• diet selection;
  • dietary quality;
  • nutrient intake;
  • plant diversity;
  • plant–herbivore interaction;
  • rangelands;
  • sheep;
  • voluntary food intake


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

1. There is general concern that local loss of plant diversity will adversely impact net primary productivity and other ecosystem properties. However, mechanisms linking plant diversity with other trophic levels, especially for large herbivores, are poorly understood.

2. We examine the responses of foraging sheep to changes in plant species richness in an indoor cafeteria experiment involving six plant species richness levels (1, 2, 4, 6, 8 and 11 species) and three plant functional group compositions within each level, and in a field experiment involving three plant species richness levels (1, 4–6 or >8 species).

3. Sheep preferred a diverse diet over a single diet even when palatable species were in the diet. Voluntary daily intake steadily rose with increases in plant species richness in both cafeteria and field experiments. The overall nutrient intake (i.e. daily energy and protein intakes) of sheep in the cafeteria also rose significantly with increased plant species richness until it reached a plateau at eight species. The quality of the diet selected by sheep was also significantly affected by plant species richness, but the variation of dietary quality was small and variable.

4. High nutrient acquisition by the sheep depended on selecting those palatable species with high nutrient content from the plant forage on offer together with the complementary effects of plant species richness, especially for plant functional group richness.

5.Synthesis and applications. Our experiments demonstrate an asymptotic relationship between plant species richness and voluntary intake by sheep. Increases in plant species richness from a low level led to increased daily nutrient intake, and presumably performance of the sheep. Natural grasslands are generally low in nutritional quality and so plant species richness will critically influence herbivore food intake and nutrition. The asymptotic relationship indicates that the maintenance of plant species richness in rangelands will benefit both domestic herbivore production and the conservation of biodiversity.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Changes in plant species richness and other components of plant diversity (see Stirling & Wilsey 2001; Wilsey et al. 2005), especially those influencing nutrient dynamics and trophic interactions, affect the structure and functional relationships of ecosystems (Chapin et al. 2000; Thebault & Loreau 2003; Hooper et al. 2005). Empirical and theoretical work has focused on functional processes within a single trophic level, such as primary productivity, community stability or nutrient utilization (Tilman et al. 1997; Hector et al. 1999; Spehn et al. 2005). The effects of plant diversity on adjacent trophic organisms, the primary consumers, which in turn directly affect secondary production, are poorly understood (Siemann et al. 1998; Scherber et al. 2006).

Herbivores have strong impacts on ecosystem processes by mediating energy transfer and nutrient cycling, and indirectly altering plant biomass, plant species composition and diversity (McNaughton, Banyikwa & McNaughton 1997; Belovsky & Slade 2000; Bakker et al. 2006). The effect of changing plant diversity on herbivore performance is therefore a key issue. Studies of the responses of small herbivores to changes in plant diversity have shown that declining plant diversity reduces the diversity of herbivorous insects (Siemann et al. 1998; Knops et al. 1999), increases the abundance of specialist insects (Andow 1990; Koricheva et al. 2000; Wilsey & Polley 2002), and decreases performance (biomass gain, survival rate and reproduction) (Giulio & Edwards 2003; Pfisterer, Diemer & Schmid 2003). Other experiments have shown that the diversity and identity of plant functional groups influence herbivore abundance and performance (Haddad et al. 2001; Scherber et al. 2006; Specht et al. 2008). The ‘resource concentration’ and the ‘enemies’ hypotheses have been proposed to explain the response of small herbivores to plant diversity change (Root 1973; Otway, Hector & Lawton 2005).

Surprisingly, little is known about the mechanisms linking plant species richness to performance of large generalist herbivores even though they control many grassland ecosystem processes and impact on the provision of ecosystem goods and services, and hence the wellbeing of humans (Owen-Smith 1988; Gordon, Hester & Festa-Bianchet 2004). Large generalist herbivores can tolerate low plant nutrient content but require greater plant abundance and energy-rich foods to maximize performance. Small herbivores, on the other hand, may be more likely to select nitrogen-rich foods and exhibit some degree of host plant specificity (Stephens & Krebs 1986; Olff, Ritchie & Prins 2002), and so the functional mechanisms are likely to differ.

Foraging is a complex process, and large generalist herbivores make decisions on which plants to consume, with the outcomes influencing their nutrient acquisition. Optimal foraging theory predicts that diet selection by large generalist herbivores is greatly influenced by trade-offs between the benefits of ingesting a given diet component and the costs of selecting for it (Stephens & Krebs 1986). Large generalist herbivores select diets among alternative plant species to better meet their need for nutrients (Simpson et al. 2004). Often, nutrient requirements cannot be met from a single plant species and so switching amongst the types of plants occurs during foraging (Westoby 1978; Wiggins, McArthur & Davies 2006). Clearly, dietary diversity is important for herbivores possibly as a consequence of nutrient balance and detoxification limitation. The nutrient balance hypothesis proposes that dietary diversity allows individuals to balance intake of different nutrients (Westoby 1978). In contrast, the detoxification limitation hypothesis predicts that the amount of food that a herbivore can safely ingest depends on the rate at which the herbivores can detoxify any plant secondary metabolites (PSMs) that the food contains (Freeland & Janzen 1974). It follows that a herbivore should be able to eat more if it selects multiple foods due to the constraints imposed by concentrations of PSMs in plants. A varied diet may also be a consequence of transient food aversions hence voluntary food intake may increase if foods of different nutritional value or different flavours are available (Provenza et al. 1996, 2003). However there is no strong empirical evidence for high plant diversity in pastures enhancing voluntary intake of herbivores.

As plant species richness increases, more frequent switching may be required to sample and assess the quality of the forage material by post-ingestive analysis (Provenza 1995). High plant species richness in grassland communities should therefore pose both opportunities and challenges to large generalist herbivores. On the one hand, herbivores have more opportunities to choose the palatable plants and optimize nutrient intake when plant species richness is higher, whereas a greater variety of plant species may make it more difficult to determine which foods best meet nutritional needs, because there are more choices and greater complexity (Duncan & Young 2002; Ginane et al. 2005; Provenza & Villalba 2006; Wang et al. 2010). The major objectives of this study were to examine the effects of plant species richness on voluntary food intake, nutrient acquisition and the quality of the overall diet selected by sheep, and to elucidate the mechanisms linking plant species richness with the foraging of large herbivores.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Two experiments were conducted; one in a semi-controlled indoor cafeteria and the other in the field. The cafeteria experiment enabled elucidation of mechanisms that are difficult to study satisfactorily in the field.

Indoor cafeteria experiment

We used nine 2-year-old male sheep (35·4 ± 1·8 kg) typical of the abundant sheep throughout north-east China. In the weeks before the experiment they had been grazing native grassland dominated by the grasses Leymus chinensis (Trin.) Tzvel. and Phragmites australis (Clav.) Trin. (Gao et al. 2008). The plant species used in the cafeteria experiment are native to and common in the meadow steppes of north-east China. They comprised 13 species from three plant functional groups (grasses: L. chinensis, P. australis, Chloris virgata Sw., Hemarthria sibirica (Gand.) Ohwi, Calamagrostis epigejos (L.) Roth., and Echinochloa crusgalli (L.) Beauv.; legumes: Lathyrus quinquenervius (Miq.) Litv. and Vicia amoena Fisch.; forbs: Kalimeris integrifolia Turcz., Artemisia scoparia Waldstem et Kitailael, Kochia sieversiana (Pall.) C. A. M., Apocynum venetum L. and Suaeda glauca Bunge).

The experiment was carried out in July/August 2005. Sheep were housed individually in 4·2 × 3·2 m pens and were acclimatized to the surroundings and feeding patterns 2 weeks before the experiment commenced. Over the 14-day pre-trial period each sheep was offered freshly cut L. chinensis foliage, and 300 g of a concentrate (components: corn 74%, soybean meal 20%, CaHPO4 1·7%, CaCO3 1·7%, NaCl 2%, premix 0·6%) each day. After the adjustment period, there was no supplementation with concentrates. Foliage from individual species was collected from the adjacent grassland every 2 days and placed in a cool room at 10 °C. It was cut into 10 cm lengths before feeding. Any foliage stored for more than 2 days was discarded.

During the experiment, twice-daily meals (at 07:30 and 14:30 hours) were presented in containers arranged adjacent to one another (Fig. 1), with one plant species per container. The containers were circular (46 cm diameter and 18 cm high) and each contained more than enough plant material for the 2-h meal. Typically, 30% or more of the plant material in each container remained. The weight of material of each plant species was recorded before and after each meal and a correction was made for water loss, based on water content measurements. Intake was calculated and expressed as g dry matter (DM)/day for each sheep. The sheep always had free access to clean drinking water.


Figure 1.  Experimental layout in a sheep pen for indoor cafeteria trial. Only one sheep was housed in a pen.

Download figure to PowerPoint

There were six levels of plant species richness – 1, 2, 4, 6, 8 or 11 species (Table 1). Within each level, there were three species combinations (termed Comb-a, -b and -c) based on plant functional type (grass, legume or forb, respectively). The plant functional group richness generally increased from Comb-a to Comb-b to Comb-c. For example, at two species richness level, the plant functional group richness was 1, 2 and 2 for Comb-a, Comb-b and Comb-c respectively, and at four species richness level, the plant functional group richness was 2, 2 and 3 for Comb-a, Comb-b and Comb-c respectively. Commonly, there were relatively large differences in nutrient levels and palatability amongst the plant functional types. Thus, the three basic combinations based on plant functional groups provided contrasting diets. For example, the nutritional value of L. chinensis and P. australis were similar but there were large differences between L. chinensis, L. quinquenervius and K. integrifolia. The order in which species were added was created by using a nested species loss approach (Zavaleta & Hulvey 2004); the most abundant and dominant plant species from each of the three functional groups in the grassland were kept in all treatments and relatively rare species were added to raise the richness levels. Each of the nine sheep was given each plant richness level offered for eight consecutive days.

Table 1.   Plant species offered to sheep in three species combinations and six levels of species richness. The values are average daily intake (g DM) of foliage of each species and total daily intake in the indoor cafeteria experiment. ‘–’ means that the species was not offered. For each species, numbers with the same superscript are not significantly different at < 0·05
  1. Lc, Leymus chinensis; Pa, Phragmites australis; Ki, Kalimeris integrifolia; Lq, Lathyrus quinquenervius; As, Artemisia scoparia; Ks, Kochia sieversiana; Av, Apocynum venetum; Cv, Chloris virgata; Va, Vicia amoena; Sg, Suaeda glauca; Hs, Hemarthria sibirica; Ce, Calamagrostis epigejos; Ec, Echinochloa crusgalli.

 1665ab 665c
 2162a634ab 796c
 455 b737a27482a1148b
 8 9b387b18120224b394147 84ab1428a
 1112b408b24243ab184 34b18711016762581508a
 1555a 555d
 2558a345b 903c
 8 60c304113e19434367a143b1011316a
 11 60c391159de4099b 3910110216515361313a
 1489 489d
 2399a495 894c
 4120b504380a140ab1142b –
 6 26b125b480377a166a139a1332ab
 8 12 b188ab16142132467a127b 39b1446a
 11 34b293a24442174b 28b20413912343131337ab

About 500 g of material of each plant species was sampled daily and combined. The bulked samples were oven-dried, ground and analysed for crude protein (CP), neutral detergent fibre (NDF), acid detergent fibre (ADF) and metabolism energy (ME). CP (N × 6·25) was determined with an automatic Kjel-Foss apparatus (2300; Foss Tecator AB, Höganäs, Sweden). NDF and ADF were analysed with an automatic fibertec apparatus (M6, Foss Tecator AB, Höganäs, Sweden) by the method of Van Soest (1976). Invitro Dry Matter Digestibility (IDMD) was determined after digestion with the 2 n HCl-cellulase-pepsin technique. Gross energy (GE) was determined with an automatic bomb calorimeter (Parr 1281; Parr Instruments, Moline, IL, USA). ME was calculated from IDMD and GE values with the formula ME (MJ kg DM−1) = GE × IVDMD × 0·815 as accepted by the British Agricultural Research Council (ARC 1965).

Field experiment

The field experiment was conducted at the Songnen Grassland Ecological Research Station (44°45′N; 123°45′E), Northeast Normal University, Changling County, China. Three plant species richness levels (1, 4–6 or >8 species), each within six replicate fenced plots (about 200 m2 each), were selected within a spatially variable grassland. Plots comprising just one species contained either L. chinensis, P. australis or C. virgata. Plots comprising 4–6 species contained L. chinensis, P. australis, K. integrifolia, A. scoparia, Puccinellia tenuiflora (Turcz.) Scribn. Et. Merr.), Thalictrum simplex L., K. sieversiana or L. quinquenervius. Plots with >8 species contained three additional species, Arundinella hirta (Thunb.) C. Tanaka, Inula japonica Thumb., and Taraxacum sinicum Kitag. All plots were hand weeded to achieve the species arrays.

Between 15 July and 20 August 2005, five adult male sheep (60·7 ± 5·5 kg) grazed for 1 day in each plot from 07:00 to 09:00 hours and from 15:30 to 17:30 hours. Sheep were fed supplementary corn (400 g per sheep) indoors each evening. This supplementary feeding is common practice in the region. Plant species intake was estimated by six observers recording the number of foraging bites taken by three sheep on each plant species (two observers per sheep). Intake mass was calculated from number of bites taken by each sheep multiplied by the bite size (the average dry matter taken in one bite).

Bite size for each of the four major plant species was estimated separately after the field experiment. An experimental enclosure of 25 m2 contained patches of L. chinensis, P. australis, C. virgata or K. integrifolia. On four separate occasions, three sheep foraged freely in the enclosure, each one in a patch where the same plant species dominated. Sheep were closely observed by three people and the number of bites was recorded (one observer per sheep). Subsequently, the shoots of grazed plants were harvested, dried and the total dry matter weight determined, and the number of the individual plants in each composite sample was counted. Shoots of similar sized ungrazed plants of each species immediately outside the enclosure were harvested, dried and weighed to determine the average weight of individual shoots. From the measurements, the mean bite size was calculated:

  • image

where BS is bite size, UN is the number of randomly selected ungrazed shoots, UW is ungrazed shoots dry weight for UN, GW is the weight of all grazed shoots, GN is the number of grazed shoots and BN is bite number.

Three replicates, based on three separate days, per sheep and per plant species, were averaged to estimate bite sizes for each of the four species. The bite size for other minor species was estimated from an average of the four species.

Statistical analyses

All statistical analyses were performed with the sas 6·12 statistical package (SAS Institute Inc., Cary, North Carolina, USA). Assumptions of normality and heteroscedasticity were tested. Two-way analyses of variance model (two-way anova) was used for all analyses. This anova model contained plant species richness as a fixed effect and individual sheep as blocks. The statistical difference between levels was determined by Duncan’s tests. The significance level was set at  0·05. For the indoor cafeteria experiment, we analysed the average daily dry matter intake, daily nutrient intake, dietary quality and daily intake of each plant species over 8 days. For the field experiment, the average daily dry matter intake from six replicate fenced plots of a species diversity level was used in the anova.

The effects of plant species richness on the above variables, except for daily intake of single plant species, were analysed in two ways. First, the three combinations (Comb-a, -b, -c) were analysed separately. Secondly, all combinations were analysed separately. Daily intake of single plant species were only analysed in three combinations (Comb-a, -b, -c) separately.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

In the cafeteria experiment, the sheep always ate foliage of all the plant species in the meals offered (Table 1). The voluntary intake of each of the 13 plant species varied widely; the most palatable species were P. australis, K. sieversiana, and L. quinquenervius; the intermediate palatable species were K. integrifolia, A. venetum, H. sibirica, V. amoena and S. glauca; and the least palatable species were L. chinensis, C. virgata, C. epigejos, E. crusgalli and A. scoparia. As the number of species was increased, the voluntary intake of almost all species significantly decreased (Table 1). For the non-palatable species, the addition of further new species reduced the voluntary intake to a greater extent than for the intermediate or most palatable species. In addition, the voluntary intake of any species was strongly influenced by the accompanying species, and varied to different degrees due to the manipulation of species richness levels.

Irrespective of the proportions of each species selected from the presented diets, the voluntary daily intake significantly increased in all three combinations with increasing plant species richness (F5,10 = 57·90, P <0·0001; F5,10 = 94·47, P <0·0001; F5,10 = 34·20, P <0·0001, respectively for Comb-a, -b, and -c). When the sheep were offered only one plant species from each of the functional groups, the sheep consumed the least amounts. When species from another functional group were added, the voluntary intake significantly increased in Comb-b and Comb-c. Addition of L. chinensis to the P. australis based meal (Comb-a) however did not significantly increase the voluntary intake (Table 1). Overall there was a 1·4-fold increase in daily intake with the increase from one to 11 species in the meals (F5,40 = 108·27, P <0·0001; Fig. 2a). The increase was highest for Comb-c (1·45-fold, 1·37-fold, 1·27-fold respectively in Comb-c, -b and -a). Voluntary daily intake did not increase with addition of species beyond eight species. In the field, there was also a significant increase in the forage consumed by the sheep when they were given access to plant communities with a higher species richness (P <0·05; Fig. 2a′).


Figure 2.  Effects of plant species richness on the diet selected by sheep. (a) The average daily dry matter (DM) intake (a′, the average daily DM intake in the field grazing experiment). (b) Energy/protein ratio. (c) The average daily metabolizable energy intake. (d) The average daily protein intake. Points are the means for nine sheep measured over 8 days and with three species combinations within each level. Different letters indicate points significantly different from each other (< 0·05).

Download figure to PowerPoint

The daily energy and protein intakes in the cafeteria experiment significantly increased with increasing plant species richness (F5,40 = 78·70, P <0·0001 for ME intake; F5,40 = 57·12, P <0·0001 for CP intake; Fig. 2c,d). However, the ME intake did not increase above eight species (Fig. 2c) and the CP intake did not increase above four species (Fig. 2d). In the single plant species diet, the ME intake and CP intake did not meet the maintenance requirements of the sheep. Overall, nutrient intake by the sheep was highest when eight species were available. In a separate analysis for three species combinations (Comb-a, -b and -c), there were similar trends for ME intake and CP intake compared to the overall analyses (Table 2). The ME/CP ratio of the eight plant species richness level was significantly higher than for the 4, 6, and 11 species richness levels; there were no significant differences amongst 1, 2, 4, 6, and 11 species richness levels (Fig. 2b).

Table 2.   Mean and anovas for energy and protein, and other measures of dietary quality for the three combinations (Comb-a, Comb-b and Comb-c) of plant species, at different plant richness levels, in the indoor cafeteria experiment. With each combination, all numbers with the same superscript are not significantly different at < 0·05
VariableSpecies combinationPlant species richness levelsF(5, 10)P
1 species2 species4 species6 species8 species11 species
Protein intake (g day−1)Comb-a77·2d87·4d128·1c156·9ab152·4b168·4a69·530·0001
Comb-c62·7d89·7c186·1a170·3ab159·9b157·6b 8·340·0005
Energy intake (MJ day−1)Comb-a 5·13c 5·91c  9·69b 11·62ab 13·26a 13·10a30·210·0001
Comb-b 5·00e 7·17d  9·31c 10·74b 12·20a 10·93b23·990·0002
Comb-c 4·94d 7·35c 11·48b 11·90b 14·03a 11·58b16·100·0007
CP (%)Comb-a10·6710·13 10·22 11·46  9·91 10·34 0·270·7742
Comb-b10·47 9·27  9·99 10·28  9·77 10·06 2·370·1417
Comb-c11·76b 9·51e 14·91a 11·81b 10·16d 10·87c 4·310·0423
ME (MJ kg−1)Comb-a7·09c 6·84c  7·73b  8·48a  8·50a  7·94b 6·160·0180
Comb-b8·20 7·18  7·26  7·85  8·56  7·70 0·540·5998
Comb-c9·26 7·80  9·20  8·25  8·92  7·95 2·320·1488
NDF (%)Comb-a70·4a70·9a 64·5b 60·2c 59·6c 63·3b 7·480·0103
Comb-b51·264·8 62·7 56·5 59·3 64·6 0·530·6065
Comb-c52·3dc62·0a 50·2d 54·1c 57·0b 61·7a 3·990·0531
ADF (%)Comb-a41·942·1 40·9 36·2 39·2 41·5 3·040·0930
Comb-b42·4a42·9a 40·3c 36·1e 38·8d 41·3b 4·070·0509
Comb-c39·841·5 33·6 35·3 38·4 41·0 0·160·8520

Overall, plant species richness significantly affected dietary quality, i.e. the nutrient content of the sheep diet (Fig. 3). The two positive indexes for dietary quality (ME and CP) were lowest at the 2 and 11 plant species richness levels. The two negative indexes for dietary quality (NDF and ADF) were also highest at the 2 and 11 species levels. Thus, the trends in various nutrients (ME, CP, NDF and ADF) were the same, namely, the best quality diet was at intermediate levels of plant species richness, apart from the single species level (Fig. 3a–d). In a separate statistical analysis of the three functional groups (Comb-a, -b and -c) there were no consistent trends for the CP, ME, NDF and ADF contents of the overall diet (Table 2).


Figure 3.  Effects of plant species richness on the quality of diet selected by sheep. The nutrient values in terms of metabolism energy (ME) (a), crude protein (CP) (b), neutral detergent fibre (NDF) (c), acid detergent fibre (ADF) (d) (where average daily nutrient intake for each component is divided by average daily dry matter intake) of the diets selected. Points are the means for nine sheep measured for three plant species combinations within each richness level. On the left of the vertical line, sheep cannot exhibit selectivity for plants because only one species is available; on the right, the nutrient quality of the diet selected is dependant on the plant species richness of the diet on offer.

Download figure to PowerPoint


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Studies of small generalist herbivores have shown that increasing plant species richness may enhance herbivore growth (Pfisterer, Diemer & Schmid 2003; Unsicker et al. 2008). For large herbivores, this relationship is difficult to study experimentally because they have long life span and require large amounts of food. Energy and protein are critical components for animal growth and were used here to predict herbivore performance (Simpson et al. 2004). In our study, increasing plant species richness significantly enhanced nutrient intake and we would expect any change in these to affect the weight gains of the sheep. Overall, the increase in nutrient intake with increasing species richness available in forage reached a plateau at eight species; the nutrient intake of sheep was not highest at the highest plant richness level (11 species). A study by Pfisterer, Diemer & Schmid (2003) on individual growth of a generalist insect herbivore (a grasshopper) demonstrated that the benefits of plant species richness came from the higher quality of plants available. In contrast, our study demonstrated that the benefits of a higher level of plant species richness for a large herbivore (sheep) mainly came from a greater amount of the forage eaten and not a higher quality of foliage.

Although previous studies have shown that generalist herbivores perform better when eating a mixture of plant species (Pennings, Nadeau & Paul 1993; Bernays et al. 1994; Burritt & Provenza 2000; Marsh et al. 2006), this is the first study to show that herbivores will consume greater quantities when grazing a plant community of higher species richness. The few studies on the effects of diet mixing on intake by generalist herbivores only compared the extremes; single species with mixed species (Burritt & Provenza 2000; Marsh et al. 2006). It is possible that herbivores will eat less of a single species diet than a mixed species diet because of monotonous flavour and nutrient or toxin limitation (Provenza 1996; Provenza et al. 2003). Thus, an important finding of this study was the steady rise in voluntary intake as the number of species offered to the sheep increased. In the field experiment, similar results were obtained but there was no evidence of a significant difference in voluntary intake between 4 and 6 species richness levels and >8 species. This may be a consequence of the provision of supplementary corn in the field, which may have increased rumen fill, thereby weakening the effects of increasing plant diversity.

The higher energy and protein intakes of the sheep as plant species richness increases may be attributed to two mechanisms: selection of palatable species with higher energy and protein content and/or the complementary effects of a diverse diet. Our study suggests both mechanisms operate together (see Table 1). Sheep can obtain great benefit from high plant species richness because there is a greater chance that palatable food will occur in species-rich rather than in species–poor plant combinations. Also, our study showed that sheep always preferred a diverse diet even over diets of one highly palatable species. Specifically, when two or more palatable species were presented to the sheep they ate less of the original palatable species, but when highly palatable species were being eaten, the addition of less palatable species did not reduce the amount they ate of the highly palatable original species; as the sheep ate more forage so the energy and protein intakes rose. We have also demonstrated that the voluntary intake of any one species depended on the other species present. For example, L. chinensis (grass) was the least palatable species. When it was combined with P. australis (grass), intake of L. chinensis was small, but intake increased when it was combined with the different functional type of L. quinquenervius (legume) or K. integrifolia (forb). Different plant functional types usually differ considerably in nutritional characteristics, which may provide a complementary nutritional advantage to the herbivore, thereby stimulating food intake. It appears that both selecting effects of palatable species and complementary effects are involved in the foraging behaviour of the sheep.

Sheep cannot obtain a high nutrient intake in all the palatable species combinations (e.g. legume). For example, although Comb-c at 1 and 2 species contained high quality L. quinquenervius, the nutrient intake of sheep was still low due to insufficient food intake resulting from a less diverse diet (Tables 1 and 2). Only when sheep have access to higher plant species richness can the presence of highly palatable species enable sheep to achieve a higher nutrient intake. Furthermore, our data indicate that plant functional group richness was a major determinant of nutrient intakes. The nutrient intake of sheep was much higher in those species combinations with high functional group richness (e.g. a combination of legumes, grasses and forbs). In the absence of high quality species, the nutrient intakes rose if there was more than one plant functional group available (e.g. a combination of L. chinensis and K. integrifolia). Therefore, high nutrient intakes depended not only on a palatable species being available, but also on there being a species-rich meal (i.e. a diverse diet). The sheep were unable to eat sufficient of a single plant species meal to satisfy their energy and protein requirements and must select foods with nutrients that complement each other (Westoby 1978). Herbivores may acquire aversions to food containing high levels of secondary metabolites, and excessive or deficient nutrients, even when the food is nutritionally adequate (Provenza 1995, 1996). The transient food aversions are attributed to satiety which is mediated by post-ingestive feedback from flavours, nutrients and secondary metabolites interacting along concentration gradients (Provenza et al. 2003; Provenza & Villalba 2006). Thus, a varied diet can improve a herbivore’s ability to meet its nutritional needs.

Large generalist herbivores exploit the heterogeneity of plant resources through selective foraging, choosing a diet of better quality than the average vegetation on offer (Jamieson & Hodgson 1979). Although dietary quality has been recognized as important, variation in voluntary food intake is considered to be more important for animal performance (Yearsley, Tolkamp & Illius 2001). It has been suggested that high plant species richness may be a means of changing plant species preference and potentially stimulating intake (Provenza 1996; Early & Provenza 1998). In our experiments, the quality of the overall diet selected by the sheep was also significantly affected by plant species richness but the variability was small (Fig. 3, Table 2). Therefore, it appears that the higher plant species richness stimulated the sheep to eat more leading to higher energy and protein intakes.

The biological value and ecological causes of mixed diets have been mainly attributed to nutritional balance (Westoby 1978), detoxification limitations (Freeland & Janzen 1974) and transient food aversion (Provenza 1996; Early & Provenza 1998; Provenza et al. 2003). Our study shows that the energy/protein ratio for the diet of sheep at the eight species level reached the ratio recognized by ARC (1965) as being balanced. At this level the voluntary intake by the sheep was highest, demonstrating that a species-rich diet contributes to nutrient balance, thereby improving voluntary intake. The energy/protein ratio did not differ among 1, 2, 4, 6 and 11 species levels (Fig. 2b), but there was still a significant increase in food intake (Fig. 2a). So the increase in voluntary intake was not only due to the nutritional balance of the diverse diet; other factors could have been involved. Physiological and behavioural experiments have shown that herbivores feeding on plants containing different classes of PSMs have multiple detoxification pathways inherently available and can therefore presumably consume more food (Marsh et al. 2006; Wiggins, McArthur & Davies 2006). In our study the plants A. venetum, A. scoparia, K. sleverslana, S. glauca, L. quinquenervius, and V. amoena all contain high but different PSMs. Thus, the increase in food intake among 1, 2, 4, 6 and 11 species levels may be attributed to detoxification limitations of the diverse diets. Although we cannot directly examine the effects of toxin dilution of different PSMs from varied diets, we concluded that nutrient balance, toxin dilution or taste modulation could have functioned jointly to improve food intake in our experiments.

A large generalist herbivore could conceivably raise energy and protein intakes by eating more forage, even if the forage is of low quality. The strong asymptotic relationship between plant species richness and voluntary food intake (Fig. 2) suggests that the maintenance of plant species richness in rangeland plant communities is critically beneficial for domestic animal production as well as the conservation of biodiversity.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank G. Chen, X. Sun, J. Liu, C. Lin, L. Ba and B. Zhang for technical assistance and D. Tilman, J. Wu and B. Liu for helpful comments on early drafts of this manuscript. This project was supported by the State Key Basic Research Program of China (2007CB106801), and the National Natural Science Foundation of China (No. 30571318, 30600427, 30590382).


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Andow, D.A. (1990) Population dynamics of an insect herbivore in simple and diverse habitats. Ecology, 71, 10061017.
  • ARC. (1965) Requirements for Energy. ARC, Nutrient Requirements of Farm Livestock, No. 2. Ruminants, pp. 193257. ARC, London.
  • Bakker, E., Ritchie, M., Olff, H., Milchunas, D. & Knops, J. (2006) Herbivore impact on grassland plant diversity depends on habitat productivity and herbivore size. Ecology Letters, 9, 780788.
  • Belovsky, G.E. & Slade, J.B. (2000) Insect herbivory accelerates nutrient cycling and increases plant production. Proceedings of the National Academy of Sciences, USA, 97, 1441214417.
  • Bernays, E.A., Bright, K.L., Gonzalez, N. & Angel, J. (1994) Dietary mixing in a generalist herbivore: tests of two hypotheses. Ecology, 75, 19972006.
  • Burritt, E.A. & Provenza, F.D. (2000) Role of toxins in intake of varied diets by sheep. Journal of Chemical Ecology, 26, 19912005.
  • Chapin, F.S. III, Zavaleta, E.S., Eviner, V.T., Naylor, R.L., Vitousek, P.M. & Reynolds, H.L. (2000) Consequences of changing biodiversity. Nature, 405, 234242.
  • Duncan, A.J. & Young, S.A. (2002) Can goats learn about foods through conditioned food aversions and preferences when multiple food options are simultaneously available? Journal of Animal Science, 80, 20912098.
  • Early, D.M. & Provenza, F.D. (1998) Food flavor and nutritional characteristics alter dynamics of food preference in lambs. Journal of Animal Science, 76, 728734.
  • Freeland, W. & Janzen, D. (1974) Strategies in herbivory by mammals: the role of plant secondary compounds. The American Naturalist, 108, 269289.
  • Gao, Y., Wang, D., Ba, L., Bai, Y. & Liu, B. (2008) Interactions between herbivory and resource availability on grazing tolerance of Leymus chinensis. Environmental and Experimental Botany, 63, 113122.
  • Ginane, C., Duncan, A.J., Young, S.A., Elston, D.A. & Gordon, I.J. (2005) Herbivore diet selection in response to simulated variation in nutrient rewards and plant secondary compounds. Animal Behaviour, 69, 541550.
  • Giulio, M.D. & Edwards, P.J. (2003) The influence of host plant diversity and food quality on larval survival of plant feeding heteropteran bugs. Ecological Entomology, 28, 5157.
  • Gordon, I.J., Hester, A.J. & Festa-Bianchet, M. (2004) The management of wild large herbivores to meet economic, conservation and environmental objectives. Journal of Applied Ecology, 41, 10211031.
  • Haddad, N.M., Tilman, D., Haarstad, J., Ritchie, M. & Knops, J.M.H. (2001) Contrasting effects of plant richness and composition on insect communities: a field experiment. The American Naturalist, 158, 1735.
  • Hector, A., Schmid, B., Beierkuhnlein, C., Caldeira, M.C., Diemer, M., Dimitrakopoulos, P.G., Finn, J.A., Freitas, H., Giller, P.S., Good, J., Harris, R., Högberg, P., Huss-Danell, K., Joshi, J., Jumpponen, A., Körner, C., Leadley, P.W., Loreau, M., Minns, A., Mulder, C.P.H., O’Donovan, G., Otway, S.J., Pereira, J.S., Prinz, A., Read, D.J., Scherer-Lorenzen, M., Schulze, E.D., Siamantziouras, A.S.D., Spehn, E.M., Terry, A.C., Troumbis, A.Y., Woodward, F.I., Yachi, S. & Lawton, J.H. (1999) Plant diversity and productivity experiments in European grasslands. Science, 286, 11231127.
  • Hooper, D.U., Chapin, F.S. III, Ewel, J.J., Hector, A., Inchausti, P., Lavorel, S., Lawton, J.H., Lodge, D.M., Loreau, M., Naeem, S., Schmid, B., Setala, H., Symstad, A.J., Vandermeer, J. & Wardle, D.A. (2005) Effects of biodiversity on ecosystem functioning: a consensus of current knowledge. Ecological Monographs, 75, 335.
  • Jamieson, W.S. & Hodgson, J. (1979) The effects of variation in sward characteristics upon the ingestive behaviour and herbage intake of calves and lambs under a continuous stocking management. Grass and Forage Science, 34, 273282.
  • Knops, J.M.H., Tilman, D., Haddad, N.M., Naeem, S., Mitchell, C.E., Haarstad, J., Ritchie, M.E., Howe, K.M., Reich, P.B., Siemann, E. & Groth, J. (1999) Effects of plant species richness on invasion dynamics, disease outbreaks, insect abundances and diversity. Ecology Letters, 2, 286293.
  • Koricheva, J., Mulder, C.P.H., Schmid, B., Joshi, J. & Huss-Danell, K. (2000) Numerical responses of different trophic groups of invertebrates to manipulations of plant diversity in grasslands. Oecologia, 125, 271282.
  • Marsh, K.J., Wallis, I.R., McLean, S., Sorensen, J.S. & Foley, W.J. (2006) Conflicting demands on detoxification pathways influence how common brushtail possums choose their diets. Ecology, 87, 21032112.
  • McNaughton, S.J., Banyikwa, F.F. & McNaughton, M.M. (1997) Promotion of the cycling of diet-enhancing nutrients by African grazers. Science, 278, 17981800.
  • Olff, H., Ritchie, M.E. & Prins, H.H.T. (2002) Global environmental controls of diversity in large herbivores. Nature, 415, 901904.
  • Otway, S.J., Hector, A. & Lawton, J.H. (2005) Resource dilution effects on specialist insect herbivores in a grassland biodiversity experiment. Journal of Animal Ecology, 74, 234240.
  • Owen-Smith, R. (1988) Megaherbivores: The Influence of Very Large Body Size on Ecology. Cambridge University Press, Cambridge.
  • Pennings, S.C., Nadeau, M.T. & Paul, V.J. (1993) Selectivity and growth of the generalist herbivore (Dolabella auricularia) feeding upon complementary resources. Ecology, 74, 879890.
  • Pfisterer, A.B., Diemer, M. & Schmid, B. (2003) Dietary shift and lowered biomass gain of a generalist herbivore in species-poor experimental plant communities. Oecologia, 135, 234241.
  • Provenza, F.D. (1995) Postingestive feedback as an elementary determinant of food preference and intake in ruminants. Journal of Range Mangement, 48, 217.
  • Provenza, F.D. (1996) Acquired aversions as the basis for varied diets of ruminants foraging on rangelands. Journal of Animal Science, 74, 20102020.
  • Provenza, F.D. & Villalba, J.J. (2006) Foraging in domestic vertebrates: linking the internal and external milieu. Feeding in Domestic Vertebrates: From Structure to Function (ed. V.L.Bels), pp. 210240. CABI Publications, Oxfordshire, UK.
  • Provenza, F.D., Scott, C.B., Phy, T.S. & Lynch, J.J. (1996) Preference of sheep for foods varying in flavors and nutrients. Journal of Animal Science, 74, 23552361.
  • Provenza, F.D., Villalba, J.J., Dziba, L.E., Atwood, S.B. & Banner, R.E. (2003) Linking herbivore experience, varied diets, and plant biochemical diversity. Small Ruminant Research, 49, 257274.
  • Root, R. (1973) Organization of a plant-arthropod association in simple and diverse habitats: the fauna of collards (I. Brassica oleracea). Ecological Monographs, 43, 95124.
  • Scherber, C., Mwangi, P.N., Temperton, V.M., Roscher, C., Schumacher, J., Schmid, B. & Weisser, W.W. (2006) Effects of plant diversity on invertebrate herbivory in experimental grassland. Oecologia, 147, 489500.
  • Siemann, E., Tilman, D., Haarstad, J. & Ritchie, M. (1998) Experimental tests of the dependence of arthropod diversity on plant diversity. The American Naturalist, 152, 738750.
  • Simpson, S.J., Sibly, R.M., Lee, K.P., Behmer, S.T. & Raubenheimer, D. (2004) Optimal foraging when regulating intake of multiple nutrients. Animal Behaviour, 68, 12991311.
  • Specht, J., Scherber, C., Unsicker, S.B., Köhler, G. & Weisser, W.W. (2008) Diversity and beyond: plant functional identity determines herbivore performance. Journal of Animal Ecology, 77, 10471055.
  • Spehn, E.M., Hector, A., Joshi, J., Scherer-Lorenzen, M., Schmid, B., Bazeley-White, E., Beierkuhnlein, C., Caldeira, M.C., Diemer, M., Dimitrakopoulos, P.G., Finn, J.A., Freitas, H., Giller, P.S., Good, J., Harris, R., Högberg, P., Huss-Danell, K., Jumpponen, A., Koricheva, J., Leadley, P.W., Loreau, M., Minns, A., Mulder, C.P.H., O’Donovan, G., Otway, S.J., Palmborg, C., Pereira, J.S., Pfisterer, A.B., Prinz, A., Read, D.J., Schulze, E.D., Siamantziouras, A.S.D., Terry, A.C., Troumbis, A.Y., Woodward, F.I., Yachi, S. & Lawton, J.H. (2005) Ecosystem effects of biodiversity manipulations in European grasslands. Ecological Monographs, 75, 3763.
  • Stephens, D. & Krebs, J. (1986) Foraging Theory. Princeton University Press, Princeton, New Jersey, USA.
  • Stirling, G. & Wilsey, B. (2001) Emperical relationships between species richness, eveness, and proportional diversity. The American Naturalist, 158, 286299.
  • Thebault, E. & Loreau, M. (2003) Food-web constraints on biodiversity-ecosystem functioning relationships. Proceedings of the National Academy of Sciences, 100, 1494914954.
  • Tilman, D., Knops, J., Wedin, D., Reich, P., Ritchie, M. & Siemann, E. (1997) The influence of functional diversity and composition on ecosystem processes. Science, 277, 13001302.
  • Unsicker, S.B., Oswald, A., Köhler, G. & Weisser, W.W. (2008) Complementarity effects through dietary mixing enhance the performance of a generalist insect herbivore. Oecologia, 156, 313324.
  • Van Soest, P.J. (1976) Influence of environment on forage composition. Proceedings of Workshop on Nutritional Evaluation of Forages (ed. V.D.Mugdal), pp. 80. Nail. Dairy Res. Inst, Karnal, India.
  • Wang, L., Wang, D., Bai, Y., Jiang, G., Liu, J., Huang, Y. & Li, Y. (2010) Spatial distributions of multiple plant species affect herbivore foraging selectivity. Oikos, 119, 401408.
  • Westoby, M. (1978) What are the biological bases of varied diets? The American Naturalist, 112, 627631.
  • Wiggins, N.L., McArthur, C. & Davies, N.W. (2006) Diet switching in a generalist mammalian folivore: fundamental to maximising intake. Oecologia, 147, 650657.
  • Wilsey, B.J. & Polley, H.W. (2002) Reductions in grassland species evenness increase dicot seedling invasion and spittle bug infestation. Ecology Letters, 5, 676684.
  • Wilsey, B.J., Chalcraft, D.R., Bowles, C.H. & Willig, M.R. (2005) Relationships amongst indices suggest that richness is an incomplete surrogate for grassland biodiversity. Ecology, 86, 11781184.
  • Yearsley, J., Tolkamp, B.J. & Illius, A.W. (2001) Theoretical developments in the study and prediction of food intake. The Proceedings of the Nutrition Society, 60, 145156.
  • Zavaleta, E.S. & Hulvey, K.B. (2004) Realistic species losses disproportionately reduce grassland resistance to biological invaders. Science, 306, 11751177.