Ecological correlates of seed survival after ingestion by ... - BES journal

Functional Ecology 2005 19, 284–290

Ecological correlates of seed survival after ingestion by Fallow Deer Blackwell Publishing, Ltd.

A. MAARTEN MOUISSIE,† CARMEN E. J. VAN DER VEEN, G. F. (CISKA) VEEN and RUDY VAN DIGGELEN Community and Conservation Ecology Group, University of Groningen, PO Box 14, 9750 AA Haren, the Netherlands

Summary 1. The survival and retention of seeds was studied by feeding known quantities of seeds of 25 species to four captive Fallow Deer (Dama dama L.). To test for ecological correlates, plant species were selected to represent large variation in seed size, seed shape, seed longevity and habitat fertility. 2. Seeds of 24 out of 25 fed plant species survived ingestion and defecation by Fallow Deer. Seed survival ranged between 0·5 and 42% of germinable seeds fed. Time to recover 50% of all seeds defecated by Fallow Deer in faeces averaged 25 h, and ranged from 13 to 38 h. 3. Seed survival was negatively related to seed mass (R = 0·65) and variance of unit seed dimensions (R = − 0·56), and positively related to seed longevity (R = 0·40), but not related to habitat fertility. Log10 of (seed mass × variance of seed dimensions) was the best predictor of seed survival (R = − 0·68). 4. The ecological correlates of seed survival presented here can help us to estimate the ability of plant species to disperse seeds over long distances. Key-words: Ellenberg indicator value, endozoochory, foliage-is-the-fruit hypothesis, herbivores, seed dispersal Functional Ecology (2005) 19, 284–290 doi: 10.1111/j.1365-2435.2005.00955.x

Introduction The ability of plant species to track rapid climate change, to persist in highly fragmented landscapes, and to reach habitat restoration sites depends critically on long-distance seed dispersal (Clark 1998; Bakker & Berendse 1999; Cain, Milligan, & Strand 2000). Some plant species have lightly plumed seeds, burrs or fleshy fruits that enable long-distance dispersal, but most species lack such diaspores and may depend on means of dispersal for which they are not specifically adapted (Higgins 2003). For example, the ant-dispersed woodland herb Trillium grandiflorum is occasionally dispersed beyond 3 km by White-tailed Deer (Vellend et al. 2003). For many plant species, large herbivores are likely candidates to serve as long-distance seed-dispersal vectors owing to their large home range sizes and long periods of seed retention inside their gut (Pakeman 2001). The ‘foliage-is-the-fruit’ hypothesis suggests that, for a large number of small-seeded herbaceous species, a normal and selected-for mode of dispersal is through ingestion of seeds as large herbivores consume their foliage (Janzen 1984).

© 2005 British Ecological Society

†Author to whom correspondence should be addressed. E-mail: [email protected]

A crucial phase in the dispersal of plant seeds by large herbivores is seed survival through the molar mills, digestive fluids and bacteria after ingestion. Seeds adapted for endozoochory should be small, round and hard. In addition, foliage edibility and composition should attract large herbivores to ingest seeds, as for example fruits are attractive to frugivores (Janzen 1984). Due to a lower carbon : nutrient ratio, plant species associated with fertile soils are more palatable than species associated with less fertile soils (Bryant, Chapin, & Klein 1983; Hobbie 1992; Iason & Hester 1993). Hence if palatability and seed survival are coevolved traits there should also be a relationship between habitat fertility and seed survival. Surveys of the seed contents of herbivore dung do show that small seeds, species with a high seed longevity index, and species associated with fertile soils are relatively frequently dispersed via dung (Pakeman, Digneffe, & Small 2002; Couvreur et al. 2004; Mouissie 2004). However, these surveys do not provide a causal explanation for this phenomenon. Does it result from higher seed survival, higher seed ingestion, or perhaps simply from higher seed production? Seed-feeding trials conducted so far do not provide an answer, because they have been directed at specific species or species from a specific system, e.g. six tropical pasture species (Simao Neto, Jones, & Ratcliff 1987); grass and legume 284

285 Ecological correlates of seed survival

seeds (Gardener, Mcivor, & Jansen 1993); legume seeds from a Mediterranean grassland (Russi, Cocks, & Roberts 1992). In the present paper we address this issue by feeding captive Fallow Deer (Dama dama) known seed quantities of 25 plant species, and recovering seedlings from their dung. Deer are among the most abundant group of wild large herbivores worldwide, and viable seeds have been recovered from pellets of several deer species (Malo & Suarez 1995; Heinken et al. 2002; Pakeman et al. 2002; Mouissie 2004; Myers, Vellend, & Gardescu 2004) Therefore deer are potentially important for longdistance dispersal of plants. We chose Fallow Deer as a model animal because they are intermediate-sized ruminants (males 70 kg, females 40 kg); are classified as intermediate selective foragers (Hofmann 1985); and are frequently kept in captivity. The selected plant species represent a range in seed mass of almost three orders of magnitude, include spherical and elongated seeds, range from transient to long-term persistent seeds, and include species associated with the full possible range in soil fertility. Most of the species selected grow in the natural habitat of Fallow Deer, and some species are of conservation interest. We test how seed survival relates to seed size, seed shape, seed longevity, seed retention and soil fertility. Such ecological correlates of seed survival may give an insight in the evolution of endozoochory, and can help to estimate the ability of plant species for long-distance dispersal.

Materials and methods

© 2005 British Ecological Society, Functional Ecology, 19, 284–290

The experiment was conducted at the deer enclosure (450 m2) ‘De Bosrand’ in Lieren, the Netherlands. Five Fallow Deer (D. dama) inhabited this enclosure: one adult male, two adult females and two yearling males. The adult male refused to be hand-fed and could not be used in the experiment, but its dung served as a control for seed ingestion. The deer grazed on short grass covering the enclosure. In addition, the animals were fed daily with commercial deer fodder. The yearlings ate the same as the adults. The selected 25 plant species were native to the European flora, and represented a large range in seed mass, seed shape, seed longevity and habitat fertility. The seed mass of each species was obtained by weighing six samples of 100 seeds. Seed dimensions (length, width and thickness), excluding appendages, were obtained from the BIOLFLOR database (Klotz, Kühn, & Durka 2003). Seed shape was quantified as the variance of unit seed dimensions, calculated by dividing seed length, seed width and seed thickness by seed length and taking the population variance of these three values (following Thompson, Band, & Hodgson 1993a). Round seeds have a low variance of unit seed dimension, whereas elongated seeds have a high variance of unit seed dimension. As an indicator of habitat fertility, Ellenberg nitrogen-indicator values (Ellenberg

1974; Ellenberg et al. 1992) were used. Ellenberg nitrogenindicator values (0 –9) are designed to estimate the relationship between vascular plant species and the availability of nutrients in the soil. Plant species with low Ellenberg nitrogen-indicator values are associated with nutrient-poor soils, whereas species with high values are associated with nutrient-rich soils. Although these estimates are derived from the distribution of species in Central Europe, they are shown to be valid in a larger area including the Netherlands and the British Isles (Thompson et al. 1993b; Schaffers & Sykora 2000). Ellenberg indicator values are known for many plant species in the European flora and need not to be calculated. As indicator of seed longevity in the soil seed bank, the seed longevity index (0–1) was used (Bekker et al. 1998). A low seed longevity index indicates transient seeds, whereas a high index indicates long-term persistent seeds. Most seeds were obtained from commercial sources, but seven species were collected in the field. Each of the four individual deer ingested 1000 seeds of most plant species. However, some species were fed in lower numbers, either because not enough seeds were available, or because the seeds were too large to feed in such quantity (Table 1). Seeds were encapsulated in bread balls, which were readily accepted by the animals. Feeding of the encapsulated seeds was divided over three days: 4 November, 18 November and 2 December 2002. The feeding scheme was a random block design, with animals and feeding days as blocks. The design was biased such that, on a single feeding day, a plant species was not fed to both adults or both yearlings. This enabled us to identify the animal to which a species was fed, as droppings of the adults were larger than droppings of the yearlings. Prior to feeding all droppings were removed from the enclosure. Over 4 days following the feeding of seeds, all droppings were collected and stored separately for each animal. The collection of dung could not be continued after daylight hours, so seeds excreted during the night were collected the next morning. Droppings were stored for 4 weeks at 5 °C for stratification of the seeds, concentrated using 2 mm and 212 µm mesh sieves stacked on top of each other, and placed in containers with layers of potting soil, sterilized potting soil and sterilized white sand, according to a standard seed-bank analysis technique (Ter Heerdt et al. 1996). Droppings from each animal and each collection period (morning, afternoon and night for each of the sampling days) were placed in separate containers. Three containers with droppings of the adult male deer served as a control for contamination of seeds ingested at the enclosure (‘seed ingestion control’). Three containers without droppings, but with the same soil layers, served as controls to test for contamination of outside seeds in the glasshouse (‘outside seed control’). Three samples of 100 seeds of each plant species, sown in similar containers, served as a test of the fraction

286 A. M. Mouissie et al.

Table 1. Plant species fed to Fallow Deer and ecological correlates tested: variance of seed dimensions; seed mass; Ellenberg nitrogen-indicator value (1–9); and seed longevity index (0 –1)

Plant species

Seeds feder animal

Variance of seed dimensions

Seed mass (mg)

Ellenberg nitrogenindicator value

Seed longevity index

Achillea millefolium Arctium lappa Calluna vulgaris Centaurea jacea Centaurea scabiosa Cerastium fontanum Erica tetralix Festuca rubra Geum urbanum Hieracium umbellatum Holcus lanatus Juncus effusus Leucanthemum vulgare Oenothera erythrosepala Plantago lanceolata Plantago major Poa trivialis Ranunculus acris Rumex obtusifolius Sanguisorba minor Silene latifolia Succisa pratensis Tragopogon pratensis Trifolium repens Vaccinium vitis-idaea

1000 300 1000 1000 200 300 1000 1000 650 500 1000 1000 1000 1000 500 1000 1000 1000 500 500 1000 200 300 1000 500

0·12 0·11 0·07 0·08 0·11 0·02 0·04 0·15 0·17 0·14 0·10 0·06 0·11 0·02 0·10 0·09 0·11 0·09 0·04 0·08 0·02 0·12 0·19 0·02 0·10

0·15 9·51 0·02 2·12 5·27 0·14 0·015 0·85 2·43 0·38 0·34 0·02 0·44 0·60 1·14 0·27 0·25 1·97 1·01 4·11 0·80 1·55 7·60 0·66 0·26

5 9 1 ? 4 5 2 6 ? 2 4 4 3 4 ? 6 9 ? 9 2 7 2 5 6 1

0·11 0·75 0·88 0·13 0·17 0·65 0·41 0·14 0·14 0·56 0·56 0·95 0·44 0·88 0·35 0·79 0·75 0·24 0·67 0·33 1 0·24 0·6 0·4 0·03

?, Ellenberg value unknown.

of germinating seeds prior to ingestion (‘preingestion control’). All containers were placed in a glasshouse and supplied with artificial lighting (Philips 400 W HPI-T bulbs) for 15 h a day, maintaining light conditions between 8 and 10 klux. Automatic daily watering with demineralized water kept the soil moist. The glasshouse was heated at 23 °C during the day and 20 °C at night. We identified and removed seedlings from the containers until no more seedlings emerged during a month (recording period was 9 months).

 


Because seeds were stratified and allowed to germinate for a long enough period, it could be assumed that all viable seeds were recovered. Hence the difference between the fraction of seeds germinating before and after ingestion is the fraction of seeds killed by the Fallow Deer. Therefore ‘seed survival’ was analysed as the ratio between the fraction of seeds germinating after ingestion and the fraction of seeds germinating in the preingestion control. Seed-survival data were log10 transformed and plotted against the independent variables: log10 of seed mass (M, mg); variance of unit seed dimension (Vsd); longevity index; and Ellenberg nitrogen-indicator values (NE). Pearson’s correlation tested how seed survival was related to these independent variables. Similarly, a series of multiple variable models were tested to explain more variance in seed survival. These models

x

y

z

followed the general equation log10 ( M Vsd N E ) , where x, y, z ranged from −10 to 10, with intervals of 0·1. Partial correlations were run to test whether variables were correlated with seed survival directly, or indirectly through interdependence with other variables. To test for ecological correlates of seed retention, the average time to recover 50% of the total seeds excreted in the faeces of Fallow Deer (t0·5) was calculated for each plant species. Whether t0·5 was correlated with seed mass, variance of unit seed dimensions, seed longevity or seed survival was tested using Pearson’s correlation. Statistical tests were carried out using  ver. 11·1 (2001).

Results     In total, 2696 seedlings of 24 species emerged from the droppings. The only species of which no seedlings were recorded in the deer droppings was Arctium lappa. Recovered fraction of seeds fed, of the other species, ranged from 0·001 for Sanguisorba minor to 0·20 for Plantago major. Data on the number of seeds recovered are not shown explicitly, but can be calculated from the number of seeds ingested (Table 1), preingestion germinating fraction and seed survival (see Fig. 3). The deer excreted the first seeds within 5 h and the last seeds after more than 90 h (Fig. 1). However, the last collected droppings (90 h since ingestion) contributed


Table 2. Pearson’s correlations between log10 seed survival, log10 seed mass (M, mg), variance of seed dimensions (Vsd), seed longevity index, Ellenberg nitrogen-indicator value, and time to recover 50% of total seeds excreted in faeces of Fallow Deer (t0·5)

Fig. 1. Fraction of total seed recovery (germinable seeds excreted after x hours/all germinable seeds recovered) in time (h) for the slowest passing species (Plantago major); the fastest passing species (Cerastium fontanum); and the average of all species (25) fed to Fallow Deer.

very little to the total seed recovery (> 0·05) (Table 2).

© 2005 British Fig. 2. Retention of seeds after ingestion by Fallow Deer. Species sorted in ascending Ecological Society, order of seed retention. Bars, mean t0·5 in four animals; error bars, SEM. Results for Arctium lappa, Festuca rubra, Geum urbanum, Sanguisorba minor, Succisa pratensis and Functional Ecology, Tragopogon 19, 284–290 pratensis are not shown because SE was equal to or greater than the mean.

Parameter

log10 seed survival

t0·5

log10 M Vsd log10(M × Vsd) Seed longevity index Ellenberg nitrogenindicator value t0·5

−0·65** −0·56** (−0·28) −0·68** (−0·18) 0·40* (0·36) 0·05 (0·49)

0·20 −0·11 0·14 −0·21 0·64**

0·32 (0·74**)

Numbers in brackets refer to partial correlations controlling for log10 M. **P < 0·01 (two-tailed); *P < 0·05 (two-tailed).

Some seedlings (mostly Poa annua) were recorded in the seed ingestion control. However, none of these seedlings was a species used in the experimental feeding. In the controls for contamination of outside seeds, no species used in the experiment emerged. Germinating fraction in the preingestion control was higher than 0·3 for most species, ranging between 0·05 for Calluna vulgaris and 0·87 for Achillea millefolium (Fig. 3).

  Seed survival of Juncus effusus was highest (0·41), followed by Erica tetralix (0·27), P. major (0·24), C. vulgaris (0·15) and Oenothera erythrosepala (0·14). The frugivorous species Vaccinium vitis-idea was ninth in the ranking of best surviving seeds. Half the species fed to the Fallow Deer had a low seed survival ( 0·05). None of the tested models combining seed mass and the seed longevity index correlated better to seed survival than seed mass alone. Seed survival was not significantly related to Ellenberg nitrogen-indicator values or t0·5 (Table 2).

Discussion Almost all species fed to the Fallow Deer germinated from the dung. Surveys of seed density in dung usually


Fig. 3. Seed survival (fraction of seeds germinating after ingestion/fraction of seeds germinating prior to ingestion) of 25 species fed to Fallow Deer. Bars represent averages of feeding trials to four individual deer and three preingestion germination tests. Error bars, SEM. The fraction of seeds germinating before ingestion is shown in brackets.


contain a large proportion of the plant species present in the study site (Welch 1985; Malo & Suarez 1995; Mouissie 2004), including fleshy-fruited and dry-fruited species, and species with and without adaptation for other dispersal vectors. These observations suggest that a strict classification of species dispersed (or not) via the digestive tract of herbivores is not feasible. Instead, we should consider the ability of plants to disperse via large herbivores as a continuous variable. Passage of seed through mouth and alimentary tract is a crucial phase in the dispersal of seeds by large herbivores. Our results show that seed characteristics can help to predict whether a plant species has high or low probability of surviving ingestion and defecation by Fallow Deer. Seed mass appears to be the best predictor of seed survival after ingestion by Fallow Deer, explaining more variability in the data than seed shape or seed longevity. Small-seeded species survive better than large-seeded species. This relationship appears to explain the high frequency of small-seeded species in the dung of large herbivores. Recent surveys show that plant species dispersed via the dung of rabbits, sheep (Pakeman et al. 2002), cattle, ponies (Mouissie 2004), donkeys (Couvreur et al. 2004) and horses (Cosyns & Hoffmann 2005) are, on average, smaller-seeded than species not dispersed via herbivore dung. Lower survival of large seeds is potentially caused by digestive fluids and bacteria, but more probably by initial mastication and rumination. The ratio of surface area to volume decreases with seed size, leading to a decreased impact of acids and digestive fluids in larger seeds. In contrast, the probability of being hit by molar teeth should increase with seed size for mechanistic reasons. A survey of the digestive tract of shot deer (Mouissie 2004) and seed-feeding trials

with cattle, sheep and goats (Simao Neto et al. 1987; Shayo & Uden 1998) also suggest that most seed damage is caused by chewing. Janzen (1984) suggests that seeds adapted for endozoochory are not only small, but also round. Seed survival in Fallow Deer appears to be negatively related to variance of seed dimensions, although when corrected for seed mass this relationship is no longer significant. Furthermore, seed mass and shape combined explain little more variability than seed mass alone. In surveys the appearance of viable seeds in herbivore dung also appears to have little relation to seed shape (Pakeman et al. 2002; Mouissie 2004; Cosyns & Hoffmann 2005). Our results do not show a relationship between soil fertility and seed survival after ingestion by Fallow Deer. Hence the relatively high frequency of species associated with fertile soils observed in the dung of large herbivores (Mouissie 2004) is probably caused not by high levels of seed survival, but by high levels of seed ingestion. Species associated with nutrient-rich soils are, in general, more palatable than species associated with nutrient-poor soils (Bryant et al. 1983; Hobbie 1992; Iason & Hester 1993). Higher seed ingestion is therefore not unlikely. In contrast, there is a positive correlation between soil fertility and t0·5, suggesting that plant species associated with nutrient-rich soils could be adapted to long-distance dispersal. But then, why do these seeds not survive better in the digestive tract? The relationship between endozoochory and soil fertility is clearly not unequivocal, and should be studied further. Seed dispersal by large herbivores has probably played an important role in the survival and distribution of plant species. Large herbivores may explain rapid plant migration between glacial periods, which cannot be explained by other dispersal agents (Cain, Damman, & Muir 1998; Pakeman 2001). Even incidental longdistance dispersal events can increase migration rates by an order of magnitude (Higgins & Richardson 1999). White-tailed Deer, which are a similar size to Fallow Deer, disperse 1% of excreted seeds beyond 1 km (Vellend et al. 2003). Model simulations suggest that movement of Fallow Deer during the period of seed retention (t50 = 25 h on average) gives similar dispersal distances (Mouissie 2004). Plants may not need specific adaptation to endozoochory to get an incidental hitch via a large herbivore. Seeds may happen to be small and durable due to other selective pressures such as disturbance, escape from predation, or persistence in the soil. Our results show that seed survival in Fallow Deer is positively related to seed longevity, suggesting that plant species adapted to build up persistent seed banks may quite serendipitously also be well suited for dispersal by large herbivores. Seed retention in Fallow Deer is short in comparison with seed retention observed in other animals. Time to recover 50% of the total seeds excreted in the faeces of cattle, sheep and goats is 63, 62 and 48 h, respectively (Simao Neto et al. 1987). Large, hard, dormant seeds


of the Guanacaste tree Enterolobium cyclocarpum are retained for more than 10 days in horses and for half that time in cattle (Janzen 1982). Although 83% of variability in seed retention of tropical grass and legume seeds after placement in the rumen of cattle is accounted for by specific gravity, the proportion of hard seeds, and seed size (Gardener et al. 1993), we found no relationship between seed retention and seed size, seed shape, seed longevity or seed survival. The ecological correlates of seed survival presented here can help us estimate the ability of plants to disperse via Fallow Deer. The general applicability of these ecological correlates in other animal species should be tested in future experiments. We expect that our results will also apply to seed dispersal via other intermediatesized ruminants, and probably also in smaller herbivores. These animals have similar digestion rates and chewing intensities, and consequently do similar damage to ingested seeds. Because of less intense chewing behaviour in large ruminants, the correlates may not be found here. Information on the dispersal capacity of plant species is needed to assess the effects of climate change, habitat fragmentation and ecological restoration efforts. We suggest that screening large numbers of species and discovering ecological correlates, rather than focusing on a limited number of species, is the key to these assessments.

Acknowledgements We thank Mrs Hertgens for kindly allowing us to use her Fallow Deer and enclosure for this experiment. We thank Jan P. Bakker, Eric Cosyns and an anonymous referee for helpful comments that have improved this manuscript. The research of A.M.M. is financially supported by the Dutch Organization for Scientific Research (NWO).

References


Bakker, J.P. & Berendse, F. (1999) Constraints in the restoration of ecological diversity in grassland and heathland communities. Trends in Ecology and Evolution 14, 63 – 68. Bekker, R.M., Schaminee, J.H.J., Bakker, J.P. & Thompson, K. (1998) Seed bank characteristics of Dutch plant communities. Acta Botanica Neerlandica 47, 15 –26. Bryant, J.P., Chapin, F.S. & Klein, D.R. (1983) Carbon/ nutrient balance of boreal plants in relation to vertebrate herbivory. Oikos 40, 357–368. Cain, M.L., Damman, H. & Muir, A. (1998) Seed dispersal and the Holocene migration of woodland herbs. Ecological Monographs 68, 325 –347. Cain, M.L., Milligan, B.G. & Strand, A.E. (2000) Longdistance seed dispersal in plant populations. American Journal of Botany 87, 1217–1227. Clark, J.S. (1998) Why trees migrate so fast: confronting theory with dispersal biology and the paleorecord. American Naturalist 152, 204 –224. Cosyns, E. & Hoffmann. M. (2005) Horse dung germinable seed content in relation to plant species abundance, diet composition and seed characteristics. Basic and Applied Ecology 6, 11–24.

Couvreur, M., Cosyns, E., Hermy, M. & Hoffmann. M. (2004) Complementarity of epi- and endozoochory of plant seeds by free ranging donkeys. Ecography 27, 1–12. Ellenberg, H. (1974) Zeigerwerte der Gefäßpflanze Mitteleuropas. Scripta Geobotanica 9, 1–97. Ellenberg, H., Weber, H.E., Düll, R., Wirth, V., Werner, W. & Paulißen, D. (1992) Zeigerwerte der Gefäßpflanze Mitteleuropas. Scripta Geobotanica 18, 1–258. Gardener, C.J., Mcivor, J.G. & Jansen, A. (1993) Passage of legume and grass seeds through the digestive tract of cattle and their survival in faeces. Journal of Applied Ecology 30, 63 –74. Heinken, T., Hanspach, H., Raudnitschka, D. & Schaumann, F. (2002) Dispersal of vascular plants by four species of wild mammals in a deciduous forest in NE Germany. Phytocoenologia 32, 627– 643. Higgins, S.I. (2003) Are long-distance dispersal events in plants usually caused by non-standard means of dispersal? Ecology 84, 1945 –1956. Higgins, S.I. & Richardson, D.M. (1999) Predicting plant migration rates in a changing world: the role of long-distance dispersal. American Naturalist 153, 464 – 475. Hobbie, S.E. (1992) Effects of plant species on nutrient cycling. Trends in Ecology and Evolution 7, 336–339. Hofmann, R.R. (1985) Digestive physiology of the deer – their morphological specialisation and adaptation. Royal Society of New Zealand Bulletin 22, 393 – 407. Iason, G.R. & Hester, A.J. (1993) The response of heather (Calluna vulgaris) to shade and nutrients – predictions of the carbon–nutrient balance hypothesis. Journal of Ecology 81, 75 – 80. Janzen, D.H. (1982) Differential seed survival and passage rates in cows and horses, surrogate Pleistocene dispersal agents. Oikos 38, 150–156. Janzen, D.H. (1984) Dispersal of small seeds by big herbivores: foliage is the fruit. American Naturalist 123, 338–353. Klotz, S., Kühn, I. & Durka, W. (2003) BIOLFLOR – eine Datenbank mit Biologisch–Ökologischen Merkmale zur Flora von Deutschland. Bundesamt für Naturschutz, Bonn, Germany. Malo, J.E. & Suarez, F. (1995) Herbivorous mammals as seed dispersers in a Mediterranean dehesa. Oecologia 104, 246– 255. Mouissie, A.M. (2004) Seed dispersal by large herbivores, implications for the restoration of plant biodiversity. PhD thesis, University of Groningen, the Netherlands. Myers, J.A., Vellend, M. & Gardescu, S.M.P.L. (2004) Seed dispersal by white-tailed deer: implications for long-distance dispersal, invasions, and migration of plants in eastern North America. Oecologia 139, 35 – 44. Pakeman, R.J. (2001) Plant migration rates and seed dispersal mechanisms. Journal of Biogeography 28, 795–800. Pakeman, R.J., Digneffe, G. & Small, J.L. (2002) Ecological correlates of endozoochory by herbivores. Functional Ecology 16, 296 –304. Russi, L., Cocks, P.S. & Roberts, E.H. (1992) The fate of legume seeds eaten by sheep from a Mediterranean grassland. Journal of Applied Ecology 29, 772–778. Schaffers, A.P. & Sykora, K.V. (2000) Reliability of Ellenberg indicator values for moisture, nitrogen and soil reaction: a comparison with field measurements. Journal of Vegetation Science 11, 225 –244. Shayo, C.M. & Uden, P. (1998) Recovery of seed of four African browse shrubs ingested by cattle, sheep and goats and the effect of ingestion, hot water and acid treatment on the viability of the seeds. Tropical Grasslands 32, 195– 200. Simao Neto, M., Jones, R.M. & Ratcliff, D. (1987) Recovery of pasture seed ingested by ruminants. 1. Seed of six tropical pasture species fed to cattle, sheep and goats. Australian Journal of Agriculture 27, 239–246.



Ter Heerdt, G.N.J., Verweij, G.L., Bekker, R.M. & Bakker, J.P. (1996) An improved method for seed-bank analysis: seedling emergence after removing the soil by sieving. Functional Ecology 10, 144 –151. Thompson, K., Band, S.R. & Hodgson, J.G. (1993a) Seed size and shape predict persistence in soil. Functional Ecology 7, 236 –241. Thompson, K., Hodgson, J.G., Grime, J.P., Rorison, I.H., Band, S.R. & Spencer, R.E. (1993b) Ellenberg numbers revisited. Phytocoenologia 23, 277–289.

Vellend, M., Myers, J.A., Gardescu, S. & Marks, P.L. (2003) Dispersal of Trillium seeds by deer: implications for longdistance migration of forest herbs. Ecology 84, 1067–1072. Welch, D. (1985) Studies in the grazing of heather moorland in north-east Scotland. IV. Seed dispersal and plant establishment in dung. Journal of Applied Ecology 22, 461– 472. Received 23 September 2004; revised 1 December 2004; accepted 3 December 2004