performed unplanned multiple comparisons between least squares means of leaf clustersizes and identified significant differences between means with sequential Bonferroni

correction.

3. Results

3.1. Forage quality

Our analysis indicated that the slope of the bite size-total forage mass density

relationship varied between forage qualities (F = 128.51, P < 0.0001, Fig. 2a).Comparison of slopes ((À0.29, 5.51, and 75.91) Â 10À4 m2, for 25%, 45%, and 100%

leaf, respectively) showed that decreasing forage quality depressed the slope of this

relationship (Table 1). Bite size increased with forage mass density for 100% and 45% leaf

(t = 11.83, P = 0.001; t = 3.60, P = 0.037, respectively), but there was no significant effect

of total forage mass density on bite size for 25% leaf (t = À0.14, P = 0.901). There was no

effect of season on bite size (F = 1.83, P = 0.269).

The relationship of bite rate with total forage mass density varied between forage

qualities (F = 5.61, P = 0.009, Fig. 2b). However, none of the slopes of this relationship

was significantly different from zero (all jt j < 2.87, all P > 0.064). Comparison of least

squares means (28.58, 36.82, and 57.21 minÀ

1, for 25%, 45%, and 100% leaf, respectively)showed that decreasing forage quality depressed bite rates for all contrasts (Table 1). There

was no effect of season on bite rate (F = 6.66, P = 0.082).





Fig. 2. Effects of forage quality on the relationship between bite size, bite rate, and forage intake rate and total

forage mass density for cattle. Shown are the experimental data and regression lines for the different forage

qualities: (a) the relationship between bite size and forage mass density (100% leaf: y = 1.8 Â 10

À1

g + 7.6 Â10À3 m2 Â x; 45% leaf: y = 1.6 Â 10À1 g + 5.5 Â 10À4 m2 Â x); (b) the relationship between bite rate and forage

mass density; (c) the functional response, i.e., the relationship between the rate of forage intake and forage mass

density (100% leaf: y = À1.7 g minÀ1 + 4.5 Â 10À1 m2 minÀ1 Â x).



The pattern of response of forage intake rate to changing total forage mass density

was very similar to the pattern shown by bite size (Fig. 2a and c). The slope of the

functional response curve varied between forage qualities (F = 128.51, P<

0.0001,Fig. 2c). Comparison of slopes ((À0.33, 1.48, and 44.99) Â 10À2 m2 minÀ1, for 25%,

45%, and 100% leaf, respectively) showed that decreasing forage quality depressed the

slope of the functional response curve (Table 1). Forage intake rate increased with total

forage mass density for 100% leaf (t = 10.15, P = 0.002), but for neither 45% leaf

nor 25% leaf the slope of the functional response curve was significantly different

from zero over the investigated ranges (t = 1.73, P = 0.182; t = À0.46, P = 0.680,

respectively). Our analysis did not indicate an effect of season (F = 2.52, P =

0.211).

Plotting the rate of forage intake against bite size again shows the effects of forage

quality on foraging behaviour (Fig. 3). The slope of the forage intake rate-bite sizerelationship can be interpreted as representing bite rate. The graph illustrates the decrease

in bite rate with forage quality, which is significantly higher for 100% leaf compared to

45% leaf (t = À3.30, P = 0.002). However, the remaining contrasts in this analysis give no

evidence for any further differences in bite rate (100% versus 25%: t = À1.52, P = 0.136;

45% versus 25%: t = À0.01, P = 0.999).

Through pairwise comparison of treatments of equal leaf mass density but different

stem mass densities, we uncoupled the effects of foraging deterrents and of high-quality

forage mass (Table 2). After controlling for individual effects and for season, increasing

stem mass density always decreased the rate of forage intake and bite rate. However, bite

size only decreased with stem mass density in the higher leaf mass density treatments.Over all comparisons, the effect of the foraging deterrents increased with leaf mass

density.


Table 1

Effects of forage quality, i.e., the proportion of leaves in the offered forage, on bite size, bite rate, and forage intake

rate of cattlea

Foraging behaviour variable Contrasted forage qualities d.f. t PBite size 100% vs. 45% 37 À14.43 <0.0001**

100% vs. 25% 37 À13.97 <0.0001**

45% vs. 25% 26 À2.04 0.026*

Bite rate 100% vs. 45% 6 À4.78 0.002**

100% vs. 25% 6 À5.96 0.001**

45% vs. 25% 6 À2.48 0.024*

Forage intake rate 100% vs. 45% 37 À14.01 <0.0001**

100% vs. 25% 37 À13.04 <0.0001**

45% vs. 25% 26 À1.21 0.118 ns

a

Unplanned multiple comparisons of slopes for bite size and forage intake rate, and of least squares means forbite rate. Uncorrected P-values for 1-tailed t -test.

ns: No significant difference after sequential Bonferroni correction.* Significant difference at a = 0.05 after sequential Bonferroni correction.

** Significant difference at a = 0.01 after sequential Bonferroni correction.



3.2. Cluster size

Bite sizes were significantly larger in the wet season than in the dry season (F = 14.77,

P = 0.031). Within seasons, bite size increased with leaf cluster size (F = 10.20, P = 0.003;

F = 5.87, P = 0.017, wet and dry season, respectively, Fig. 4a). Comparison of bite sizes

between leaf cluster sizes indicated that in both seasons bite sizes increased from 25 to100 cm2 leaf cluster size (0.26–0.50 g, and 0.23–0.38 g, for the wet and dry season,

respectively), but not thereafter (Table 3).


Table 2

Effects of decreasing forage quality, i.e., increasing stem mass density while keeping leaf mass density constant,

on forage intake rate, bite size, and bite rate of cattlea

Forage mass density Proportional change in foraging behaviourvariables from higher to lower forage quality

Higher forage quality Lower forage quality

Leaves

(g mÀ2)

Stems

(g mÀ2)

Leaves

(g mÀ2)

Stems

(g mÀ2)

Forage intake

rate (%)

Bite

size (%)

Bite

rate (%)

19.0 (1.5) 0.0 (0.0) 21.6 (1.2) 67.4 (4.3) À0.29 ns (0.29) 0.07 ns (0.16) À0.30* (0.17)

41.2 (1.6) 50.9 (3.5) 39.4 (1.7) 141.7 (6.1) À0.54** (0.11) À0.30* (0.15) À0.35** (0.09)

77.7 (3.6) 0.0 (0.0) 79.0 (4.8) 108.1 (3.9) À0.76*** (0.04) À0.60*** (0.04) À0.40*** (0.05)

a Averaged over animals and seasons, standard errors in parentheses. Each pair of compared treatments (higher

forage quality vs. lower forage quality) has the same leaf mass density, but different stem mass densities.

ns: P > 0.05.*P < 0.05.

**P < 0.01.

***P < 0.001.

Fig. 3. Effect of forage quality on the relationship between forage intake rate and bite size of cattle. Shown are theexperimental data and regression lines for the different forage qualities (100% leaf: y = À3.0 g minÀ1 + 59.4 minÀ1

Â x; 45% leaf: y = À0.3 g minÀ1 + 37.6 minÀ1 Â x; 25% leaf: y = À1.8 g minÀ1 + 38.0 minÀ1 Â x).




Fig. 4. Effects of leaf cluster size on bite size, bite rate, and forage intake rate of cattle. Shown are means with

standard deviation. Means were compared within each season (pooled over both seasons for bite rate) and



There was no significant effect of season on bite rate (F = 3.94, P = 0.141), but bite rate

changed with leaf cluster size (F = 5.87, P = 0.017, Fig. 4b). The analysis indicated that

bite rate for the 25 cm2 leaf cluster size was higher than for the 100 cm2 leaf cluster size

(39.9 and 27.0 minÀ1, for 25 and 100 cm2, respectively), but not different from the 200 to

400 cm2 leaf cluster size treatments. Bite rates did not differ between any of the 100, 200,

and 400 cm2 leaf cluster size treatments (Table 3).

Wet season rates of forage intake were significantly higher than in the dry season

(F = 12.92, P = 0.037). The analysis of variance did not indicate an effect of leaf cluster

size on the rate of forage intake within seasons (F = 1.27, P = 0.341; F = 1.51, P = 0.277,

wet and dry season, respectively, Fig. 4c). Comparison of mean forage intake rate

between leaf cluster sizes (wet season: 11.7 and 16.9 g minÀ1, dry season: 7.7 and

11.4 g minÀ1, for 25 and 400 cm2, respectively), did not indicate any significant

differences either (Table 3).


significant differences after sequential Bonferroni correction (P < 0.05) are indicated by different letters: (a) the

relationship between bite size and leaf cluster size for the wet and for the dry season; (b) the relationship between

bite rate and leaf cluster size pooled over both seasons; (c) the relationship between forage intake rate and leaf

cluster size for the wet and for the dry season.

Table 3

Effects of leaf cluster size on bite size, bite rate, and forage intake rate of cattle during the wet or dry season, or

during both seasonsa

Foraging behaviourvariable

Contrasted leaf cluster sizes

d.f. Both seasons Wet season Dry seasont P(2) t P(1) t P(1)

Bite size 25 cm2 vs. 100 cm2 9 À4.58 0.001** À3.58 0.003*

25 cm2 vs. 200 cm2 9 À4.97 0.001** À3.28 0.005*

25 cm2 vs. 400 cm2 9 À3.27 0.005* À3.40 0.004*

100 cm2 vs. 200 cm2 9 À0.39 0.353 ns 0.30 0.618 ns

100 cm2 vs. 400 cm2 9 1.32 0.890 ns 0.18 0.570 ns

200 cm2 vs. 400 cm2 9 1.71 0.939 ns À0.12 0.453 ns

Bite rate 25 cm2 vs. 100 cm2 9 3.93 0.004*

25 cm2 vs. 200 cm2 9 2.96 0.016 ns

25 cm2 vs. 400 cm2 9 1.55 0.155 ns

100 cm2 vs. 200 cm2 9 À0.97 0.359 ns

100 cm2 vs. 400 cm2 9 À2.38 0.041 ns

200 cm2 vs. 400 cm2 9 À1.41 0.192 ns

Forage intake rate 25 cm2 vs. 100 cm2 9 À1.00 0.171 ns À0.80 0.223 ns

25 cm2 vs. 200 cm2 9 À1.63 0.069 ns À1.67 0.065 ns

25 cm2 vs. 400 cm2 9 À1.74 0.058 ns À1.90 0.049 ns

100 cm2 vs. 200 cm2 9 À0.62 0.275 ns À0.87 0.203 ns

100 cm2 vs. 400 cm2 9 À0.74 0.239 ns À1.11 0.149 ns

200 cm2 vs. 400 cm2 9 À0.12 0.455 ns À0.23 0.410 ns

a Planned multiple comparisons of bite size, bite rate, and forage intake rate. P(1) and P(2): 1-tailed and 2-tailed

t -test, respectively. Uncorrected P-values.ns: No significant difference after sequential Bonferroni correction.* Significant difference at a = 0.05 after sequential Bonferroni correction.** Significant difference at a = 0.01 after sequential Bonferroni correction.



4. Discussion

The results of this study are in line with our prediction that decreasing forage quality

has negative effects on bite size, bite rate, and the rate of forage intake of cattle, thusreaching our first objective. Decreasing forage quality decreased the slope of the

functional response curve (Table 1 and Fig. 2c). Whereas the rate of forage intake

responded strongly to an increase in forage mass density for a forage quality of 100% leaf,

there was no significant increase in forage intake rate for a forage quality of either 45% leaf

or of 25% leaf. In the light of the already high densities of total forage mass offered in the

45% leaf and 25% leaf treatments, which scale up to about 2 tonnes of dry mass per

hectare, we do not anticipate that there would be any increases in forage intake rate if total

forage mass density was to be increased further. This implies that decreasing forage

quality did not only decrease the slope of the functional response curve, but likely also

depressed its asymptotic maximum. Also in agreement with our predictions is the

decreasing slope of the bite size-total forage mass density-relationship, and the decrease

in mean bite rate with decreasing forage quality. Our results are supported by the study of

Hongo and Akimoto (2004), who investigated the grazing behaviour of horses and cattle.

They found that exchanging half of the leaves of a uniformly leafy sward for grass stems,

reduced bite size, bite rate, and the rate of forage intake of both species. When comparing

the pattern of variation between the different behavioural variables (Fig. 2a–c), the similar

responses of forage intake rate and bite size give evidence that changes in the functional

response curve were mainly caused by the response of bite size to forage quality. The plot

of forage intake rate against bite size (Fig. 3) supports this latter notion, as indicated by theincreasingly strong limitation of bite size with decreasing forage quality and the

comparatively weaker response of bite rate.

For a given total forage mass density, we decreased forage quality by decreasing the

mass density of high quality forage (leaves) and at the same time increasing the mass

density of foraging deterrents (stems) in the grass sward. This caused the depression of bite

size, bite rate, and of the rate of forage intake. However, these changes were not merely a

response to a decreasing availability of high-quality forage. We also compared grass

swards, which had equal densities of leaves and differed only in the amount of stems. The

results show that even when the mass density of leaves was kept constant, just increasing

the amount of stems already depressed bite size, bite rate, and the rate of forage intake(Table 2). Thus, the observed changes in the foraging behaviour are not simply the

consequence of decreased availability of high-quality forage parts, but are at least in part

caused by the interference of stems with the foraging process.

We also partially reached our second objective, which was to show that the spatial

arrangement of forage parts, i.e., different degrees of clustering of leaves, affects

foraging behaviour of cattle. Increasing leaf cluster size had the predicted positive effect

on bite size (Table 3 and Fig. 4a). Bite size increased from leaf cluster sizes of 25–

100 cm2 and reached a plateau at a cluster size of 100 cm2. Also as predicted, we found

that bite rates responded to changes in leaf cluster size. Bite rate was highest for the

smallest leaf clusters (25 cm2), decreasing with increasing cluster size (100 cm2), butthen increasing again (200–400 cm2). However, based on our analysis, there is no

evidence that increasing leaf cluster size has a positive effect on the rate of forage intake,




though visual inspection suggests such an effect (Table 3 and Fig. 4c). It appears that

despite the negative response of bite size to decreasing cluster size, the high bite rates at

small leaf cluster sizes alleviated the negative effect of decreasing cluster size on the rate

of forage intake. Though we see a consistent trend of increasing forage intake rate withcluster size in both seasons, given the observed mean difference and sample variance, we

would have had to increase our sample size at least eight-fold to have a probability of

about 75% of finding a significant effect.

Bite size reached its maximum at a leaf cluster size area of 100 cm 2, which was

approximately equal to the maximum bite area of our cattle. We interpret this result as an

indication that stems surrounding clusters of leaves can impose an upper limit on bite size

as long as leaf cluster size is smaller than the maximum bite area. Because maximum bite

area is largely determined by muzzle width (or incisor arcade width: Gordon and Illius,

1988), it can be expected that the interaction between grass sward structure and grazers

depends on muzzle size. Because of the allometric relationship of muzzle size and body

size (Gordon and Illius, 1988), different sized foragers may be affected differently by grass

stems: smaller grazers may reach their maximum bite size at higher densities of foraging

deterrents and at smaller leaf cluster sizes. Therefore, we suggest that the observed

interaction between grass sward structure and forager may vary during an individual’s

ontogeny and between species.

We propose, that the negative effect of grass stems on bite size in cattle is comparable to

the effects of thorns and spines on bite size in browsers. In browsers, larger bites tend to

contain more woody support material. The large differences in structural and chemical

properties between plant parts in shrubs (i.e., leaves and branches), therefore lead to apositive correlation of fibre content with bite size. Consequently, bite size selection is

effectively limited by a trade-off between forge intake and diet quality (Shipley et al., 1999;

Wilson and Kerley, 2003). Structural foraging deterrents, like thorns or spines, can reduce

bite size to below its optimum as set by the intake-quality trade-off (Gowda, 1996;

Haschick and Kerley, 1997; Illius et al., 2002; Wilson and Kerley, 2003). Bite size can also

be affected by branch architecture and the spatial distribution of leaves along the branches

(Dziba et al., 2003). We caution, however, that the negative effect of grass stems on bite

size and forage intake could be less pronounced in softer grass species. Smit et al. (2005)

found no negative effect of stem mass on forage intake in cattle grazing ryegrass, though

this might also be explained by the low proportion of stems in their study (around 15%) andthe positive correlation of pseudostem mass and total forage availability.

The mixed response of bite rate to leaf cluster size, which is defined as the surface area

of leaf clusters, can be interpreted as an indication of two distinct domains of influence of

cluster size (Fig. 4b). We speculate that bite rate was highest for the smallest leaf clusters

(25 cm2) because small bites were located close to each other. If the time between

cropping of bites depends on the distance between bites, this would mean that the time

between cropping was shortest for the smallest leaf clusters and could thus explain the

higher bite rates. For larger leaf clusters, bite size and the distance between bites

increased, thus increasing the time between cropping and decreasing bite rate. For the

largest cluster size however, we speculate that several bites can be located next to eachother within one leaf cluster, thus decreasing the distance between bites and the time

between cropping, as well as that cropping time is minimal, thus increasing bite rate.




A manuscript of a detailed investigation of the effects of cropping time and time between

cropping on bite rate is in preparation.

The sward structure of natural grasslands commonly changes in the course of the

seasons due to phenology and forage consumption. Especially towards the end of and afterthe growth season, a general pattern is the decrease in the proportion of green and of leafy

material. During this time free-ranging, non-supplemented cattle tend to show a decrease

in live weight gain or even a loss of live weight (e.g., WallisDeVries, 1996). Our results

show the negative effects of decreased proportion of grass leaves on the rate of forage

intake. If cattle are not able to sufficiently increase daily grazing time after the growth

season, the decreased rate of forage intake will therefore cause the daily forage intake to

drop. Because of the positive relationship of forage intake with live weight gain, we can

also expect live weight gain to decrease during this period. Our findings therefore illustrate

an additional link between seasonal changes in sward structure and decreased live weight

gain after the growth season, besides of the commonly recognized effect of a drop in

nutritive value of the plants.

The results of the current study have implications for the modeling of herbivore foraging

behaviour. In their landmark paper, Spalinger and Hobbs (1992) described models of the

functional response in sparse and food-concentrated patches. According to their definition,

forage intake in food-concentrated patches is exclusively regulated by forage handling.

They proposed that the dynamics of forage intake in such patches arise from the

competition between cropping and processing of bites. Because larger bites need to be

chewed longer than smaller ones, bite size and bite rate show a negative relationship.

However, they also mentioned that it is believed that the degree of protection of bites byspines and thorns influences handling time and thus bite rate. The results of our study

illustrate circumstances in which bites are protected by grass stems to various degrees.

Fig. 5 shows the relationship of bite rate with bite size for a given total forage mass density

of about 80 g mÀ2 and changing forage quality. Higher forage quality means an increase in

the availability of preferred leaves and leads to larger bite sizes. If the effect of bite size on


Fig. 5. Effect of forage quality on the relationship between bite rate and bite size of cattle, given a total forage

mass density of 80 g mÀ2.



handling time were predominant as assumed by Spalinger and Hobbs (1992), then we

would expect a decrease in bite rate with increasing bite size. However, our results show an

increase of bite rate with increasing bite size. This finding is line with the study of Orr et al.

(2004), who found that both bite size and bite rate of cattle were decreasing duringincreasing depletion of a ryegrass sward. It appears that in our study the mechanistic

control of foraging deterrents over bite size and bite rate was so strong that it outweighed

the proposed effect of the competition between cropping and processing of bites. We

speculate that the presence of grass stems suppressed bite size so far and inflated the time

needed to grasp bites so much, that the chewing time of bites was shorter than the minimal

time between the cropping of bites. Then, with decreasing stem density, bite size increases

and grasping time decreases, leading to the observed positive relationship between bite rate

and bite size. We therefore suggest that the presence of stems in the grass sward in our

experiment transformed the foraging process from a situation where bite rate was limited

by chewing time, to a situation where bite rate was limited by grasping time. The density

and spatial pattern of high-quality plant parts (leaves) and foraging deterrents (stems),

therefore deserves more attention in the modeling of herbivore foraging behaviour.

5. Conclusion

Forage quality, as the proportion of high-quality forage parts, and the spatial pattern of

high-quality forage parts have strong effects on the rate of forage intake in cattle.

Primarily, these effects are caused by the strong negative response of bite size to thepresence of structural foraging deterrents, but to some extent also by the negative response

of bite rate. The presence of grass stems in the grass sward can transform the relationship

of bite rate with bite size from the familiar negative form to a positive one. The proportion

and spatial pattern of preferred forage parts and of foraging deterrents therefore deserves

more attention in the modeling of foraging behaviour of large herbivores in complex

grasslands.

Acknowledgements

We are grateful for the help provided by the University of Fort Hare. Special thanks go

out to W.S.W. Trollope, the staff of the Faculty of Agricultural and Environmental

Sciences, and the staff of the Research Farm of the UFH. The comments of F. van

Langevelde, W.F. de Boer, and an anonymous reviewer helped improving this manuscript.

This study was financially supported by a grant of The Netherlands Foundation for the

Advancement of Tropical Research (WOTRO, W84-441).

References

Arias, J.E., Dougherty, C.T., Bradley, N.W., Cornelius, P.L., Lauriault, L.M., 1990. Structure of tall fescue swards

and intake of grazing cattle. Agron. J. 82, 545–548.




Bakker, M.L., Gordon, I.J., Milne, J.A., 1998. Effects of sward structure on the diet selected by guanacos ( Lama

guanicoe) and sheep (Ovis aries) grazing a perennial ryegrass-dominated sward. Grass Forage Sci. 53, 19–30.

Bergman, C.M., Fryxell, J.M., Gates, C.C., 2000. The effects of tissue complexity and sward height on the

functional response of Wood Bison. Funct. Ecol. 14, 61–69.

Black, J.K., Kenney, P.A., 1984. Factors effecting diet selection by sheep. II. Height and density of pasture. Aust. J.

Agri. Res. 35, 565–578.

Durant, D., Fritz, H., Blais, S., Duncan, P., 2003. The functional response in three species of herbivorous Anatidae:

effects of sward height, body mass and bill size. J. Anim. Ecol. 72, 220–231.

Dziba, L.E., Scogings, P.F., Gordon, I.J., Raats, J.G., 2003. Effects of season and breed on browse species intake

rates and diet selection by goats in the False Thornveld of the Eastern Cape, South Africa. Small Ruminant

Res. 47, 17–30.

Farnsworth, K.D., Illius, A.W., 1996. Large grazers back in the fold: generalizing the prey model to incorporate

mammalian herbivores. Funct. Ecol. 10, 678–680.

Ginnet, T.F., Dankosky, G., Demment, M.W., 1999. Patch depression in grazers: the roles of biomass distribution

and residual stems. Funct. Ecol. 13, 37–44.

Ginnet, T.F., Demment, M.W., 1995. The functional response of herbivores: analysis and test of a simplemechanistic model. Funct. Ecol. 9, 376–384.

Gong, Y., Hodgson, J., Lambert, M.G., Gordon, I.L., 1996. Short-term ingestive behaviour of sheep and goats

grazing grasses and legumes. 2. Quantitative relationships between sets of swards and ingestive behaviour

variables. NZ J. Agri. Res. 39, 75–82.

Gordon, I.J., Illius, A.W., 1988. Incisor arcade structure and diet selection in ruminants. Funct. Ecol. 2, 15–22.

Gowda, J.H., 1996. Spines of Acacia tortilis: what do they defend and how? Oikos 77, 279–284.

Gross, J.E., Hobbs, T.N., Wunder, B.A., 1993a. Independent variables for predicting intake rate of mammalian

herbivores: biomass density, plant density, or bite size? Oikos 68, 75–81.

Gross, J.E., Shipley, L.A., Hobbs, N.T., Spalinger, D.E., Wunder, B.A., 1993b. Functional response of herbivores

in food-concentrated patches: test of a mechanistic model. Ecology 74, 778–791.

Haschick, S.L., Kerley, G.I.H., 1997. Browse intake rates by bushbuck (Tragelaphus scriptus) and boergoats

(Capra hircus). Afr. J. Ecol. 35, 146–155.Holling, C.S., 1959. Some characteristics of simple types of predation and parasitism. Can. Entomol. 91, 385–398.

Hongo, A., 1998. Selective grazing in pure leaf and leaf/culm mixtures of herbage grasses by sheep. J. Agri. Sci.

131, 353–359.

Hongo, A., Akimoto, M., 2004. The role of incisors in selective grazing by cattle and horses. J. Agri. Sci. 140, 469–

477.

Illius, A.W., Duncan, P., Richard, C., Mesochina, P., 2002. Mechanisms of functional response and resource

exploitation in browsing roe deer. J. Anim. Ecol. 71, 723–734.

Laca, E.A., Ungar, E.D., Seligman, N.G., Ramey, M.R., Demment, M.W., 1992a. Effects of sward height and bulk

density on bite dimensions of cattle grazing homogeneous swards. Grass Forage Sci. 47, 91–102.

Laca, E.A., Ungar, E.D., Seligman, N.G., Ramey, M.R., Demment, M.W., 1992b. An integrated methodology for

studying short-term grazing behavior of cattle. Grass Forage Sci. 47, 82–90.

Laca, E.A., Ungar, E.D., Demment, M.W., 1994. Mechanisms of handling time and intake rate of a largemammalian grazer. Appl. Anim. Behav. Sci. 39, 3–19.

Landete-Castillejos, T., Garcia, A., Gomez, J.A., Gallego, L., 2003. Lactation under food constraints in Iberian red

deer Cervus elaphus hispanicus. Wildlife Biol. 9, 131–139.

Orr, R.J., Rutter, S.M., Yarrow, N.H., Champion, R.A., Rook, A.J., 2004. Changes in ingestive behaviour of

yearling dairy heifers due to changes in sward state during grazing down of rotationally stocked ryegrass or

white clover pastures. Appl. Anim. Behav. Sci. 87, 205–222.

Parsons, A.J., Thornley, J.H.M., Newman, J., Penning, P.D., 1994. A mechanistic model of some physical

determinants of intake rate and diet selection in a two-species temperate grassland sward. Funct. Ecol. 8, 187–

204.

Rice, W.R., 1989. Analyzing tables of statistical tests. Evolution 43, 223–225.

Romney, D.L., Blunn, V., Sanderson, R., Leaves, J.D., 2000. Feeding behaviour, food intake and milk production

responses of lactating dairy cows to diets based on grass silage of high or low dry-matter content,

supplemented with quickly and slowly fermentable energy sources. Anim. Sci. 71, 349–357.




Ruyle, G.B., Hasson, O., Rice, R.W., 1987. The influence of residual stems on biting rates of cattle grazing

Eragrostis lehmanniana Nees. Appl. Anim. Behav. Sci. 19, 11–17.

Shipley, L.A., Illius, A.W., Danell, K., Hobbs, N.T., Spalinger, D.E., 1999. Predicting bite size selection of

mammalian herbivores: a test of a general model of diet optimization. Oikos 84, 55–68.

Smallegange, I.M., Brunsting, A.M.H., 2002. Food supply and demand, a simulation model of the functional

response of grazing ruminants. Ecol. Model. 149, 179–192.

Smit, H.J., Tas, B.M., Taweel, H.Z., Tamminga, S., Elgersma, A., 2005. Effects of perennial ryegrass ( Lolium

perenne L.) cultivars on herbage production, nutritional quality and herbage intake of grazing dairy cows.

Grass Forage Sci. 60, 297–309.

Spalinger, D.E., Hanley, T.A., Robbins, C.T., 1988. Analysis of the functional response in foraging in the Sitka

black-tailed deer. Ecology 69, 1166–1175.

Spalinger, D.E., Hobbs, N.T., 1992. Mechanisms of foraging in mammalian herbivores: new models of functional

response. Am. Nat. 140, 325–348.

Ungar, E.D., Genizi, A., Demment, M.W., 1991. Bite dimensions and herbage intake by cattle grazing short hand-

constructed swards. Agron. J. 83, 973–978.

Van Soest, P.J., 1994. Nutritional Ecology of the Ruminant. Comstock Publishing Associates, Ithaca, NY, USA.WallisDeVries, M.F., 1996. Nutritional limitations of free-ranging cattle: the importance of habitat quality. J.

Appl. Ecol. 33, 688–702.

WallisDeVries, M.F., Laca, E.A., Demment, M.W., 1998. From feeding station to patch: scaling up food intake

measurements in grazing cattle. Appl. Anim. Behav. Sci. 60, 301–315.

Wilmshurst, J.F., Fryxell, J.M., Colucci, P.E., 1999. What constraints daily intake in Thomson’s gazelles? Ecology

80, 2338–2347.

Wilson, S.L., Kerley, G.I.H., 2003. Bite diameter selection by thicket browsers: the effect of body size and plant

morphology on forage intake and quality. Forest Ecol. Manag. 181, 51–65.


drescher 2006 the role of grass stems as structural foraging deterrents and their effects on the...

Documents