SEARCH

SEARCH BY CITATION

Keywords:

  • foraging behaviour;
  • patch choice;
  • plant–herbivore interaction

Summary

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

1. Foraging decisions by herbivores depend on variation in food types, the scale(s) at which this variation occurs and the opportunity and capacity for herbivores to respond to such variation. These decisions affect not only the herbivores themselves, but also the vulnerability of individual plants to being eaten. Associational plant refuges, in which neighbouring plants alter focal plant vulnerability, are an emergent property of foraging decisions.

2. Using the red-bellied pademelon (Thylogale billardierii) as a model generalist mammalian herbivore, we investigated the spatial scale(s) at which animals made foraging decisions and the resultant effect on focal plant vulnerability. In a replicated design, we varied vegetation at the individual plant scale, generating intraspecific differences in Eucalyptus nitens seedlings by altering their nutrient status (high, low). We varied vegetation at the patch scale, in which seedlings were planted, using high- (grass) and low- (herbicided) quality patches. Animals were allowed to choose where they fed and what they ate. Animal behaviour was recorded and intake of seedlings measured.

3. We found that animals made foraging decisions first at the patch scale then at the scale of individual plants; both patch and focal seedling characteristics influenced browsing. Pademelons spent most of their time in high-quality patches, and seedlings were consequently more vulnerable there than in low-quality patches. Pademelons also ate more foliage from high- than from low-nutrient status seedlings. This behaviour concentrated resources, increasing foraging efficiency and making focal plants more vulnerable to browsing.

4. The opportunity and capacity to choose at both plant and patch scales resulted in a pattern of focal plant vulnerability consistent with the repellent-plant hypothesis. This contrasts with our previous study, in which animals were only provided with choice at the plant level and plant vulnerability followed the attractant-decoy hypothesis. These combined results demonstrate that the influence of neighbouring vegetation on consumption of a focal plant depends on the spatial scale of selection and on opportunities (and capacity) for herbivores to choose.


Introduction

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

Foraging theory predicts that herbivores maximize their foraging efficiency by consuming the most profitable food sources while minimizing the associated costs of obtaining them (Charnov 1976; Stephens & Krebs 1986). More specifically, foragers are expected to maximize energy intake by focusing their browsing effort at sites where resources are highly concentrated (the resource-concentration hypothesis; Root 1972). The foraging decisions leading to resource concentration and the maximization of foraging efficiency depend on variation in food types in the landscape, the scale(s) at which this variation occurs, and the opportunity and capacity for herbivores to respond to such variation. Two obvious scales of variation in food types relevant to herbivores occur at the individual plant and vegetation patch levels, although variation and subsequent selection by herbivores can occur at finer and broader levels (Senft et al. 1987; Kotliar & Wiens 1990).

Plants vary enormously in their nutritional quality and their chemical and structural defences, and the quality of plants to herbivores reflects a trade-off between these characteristics (Villalba, Provenza & Bryant 2002). Plant secondary compounds reduce intake either by reducing digestibility (e.g. tannins; Robbins et al. 1987), or by incurring a toxic load (e.g. essential oils; Boyle et al. 2005). Physical defences, such as thorns, spines or ‘tough’ leaves and stems, can increase the time required for food handling and reduce intake (Lucas et al. 2000). Herbivores often choose plants with higher levels of beneficial nutrients such as nitrogen, and lower levels of toxic secondary chemicals (e.g. Cooper, Owen-Smith & Bryant 1988; McArthur et al. 1993).

Vegetation patches arise from variation in quality and quantity of vegetation at a spatial scale larger than among the individual plants (Kotliar & Wiens 1990; Laca & Demment 1991). Herbivores make foraging decisions at the patch scale in response to a range of patch characteristics. These characteristics include overall patch quality (Ball, Danell & Sunesson 2000), size (Wallis de Vries, Laca & Demment 1999; Pietrzykowski et al. 2003), location and spatial distribution (Clarke, Welch & Gordon 1995), risk of predation (Newman & Caraco 1987; Kotler & Blaustein 1995), proximity to other patches (Palmer et al. 2003) and patch intake rate (Distel et al. 1995; Wallis de Vries et al. 1999). Patch selection is clearly an important component in foraging decisions for many generalist herbivores.

The net foraging behaviour of herbivores comprises the interaction between decisions made at a range of spatial scales, including the plant and vegetation patch levels (Newman 2007). Herbivore responses to a spatial scale can be modified by the existence of other spatial scales (Edwards et al. 1994). For example, browsing of trees by roe deer Capreolus capreolus (Bergman, Iason & Hester 2005) and the probability of trees of different species being attacked by wild herbivores (Baraza, Zamora & Hódar 2006) was influenced by patch type. Similarly, in a field survey focal plants of the same species were more heavily browsed by red deer Cervus elaphus in patches with high- rather than low-quality neighbours, and, at the fine scale within a patch, there was differential browsing among plants (Bee et al. 2009). These studies demonstrate that herbivores can, and do, make decisions at multiple scales simultaneously.

Variation in foraging decisions has direct and important consequences for the herbivores themselves, but logically, it also results in variation in vulnerability of individual plants to being eaten. There are a plethora of hypotheses to explain such variation in plant vulnerability. These predict the outcome of browsing when plants are more or less palatable than their neighbours, and collectively can be termed associational hypotheses (key conceptual developments outlined in Tahvanainen & Root 1972; Atsatt & O’Dowd 1976; McNaughton 1978; Hay 1986; Pfister & Hay 1988; Hjältén, Danell & Lundberg 1993; Alm Bergvall et al. 2006). The attractant-decoy hypothesis, for example, predicts lower consumption of a focal plant in high- rather than low-quality patches because the neighbouring within-patch vegetation of the former provides an alternative food source (Tahvanainen & Root 1972; Atsatt & O’Dowd 1976). In contrast, the repellent-plant hypothesis (McNaughton 1978; Pfister & Hay 1988) predicts that a focal plant is more vulnerable to herbivory in high- rather than low-quality patches because herbivores preferentially choose to feed in such patches. This difference in plant vulnerability depends on whether animals make choices within- or between-patches, and can be predicted using optimal patch use theory (Danell, Edenius & Lundberg 1991; Hjältén et al. 1993).

Understanding the decisions of foraging herbivores, and the outcomes for both of them and the plants they consume under different landscapes of choice, has important ecological and evolutionary consequences. These decisions help define how herbivores value landscapes, how particular plants are likely to be affected by this, and ultimately, how these in turn affect the dynamics of animal and plant populations and hence ecological communities.

Here, we use red-bellied pademelons, Thylogale billardierii (Desmarest, 1822), as a model generalist mammalian herbivore, and Eucalyptus nitens (H. Deane & Maiden) Maiden as the focal plant species, to explore the interactions between foraging decisions, two spatial scales of vegetation heterogeneity (variation in individual plants and in vegetation patches) and consequent focal plant vulnerability to being eaten. We have already established that pademelons make foraging decisions at a vast range of spatial scales. They choose at the intraspecific level, both between leaves within plants (Loney et al. 2006) and between plants (O’Reilly-Wapstra, McArthur & Potts 2002); and at the interspecific plant level (McArthur, Goodwin & Turner 2000). They also make foraging decisions between vegetation patches (Pietrzykowski et al. 2003) and habitats (le Mar & McArthur 2005). Patch characteristics that influence their feeding on a focal plant include vegetation palatability, height and abundance (Miller, McArthur & Smethurst 2006, 2007).

In Miller et al. (2007), we demonstrated that in the absence of choice at the patch level, pademelons consume more from high-nutrient E. nitens seedlings in vegetation patches of high rather than low palatability, consistent with the attractant-decoy hypothesis. Low-nutrient seedlings were eaten very little in either patch type. Here, we extend this work using the same study system, by simultaneously manipulating intraspecific focal plant quality and vegetation patch quality available to pademelons. Other studies have investigated the simultaneous effect of variation in focal plant and patch characteristics, but none using intraspecific variation at the focal plant level. This is an important component of ecological and evolutionary variation, occurring from genetic differences (O’Reilly-Wapstra et al. 2002) and environmental influences including soil nutrient status and shading (Rowe & Potter 2000; Close et al. 2003).

The specific questions we ask in this study are:

  • 1
     What foraging decisions do generalist herbivores make when given the opportunity to choose at both plant and patch scales?
  • 2
     What is the resulting vulnerability of a focal plant to being eaten?
  • 3
     Do these foraging decisions and consequent plant vulnerability differ to those when herbivores only have the opportunity to choose at the plant scale?

We then place our results in the context of foraging theory as it relates to generalist herbivores and consider the implications for associational hypotheses on plant vulnerability to herbivory.

Materials and methods

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

Animals and basal diets

Red-bellied pademelons are medium-sized (adults 4–13 kg) macropodid marsupials, now confined to Tasmania, Australia (Rounsevell, Taylor & Hocking 1991). They are most commonly found in dense vegetation adjacent to night-time feeding areas (le Mar & McArthur 2005). They are mixed feeders or generalist browsers; feeding on grasses, forbs and leaves of trees and shrubs (Sanson 1989; Sprent & McArthur 2002).

A feeding trial was conducted with captive pademelons in animal enclosures at the School of Zoology, University of Tasmania, Hobart. The pademelons (six males and one female, mean body weight 8·8 ± 0·9 kg) were part of a captive colony maintained at the university. Each pademelon was housed in an individual open-roofed holding pen (4 m × 8 m) with a shelter at one end for 4 days before the trial for acclimation. During this period they were fed a basal diet of ‘wallaby pellets’ (c. 2·3% nitrogen; Roberts Limited, Hobart), fresh apple, carrot and silverbeet, and were offered eucalypt seedlings to sample. They had all had prior, intermittent exposure to eucalypt seedlings over several years. The animals were provided with enough food to maintain body weight; they were fed daily and weighed regularly to monitor condition. Fresh drinking water was constantly available to all animals.

Focal plant (seedling) characteristics

Two types of E. nitens were prepared, using previously established fertilizer regimes (Close et al. 2003, 2004). The seedlings were obtained from Forestry Tasmania’s tree nursery, Perth, Tasmania. They were grown in Lännen trays (81 cells per tray, each 41 mm wide by 73 mm deep). The seedlings were either starved (low-nutrient status) or fertilized (high-nutrient status) regularly for 3 months before the trial. Previous studies have found fertilized seedlings to be consumed in preference to unfertilized ones, that is, are of higher quality (Close et al. 2003, 2004). High-nutrient seedlings were fertilized thrice a week with Peters Excel® water-soluble fertilizer (NPK 20 : 2·2 : 6·6), receiving c. 2·25 mg each time. All plants were watered for 20 min twice a day. Levels of nitrogen, neutral detergent fibre (NDF), acid detergent fibre (ADF), acid detergent lignin (ADL), total oils, cineole and formyl phloroglucinol compounds (FPCs) were all significantly higher in high- than in low-nutrient seedlings (Appendix S1). Low-nutrient seedlings had significantly more total phenolics and were also much smaller (Appendix S1). This difference in height was the only component of the plant structure that was altered by the fertilizer. The seedlings in both treatments retained the same form of growth (shown in Fig. 1b), with leaves growing directly from the main stem and no branching. Seedling nutrient status, therefore, reflected both plant chemistry and plant height.

image

Figure 1.  (a) The experimental arena and an example of the treatment layout are shown. Rectangles represent vegetation patches: stippling, high-quality patches; no stippling, low-quality patches. Symbols represent seedlings: O, high-nutrient status seedlings; X, low-nutrient status seedlings. The patches were moved for each animal; there was always one replicate of each treatment east and west of the path as well as north and south. Black triangles show the positions of cameras. Solid grey indicates shelter and bedding straw. (b) A high-quality patch with low-nutrient status seedlings. (c) A pademelon within one of the eight patches offered simultaneously in the experimental arena.

Download figure to PowerPoint

Patch characteristics

We defined ‘patch’ as an area of 12 m2, with less contrast within it than between it and the surrounding matrix. Instant turf, a mixture of bluegrass Poa pratensis L. and ryegrass Lolium rigidium Gaudin (‘Easi Green’ Strathayr instant lawn, Richmond), was rolled out to create high-quality patches. It was short (<5 cm; ≤20% of seedling height) and so did not reduce seedling apparency (sensu Feeny 1976). The low-quality patches were sprayed with herbicide (360 g L–1 of glyphosate) to remove most vegetation; a very small amount of native grass was still present. It could be argued that the two patch types provided a different foraging substrate, but the most important differences between the patches were the amount and type of vegetation.

Effect of seedling nutrient status and patch quality on foraging

The experimental arena consisted of eight 4 × 8 m pens, with four pens on either side of a 1-m wide dirt path. A 3 × 4 m area in each pen was set up as a vegetation patch (Fig. 1). Gates between the eight pens were tied open to allow the test animal to move freely among them. Fencing between the pens meant that patches were spatially separated, but all were accessible via the central path. Three water bowls were placed on the path between the pens. Bedding straw and shelter was available in all pens.

There were two seedling qualities and two patch qualities, resulting in a two-factor design with four different treatments. There were two replicates of each treatment. The experimental layout was a randomized block design, with treatments blocked north/south, and east/west (Fig. 1). For the seedling treatment, six holes were dug in each patch, with 1 m spacing. At the start of each testing period, six pots, each containing one E. nitens seedling, of known height and of one nutrient status, were placed in the holes in each patch to give 6 seedlings per patch and 48 seedlings in total. One animal was tested in the arena at a time. Patch and seedling treatments were switched between pens when testing different animals to avoid a pen effect.

Each animal was kept in the experimental arena for 3 days, and was given access to seedlings for c. 22 h each day. The seedlings were removed and scored for browsing damage (see later), and replaced with fresh seedlings each day. The animals were tested sequentially.

Behavioural data

We investigated animal behaviour by filming pademelons in vegetation patches. One camera was positioned above each of the western four pens (one replicate of each treatment; Fig. 1). Cameras were connected to a Panasonic® Video Cassette Recorder (NV-FJ630 Series, Matsushita Electric Industrial Co. Ltd, Osaka, Japan). Video footage was recorded onto EMTEC BASF EQ-300 cassette tapes (Emtec Magnetics, Oxfordshire, UK). Filming began each day as soon as pademelons were allowed access to seedlings and continued until sunset (c. 7 h), as no lighting was available. The amount of time, in minutes, each animal spent within patches was determined. We also determined the amount of time spent feeding on grass and on seedlings.

Browsing of plants and intake of foliage

The seedlings were scored for browsing damage immediately upon removal from patches. Browsing damage was scored as percentage foliage removed on a scale from 0 to 6, where 0 is 0%, 1 is 1–5%, 2 is 6–25%, 3 is 26–50%, 4 is 51–75%, 5 is 76–95% and 6 is 96–100%. To estimate intake from the damage data, a subsample of seedlings (10/seedling nutrient status/animal/night) was measured (height to the nearest 0·5 cm), cut at soil level, dried (55 °C oven for 48 hours) and weighed. The relationship between seedling height and dry matter of foliage was then examined. There were significant regressions (< 0·01) between plant height and foliage dry matter for both treatments (low-nutrient, r2 = 0·33; high-nutrient, r2 = 0·15). Regression equations were used to calculate the dry matter of each plant offered to animals from its starting height. The browsing score for each plant was converted to the midpoint of the percentage of foliage removed (e.g. score of 2 = 15·5%). Consumption was then calculated as the percentage of the dry matter of foliage offered.

Statistical analysis

To estimate plant biomass offered to animals, regressions of plant height vs. plant dry matter were performed in sas (proc reg; SAS Institute Inc. 1990). To allow for the size range of animals used and the different moisture contents of high- and low-nutrient seedlings (and therefore different percentage of dry matter), analysis of consumption (total amount of foliage consumed from all seedlings within a given treatment averaged per day for each pademelon) was based on grams of dry plant matter consumed per kilogram body mass (gDM kgBM−1).

The amount of time, in minutes, that pademelons spent in each patch type was summed over the last 3 hours before sunset [when most (76 ± 8%) feeding took place] and averaged over the 3 days of the trial. Time spent in patches and time spent feeding on grass were then compared between seedling and patch treatments (both fixed factors) with the individual animal included in the model as a random factor (proc mixed; SAS Institute Inc. 2004). Time spent in feeding on seedlings could not be normalized and so was analysed nonparametrically with the four seedling–patch combinations tested as four treatment levels (proc npar1way; SAS Institute Inc. 1989).

The percentage of seedlings browsed, the percentage of foliage consumed (a measure of selection) and the consumption of foliage (gDM kgBM−1) from seedlings for each animal was averaged over the 3 days for analysis. The effects of individual animal (random factor), seedling and patch (both fixed factors), and the interaction between seedling and patch, on these variables were examined (proc mixed; SAS Institute Inc. 2004).

For all statistical tests, residuals were checked for homoscedasticity and normality, and transformations performed where necessary (Zar 1996). Time spent feeding on grass was square-root-transformed; percentage of foliage consumed was arc-sine square-root-transformed. Values are reported as least square means ± SE.

Results

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

Foraging behaviour

Pademelons spent significantly more time in high-quality (15·5 ± 4·0 min) than in low-quality (0·86 ± 4·0 min) patches, irrespective of seedling nutrient status (Table 1a). Within high-quality patches, most of this time (89–92%) was spent feeding on grass, and pademelons spent significantly more time feeding on grass in high- than in low-quality patches (Table 1b and Fig. 2a). Pademelons never spent more than 5% of the total time in patches feeding on seedlings. Although the mean time spent in feeding on high-nutrient seedlings was greater than that on low-nutrient seedlings (Fig. 2b), there was no significant treatment effect (Kruskal–Wallis inline image; = 0·606).

Table 1.   Results of the mixed model analysis testing the effects of time spent in patches, time spent in feeding on grass, percentage of seedlings browsed, percentage of foliage consumed and gDM kgBM−1 of foliage consumed
Dependent variablesFactorsd.f.Z- or F-valueP
  1. Variables in italics are random and show Z-values; others are fixed and show F-values.

  2. n.e., not estimated; DM, dry matter; BM, body mass.

(a) Time in patchesSeedling nutrient status10·720·408
Patch quality18·760·008
Seedling nutrient status × patch quality10·540·472
Individual animal60·630·266
Residual183·00 
(b) Time in feeding on grassSeedling nutrient status10·480·500
Patch quality19·910·006
Seedling nutrient status × patch quality10·570·460
Individual animal60·570·286
Residual183·00 
(c) Seedlings browsed (%)Seedling nutrient status114·860·001
Patch quality118·23<0·001
Seedling nutrient status × patch quality10·500·490
Individual animal60·230·411
Residual183·00 
(d) Foliage consumed (%)Seedling nutrient status11·280·273
Patch quality112·080·003
Seedling nutrient status × patch quality10·370·551
Individual animal6n.e.n.e.
Residual183·46 
(e) Foliage consumed (gDM kgBM−1)Seedling nutrient status114·860·001
Patch quality17·800·012
Seedling nutrient status × patch quality10·640·433
Individual animal6n.e.n.e.
Residual183·46 
image

Figure 2.  Time spent: (a) feeding on grass and (b) feeding on seedlings in the 3 hours before sunset, averaged over the 3 days of the trial, by pademelons in four different patch types. Values are least-squares means (SE). Superscript letters indicate a significant difference at = 0·05. Note the 10-fold difference in the scale of the y-axis between (a) and (b).

Download figure to PowerPoint

Browsing of plants and intake of foliage

Both seedling nutrient status and patch quality had a significant influence on the percentage of seedlings browsed (Table 1c). Pademelons browsed an average of 36 ± 4% of the high-nutrient seedlings, but only 18 ± 4% of the low-nutrient seedlings. Likewise, they browsed 37 ± 4% of the seedlings located in high-quality patches and 17 ± 4% in low-quality patches (Fig. 3a). Pademelons consumed (hence selected) a significantly higher percentage of foliage from seedlings in high-quality (5·5 ± 0·8%) than in low-quality (2·1 ± 0·8%) patches; there was no effect of seedling nutrient status on the percentage of foliage consumed (Table 1d and Fig. 3b). Pademelons ate significantly more foliage from high-nutrient (0·18 ± 0·02 gDM kgBM−1) than from low-nutrient (0·05 ± 0·02 gDM kgBM−1) seedlings, and in high-quality (0·16 ± 0·02 gDM kgBM−1) than in low-quality (0·07 ± 0·02 gDM.kgBM−1) patches (Table 1e and Fig. 3c). There was no significant interaction between seedling nutrient status and patch quality (Table 1e).

image

Figure 3.  Average daily (a) percentage of seedlings browsed, (b) percentage of foliage consumed and (c) foliage consumed (gDM kgBM−1) by pademelons over 3 days, when offered a choice of high- or low-nutrient seedlings in high- or low-quality patches. Values are least-squares means (SE). Superscript letters indicate a significant difference at = 0·05.

Download figure to PowerPoint

Discussion

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

Foraging decisions at multiple spatial scales

Our results support the concept that, when given the opportunity to do so, generalist herbivores can make foraging decisions at multiple scales (Senft et al. 1987; Laca & Ortega 1995) based on a hierarchy of vegetation structure (Kotliar & Wiens 1990). Red-bellied pademelons clearly chose between patches and did not simply forage randomly. They spent more time in high-quality patches and more time feeding in them; and they ate more grass, browsed a greater proportion of seedlings (% of seedlings browsed), consumed a greater proportion of foliage from seedlings (% of foliage consumed) and ate more from seedlings (gDM of foliage consumed) in high- than in low-quality patches. Pademelons also made foraging decisions at the individual plant scale: although they did not consume a greater proportion of foliage from high- than from low-nutrient seedlings, they did browse a greater proportion of the high-nutrient seedlings (i.e. ate from more of them) and they consumed a greater amount of foliage from them (i.e. ate more in absolute terms). The choice of high-quality patches and high-nutrient seedlings were two separate, additive decisions, as reflected in the significant main effects and lack of interactions between patch and seedling quality for any of the aforementioned variables. We suggest that pademelons chose first at the patch scale and then at the individual plant scale, given that they spent most of their time feeding on grass, with only the occasional nibble taken from the seedlings. The initial decision therefore appears to be for high-quality patches. Within patches, animals then consumed more from high- than from low-nutrient status seedlings. The additive effects of plant and patch characteristics on foraging, with pademelons spending most time and consuming most food in high-quality patches with high-nutrient seedlings, is consistent with the resource concentration hypothesis (Root 1972).

The greater consumption of high- than of low-nutrient seedlings was expected and is consistent with other studies using pademelons and E. nitens seedlings. We found the same pattern in a similar experimental system, but without choice between patches (Miller et al. 2007). We also found the same pattern when testing for preferences under controlled conditions with equal and ad libitum availability of high- and low-nutrient seedlings (Close et al. 2003). The extremely low-nitrogen levels in the low-nutrient seedlings along with the higher phenolic levels probably explains this preference, despite the higher concentration of toxic FPCs and oils in the high-nutrient seedlings. FPCs in eucalypts often reduce intake (Lawler et al. 1999; O’Reilly-Wapstra et al. 2002), but not always, as other constituents can over-ride their influence (Close et al. 2003; O’Reilly-Wapstra et al. 2005). High- and low-nutrient seedlings not only differed in their chemistry; the former were taller and had greater foliage biomass, and so we cannot discount that this structural difference contributed to their greater consumption, as seen with red deer (C. elaphus) offered Sitka spruce saplings (Picea sitchensis; Hartley et al. 1997). Nevertheless, we can discount an effect of plant structure via apparency, because the grass was essentially at ground level in both our patch treatments and both seedling types were well above this height. The greater consumption of high- than low-nutrient seedlings also highlights the importance of intraspecific variation in plants in their vulnerability to herbivory, although the fact that the same proportion of foliage was removed from both seedling types adds an interesting twist. It remains to be seen whether this translates into similar or different growth rates and hence the ultimate costs to the plants.

Many studies have explored the effect of intraspecific variation in plants on herbivory (e.g. O’Reilly-Wapstra et al. 2002; Glynn et al. 2004; Prittinen et al. 2006; Lindroth et al. 2007), but not in the context of multiple spatial scales of selection as we have here. The effects of our phenotypic manipulation of seedlings through nutrient application is highly relevant to ecosystems in which soil types vary naturally, where nutrient levels to plants are enhanced by faecal or urinary input from herbivores, and even more broadly where, for example, some plants are shaded and others not. We would anticipate a similar effect of intraspecific genetic variation too, as it also leads to differences in leaf chemistry and structure. The significance of our results, however, lies in demonstrating that the net vulnerability of a focal plant to herbivory will be determined as much by the vegetation among which it occurs, as by its own genotype/phenotype.

Implications of foraging behaviour on focal plant vulnerability

The scale-dependent foraging behaviour of pademelons has clear implications for focal plant vulnerability to being eaten (as distinct from its capacity to recover from browsing). When choosing at the patch scale, pademelons were attracted to high-quality patches, reducing vulnerability of focal plants within low-quality patches. This pattern is consistent with the repellent-plant hypothesis. Such ‘repellency’ is often argued to arise by low-quality vegetation reducing the ability of herbivores to find and/or use the more palatable focal plant (Wahl & Hay 1995). Our results indicate that this general pattern can also occur simply through animals assessing the nutritional value of the patch as a whole, as the limited vegetation in our low-quality patches could not have actually hindered foraging capacity.

Opportunities to choose affect foraging outcomes

For a focal plant to be least vulnerable to herbivory when both the plant and the patch within which it occurs are of low quality, two foraging conditions must be met. First, animals must have the opportunity to choose between patches and second, they must have the capacity to do so. If one of these conditions is not met, then foraging outcomes can alter markedly, and in fact, focal plant vulnerability can be completely reversed. For example, in our system, pademelons prefer grass to E. nitens seedlings (this study, Miller et al. 2007). In the current study they were able to – and did – choose high-quality (grass) patches. Seedlings within such patches were more vulnerable. However, when choice only occurs within patches, for example when an animal cannot move to another patch, or does not choose between patches despite the opportunity, seedlings may actually be less vulnerable in high- than in low-quality patches because the major/most profitable vegetation (e.g. grass) is eaten in preference to seedlings. In such a situation (Miller et al. 2007), we found that seedlings are more vulnerable in low-quality patches, as in this case the seedling is the most profitable vegetation. Thus, the same patch type can act to either increase (repellent-plant hypothesis) or decrease (attractant-decoy hypothesis) focal plant vulnerability depending on the foraging scenario.

Generalist herbivores do not always make choices at the patch level despite the opportunity. Fallow deer (Dama dama) did not choose between high-quality (food pellets with low tannin) and low-quality (high tannin) patches when offered the choice (Alm Bergvall et al. 2006), but did choose within patches, preferring the high-quality (low tannin) buckets. Whether this difference reflects different herbivore species (deer vs. pademelon) or different types of patches (artificial vs. natural plants) is unclear.

Conclusions

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

Foraging decisions have direct short-term effects on food intake and longer-term consequences in terms of nutritional status, reproductive success and therefore fitness of herbivores. But there is also an important link between foraging behaviour and resulting plant vulnerability that is often overlooked. We advocate greater emphasis in understanding the foraging decisions made by herbivores for deriving associational hypotheses and predicting outcomes of plant vulnerability. The conclusion from our results, combined with those of other studies, is that the influence of neighbouring vegetation on consumption of a focal food/plant depends upon the spatial scale(s) of heterogeneity and hence on opportunities (and capacity) for choice by herbivores.

Acknowledgements

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

This study was funded by the Forests and Forest Industry Council of Tasmania, the Holsworth Wildlife Research Fund and the Maxwell Ralph Jacobs Fund. The authors thank Hugh Fitzgerald, Keith Churchill and Haydn Beck for assistance with enclosure preparation and pademelon handling. Simone Janney, Hugh Fitzgerald and Noel Davies conducted chemical analyses. Captive animal trials were run with approval from the University of Tasmania Animal Ethics Committee, Project No. A0008196. Thanks are also due to Julianne O’Reilly-Wapstra and Barrie May for comments on the manuscript. A.M. Miller was supported by an Australian Postgraduate Award, CRC-SPF scholarship and the Dr Joan Woodberry Postgraduate Fellowship.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information
  • Alm Bergvall, U., Rautio, P., Kesti, K., Tuomi, J. & Leimar, O. (2006) Associational effects of plant defences in relation to within- and between-patch food choice by a mammalian herbivore: neighbour contrast susceptibility and defence. Oecologia, 147, 253260.
  • Atsatt, P.R. & O’Dowd, D.J. (1976) Plant defense guilds. Science, 193, 2429.
  • Ball, J.P., Danell, K. & Sunesson, P. (2000) Response of a herbivore community to increased food quality and quantity: an experiment with nitrogen fertilizer in a boreal forest. Journal of Applied Ecology, 37, 247255.
  • Baraza, E., Zamora, R. & Hódar, J.A. (2006) Conditional outcomes in plant–herbivore interactions: neighbours matter. Oikos, 113, 148156.
  • Bee, J.N., Tanentzap, A.J., Lee, W.G., Lavers, R.B., Mark, A.F., Mills, J.A. & Coomes, D.A. (2009) The benefits of being in a bad neighbourhood: plant community composition influences red deer foraging decisions. Oikos, 118, 1824.
  • Bergman, M., Iason, G.R. & Hester, A.J. (2005) Feeding patterns by roe deer and rabbits on pine, willow and birch in relation to spatial arrangement. Oikos, 109, 513520.
  • Boyle, R.R., McLean, S., Brandon, S. & Wiggins, N. (2005) Rapid absorption of dietary 1,8-cineole results in critical blood concentration of cineole and immediate cessation of eating in the common brushtail possum (Trichosurus vulpecula). Journal of Chemical Ecology, 31, 27752790.
  • Charnov, E.L. (1976) Optimal foraging, the marginal value theorem. Theoretical Population Biology, 9, 129136.
  • Clarke, J.L., Welch, D. & Gordon, I.J. (1995) The influence of vegetation pattern on the grazing of heather moorland by red deer and sheep. I. The location of animals on grass/heather mosaics. Journal of Applied Ecology, 32, 166176.
  • Close, D., McArthur, C., Paterson, S., Fitzgerald, H., Walsh, A. & Kincade, T. (2003) Photoinhibition: a link between effects of the environment on eucalypt leaf chemistry and herbivory. Ecology, 84, 29522966.
  • Close, D.C., McArthur, C., Pietrzykowski, E., Fitzgerald, H. & Paterson, S. (2004) Evaluating effects of nursery and post-planting nutrient regimes on leaf chemistry and browsing of eucalypt seedlings in plantations. Forest Ecology and Management, 200, 101112.
  • Cooper, S.M., Owen-Smith, N. & Bryant, J.P. (1988) Foliage acceptability to browsing ruminants in relation to seasonal changes in leaf chemistry of woody plants in a South African savannah. Oecologia, 75, 336342.
  • Danell, K., Edenius, L. & Lundberg, P. (1991) Herbivory and tree stand composition: moose patch use in winter. Ecology, 72, 13501357.
  • Distel, R.A., Laca, E.A., Griggs, T.C. & Demment, M.W. (1995) Patch selection by cattle: maximization of intake rate in horizontally heterogeneous pastures. Applied Animal Behaviour Science, 45, 1121.
  • Edwards, G.R., Newman, J.A., Parsons, A.J. & Krebs, J.R. (1994) Effects of the scale and spatial distribution of the food resource and animal state on diet selection: an example with sheep. Journal of Animal Ecology, 63, 816826.
  • Feeny, P. (1976) Plant apparency and chemical defense. Recent Advances in Phytochemistry, 10, 140.
  • Glynn, C., Ronnberg-Wastljung, A.C., Julkunen-Tiitto, R. & Weih, M. (2004) Willow genotype, but not drought treatment, affects foliar phenolic concentrations and leaf-beetle resistance. Entomologia Experimentalis et Applicata, 113, 114.
  • Hartley, S.E., Iason, G.R., Duncan, A.J. & Hitchcock, D. (1997) Feeding behaviour of Red Deer (Cervus elaphus) offered Sitka Spruce saplings (Picea sitchensis) grown under different light and nutrient regimes. Functional Ecology, 11, 348357.
  • Hay, M.E. (1986) Associational plant defenses and the maintenance of species diversity: turning competitors into accomplices. The American Naturalist, 128, 617641.
  • Hjältén, J., Danell, K. & Lundberg, P. (1993) Herbivore avoidance by association: vole and hare utilization of woody plants. Oikos, 68, 125131.
  • Kotler, B.P. & Blaustein, L. (1995) Titrating food and safety in a heterogeneous environment: when are the risky and safe patches of equal value? Oikos, 74, 251258.
  • Kotliar, N.B. & Wiens, J.A. (1990) Multiple scales of patchiness and patch structure: a hierarchical framework for the study of heterogeneity. Oikos, 59, 253260.
  • Laca, E.A. & Demment, M.W. (1991) Herbivory: the dilemma of foraging in a spatially heterogeneous food environment. Plant Chemical Defences Against Mammalian Herbivory (eds R.T.Palo & C.T.Robbins), pp. 2944. CRC Press, Boca Raton, FL, USA.
  • Laca, E.A. & Ortega, I.M. (1995) Integrating foraging mechanisms across spatial and temporal scales. Rangelands in a sustainable biosphere. Proceedings of the Fifth International Rangeland Congress (ed. N.E.West), pp. 129132. Salt Lake City, Utah.
  • Lawler, I.R., Eschler, B.M., Schliebs, D.M. & Foley, W.J. (1999) Relationship between chemical functional groups on Eucalyptus secondary metabolites and their effectiveness as marsupial antifeedants. Journal of Chemical Ecology, 25, 25612573.
  • Lindroth, R.L., Donaldson, J.R., Stevens, M.T. & Gusse, A.C. (2007) Browse quality in quaking aspen (Populus tremuloides): effects of genotype, nutrients, defoliation, and coppicing. Journal of Chemical Ecology, 33, 10491064.
  • Loney, P.E., McArthur, C., Sanson, G.D., Davies, N.W., Close, D.C. & Jordan, G.J. (2006) How do soil nutrients affect within-plant patterns of herbivory in seedlings of Eucalyptus nitens? Oecologia, 150, 409420.
  • Lucas, P.W., Turner, I.M., Dominy, N.J. & Yamashita, N. (2000) Mechanical defences to herbivory. Annals of Botany, 86, 913920.
  • Le Mar, K. & McArthur, C. (2005) Comparison of habitat selection by two sympatric macropods, Thylogale billardierii and Macropus rufogriseus, in a patchy eucalypt-forestry environment. Austral Ecology, 30, 674683.
  • McArthur, C., Robbins, C.T., Hagerman, A.E. & Hanley, T.A. (1993) Diet selection by a ruminant generalist browser in relation to plant chemistry. Canadian Journal of Zoology, 71, 22362243.
  • McArthur, C., Goodwin, A. & Turner, S. (2000) Preferences, selection and damage to seedlings under changing availability by two marsupial herbivores. Forest Ecology and Management, 139, 157173.
  • McNaughton, S.J. (1978) Serengeti ungulates: feeding selectivity influences the effectiveness of plant defense guilds. Science, 199, 806807.
  • Miller, A.M., McArthur, C. & Smethurst, P.J. (2006) Characteristics of tree seedlings and neighbouring vegetation have an additive influence on browsing by generalist herbivores. Forest Ecology and Management, 228, 197205.
  • Miller, A.M., McArthur, C. & Smethurst, P.J. (2007) Effects of within-patch characteristics on the vulnerability of a plant to herbivory. Oikos, 116, 4152.
  • Newman, J.A. (2007) Herbivory. Foraging: Behaviour and Ecology (eds D.W.Stephens, J.S.Brown & R.C.Ydenberg), pp. 175220. University of Chicago Press, Chicago.
  • Newman, J.A. & Caraco, T. (1987) Foraging, predation hazard and patch use in grey squirrels. Animal Behaviour, 35, 18041813.
  • O’Reilly-Wapstra, J.M., McArthur, C. & Potts, B.M. (2002) Genetic variation in resistance of Eucalyptus globulus to marsupial browsers. Oecologia, 130, 289296.
  • O’Reilly-Wapstra, J.M., Potts, B.M., McArthur, C. & Davies, N.W. (2005) Effects of nutrient variability on the genetic based resistance of Eucalyptus globulus to a mammalian herbivore and on plant defensive chemistry. Oecologia, 142, 597605.
  • Palmer, S.C.F., Hester, A.J., Elston, D.A., Gordon, I.J. & Hartley, S.E. (2003) The perils of having tasty neighbours: grazing impacts of large herbivores at vegetation boundaries. Ecology, 84, 28772890.
  • Pfister, C.A. & Hay, M.E. (1988) Associational plant refuges: convergent patterns in marine and terrestrial communities result from differing mechanisms. Oecologia, 77, 118129.
  • Pietrzykowski, E., McArthur, C., Fitzgerald, H. & Goodwin, A.N. (2003) Influence of patch characteristics on browsing damage of tree seedlings by mammalian herbivores. Journal of Applied Ecology, 40, 458469.
  • Prittinen, K., Pusenius, J., Tahvanainen, J., Rousi, M., Heinonen, J. & Roininen, H. (2006) Herbivory modifies the genetic structure of birch populations. Oikos, 114, 465470.
  • Robbins, C.T., Hanley, T.A., Hagerman, A.E., Hjeljord, O., Baker, D.L., Schwartz, C.C. & Mautz, W.W. (1987) Role of tannins in defending plants against ruminants: reduction in protein availability. Ecology, 68, 98107.
  • Root, R.B. (1972) Organisation of a plant–arthropod association in simple and diverse habitats: the fauna of collards (Brassica oleracea). Ecological Monographs, 43, 95124.
  • Rounsevell, D.E., Taylor, R.J. & Hocking, G.J. (1991) Distribution records of native terrestrial mammals in Tasmania. Wildlife Research, 18, 699717.
  • Rowe, W.J. & Potter, D.A. (2000) Shading effects on susceptibility of Rosa spp. to defoliation by Popillia japonica (Coleoptera: Scarabaeidae). Environmental Entomology, 29, 502508.
  • Sanson, G.D. (1989) Morphological adaptations of teeth to diets and feeding in the Macropodoidea. Kangaroos, Wallabies and Rat-Kangaroos (eds G.Grigg, P.Jarman & I.Hume), pp. 151168. Surrey Beatty & Sons Pty Limited, New South Wales.
  • SAS Institute Inc. (1989) SAS/STAT User’s Guide. SAS Institute, Cary, NC, USA.
  • SAS Institute Inc. (1990) SAS Procedures Guide. SAS Institute, Cary, NC, USA.
  • SAS Institute Inc. (2004) SAS 9.1.2. SAS Institute, Cary, NC, USA.
  • Senft, R.L., Coughenour, M.B., Bailey, D.W., Rittenhouse, L.R., Sala, O.E. & Swift, D.M. (1987) Large herbivore foraging and ecological hierarchies – landscape ecology can enhance traditional foraging theory. BioScience, 37, 789799.
  • Sprent, J.A. & McArthur, C. (2002) Diet and diet selection of two species in the macropodid browser–grazer continuum: do they eat what they ‘should’? Australian Journal of Zoology, 50, 183192.
  • Stephens, D.W. & Krebs, J.R. (1986) Foraging Theory. Princeton University Press, Princeton, New Jersey.
  • Tahvanainen, J.O. & Root, R.B. (1972) The influence of vegetational diversity on the population ecology of a specialized herbivore, Phyllotreta cruciferae (Coleoptera: Chrysomelidae). Oecologia, 10, 321346.
  • Villalba, J.J., Provenza, F.D. & Bryant, J.P. (2002) Consequences of the interaction between nutrients and plant secondary metabolites on herbivore selectivity: benefits or detriments for plants? Oikos, 97, 282292.
  • Wahl, M. & Hay, M.E. (1995) Associational resistance and shared doom: effects of epibiosis on herbivory. Oecologia, 102, 329340.
  • Wallis de Vries, M.F., Laca, E.A. & Demment, M.W. (1999) The importance of scale of patchiness for selectivity in grazing herbivores. Oecologia, 121, 355363.
  • Zar, J.H. (1996) Biostatistical Analysis. Prentice-Hall, Upper Saddle River, New Jersey.

Supporting Information

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

Appendix S1. Chemical characteristics of foliage from Eucalyptus nitens seedlings, by nutrient status, used in feeding trial. Values are arithmetic means (SE). Chemical analyses were performed as in O’Reilly-Wapstra et al. (2005), and levels in high- and low-nutrient seedlings were compared using one-way anovas (proc univariate; SAS Institute Inc. 2004). Superscript letters indicate a significant difference at P = 0·05 for each seedling characteristic.

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
JANE_1589_sm_appendixs1.doc60KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.