Jacobs syndrome

The Ultimate Basis of the Caching Preferences of Rodents, and the Oak-Dispersal Syndrome: Tannins, Insects, and Seed Germination'

SYNOPSIS. We have shown that eastern gray squirrels and other animals consistently prefer to store intact acorns from the red oak group rather than those from the white oak group. We hypothesized that the ultimate advantage to this behavior comes from the dormancy of red oak acorns. Acorns of the white oak group germinate early in the autumn; thus, we hypothesize that avoiding germination is the primary selective advantage to the preference for caching red oak acorns. Here, we test two alternative hypotheses to explain the benefits of this caching preference: 1) storing red oak acorns allows the high tannin concentrations in red oak acorns to decline (making them more palatable), and 2) storing red oak acorns minimizes losses to insects, presuming they are less infested with insects. We also test the effect of germination schedule on squirrel caching preferences directly, by presenting them with dormant red oak acorns, and red oak acorns about to germinate. We find no evidence that tannin concentrations in red oak acorns decline, although tannin levels did decline in our white oak acorns. We found high losses to insect infestations in one white oak species, but a second white oak species lost very little mass to insects. Finally, we found that germination schedule directly affects squirrel caching preferences: red oak acorns that are near germination are treated like white oak acorns. We conclude that the primary advantage to the preference for caching red oak acorns is that they are less perishable, due to their dormancy. We discuss the effects of this preference on the dispersal of red and white oak acorns, and the subsequent effects of differential dispersal on the ecology and evolution of oaks.

INTRODUCTION

Few have stated it as concisely as Howe and Miriti (2000, p. 434): "No question: seed dispersal matters." They argue that despite wide acceptance of this premise, it has been an open question until very recently. More than a generation ago, Janzen (1970, 1971) and Connell (1971) focused the attention of ecologists on seed dispersal as a critical step in the ecology and evolution of plants. They hypothesized that density-dependent mortality near parent plants gave a substantial advantage to seeds dispersed away from the parents, driving the evolution of the myriad of methods employed by plants to disperse their seeds. There was little data to support this hypothesis, but it made compelling, intuitive sense. This hypothesis was immediately persuasive, inspiring a large body of theory, and influencing field studies and textbooks-even though there was very little empirical evidence (Herrera, 1986; Levey and Benkman, 1999; Howe and Miriti, 2000). Evidence has been slow in coming, in part because of the difficulties in determining dispersal patterns from post-dispersal distributions (Schupp and Fuentes, 1995; Nathan and Muller-Landau, 2000). Only recently have comprehensive field studies begun to confirm the Janzen-Connell hypothesis (Harms et al., 2000; Packer and Clay, 2000). It appears that for many species, seed dispersal does matter.

For a large fraction of plant species, it is animals that disperse their seeds (Howe and Smallwood, 1982; Smith and Reichman, 1984; Stiles, 1989; Willson, 1992). The most familiar strategy employed by plant species involves some sort of food reward attached to the seed, such as elaisomes (Culver and Beattie, 1978) or fruits (Stiles, 1980). Here, the seeds-so vital to the plants-are incidental to the animals. They are either discarded or passed through the gut, often intact and dispersed to a suitable place for germination (Wenny and Levey, 1998).

Many nut-bearing trees and some conifers employ a different strategy; the seed itself is the reward for the animal dispersal agent (Vander Wall, 1990), which may sharpen the conflict of interest between the plants and their dispersal agents. Here, dispersal occurs when animals store the seeds in a place suitable for germination (i.e., directed dispersal, sensu Howe and Smallwood, 1982; Wenny, 2000). Successful dispersal depends on the animal failing to recover some fraction of its cache of seeds, either because the animal has died or moved, stored more than it could use, or lost track of its stores (Vander Wall, 1990). Oaks are a typical example of this strategy (Steele and Smallwood, 1994, 2001). A number of animals store acorns in scatterhoards (one seed per cache), with the acorns well spaced apart (e.g., Stapanian and Smith, 1984). Buried acorns are more likely to germinate and survive to seedlings than acorns on the surface (Barnett, 1977). Both the animals and the oaks appear to benefit, providing a sufficient number of acorns are not recovered. However, the interaction between oaks and animals is more complicated. Acorns from the red oak species group are far more likely to be dispersed and cached intact than those from the white oak group (Steele and Smallwood, 1994; Hadj-Chikh et al., 1996; Smallwood et al., 1998). The purpose of this paper is to test alternative hypotheses explaining the ultimate causation (i.e., selective advantage) for animals to preferentially disperse red oak acorns, and to directly test the influence of germination schedule on squirrel caching behavior.

If the differential dispersal of red oak acorns results in more widely dispersed seedlings, then it has many important consequences for the population structure, life history, ecology, evolution, seedling physiology, and even the management of oaks. We discuss some of the evidence for, and implications and questions arising from this behavior at the end of this paper. The subsequent paper in this symposium discusses the proximate causation of this behavior. First, we review what is currently known about the oak-animal system, and present the problems in determining ultimate causation.

The oak-animal system

Oak taxonomy (genus Quercus) remains problematic (Loreto et al., 1998). Estimates of the number of species world-wide range into the hundreds, but most of the 35-40 oak species in deciduous forests of eastern USA can be placed into one of two subgeneric groups (Kaul, 1985; Miller and Lamb, 1985). The red oak group (also called the black oak group; section Erythrobalanus), is typified by Q. rubra, the northern red oak. The white oak group (section Quercus, formerly Lepidobalanus or Leucobalanus), is typified by Q. alba, the white oak. Hereafter, we will refer to these groups as the RO and WO groups, respectively. While the size, shape, and color of an acorn are often helpful in species identification, these characteristics are not correlated with RO or WO species groups (Jensen, 1992). However, other acorn properties differentiate the two species groups (Table 1). RO acorns usually have higher concentrations of fat than WO acorns and higher concentrations of tannins. There are some exceptions to this pattern, especially outside the eastern deciduous forest biome (e.g., in Florida; Fleck and Layne, 1990), but most studies confirm this pattern for most oak species (see references in Table 1). In addition, WO acorns germinate early in the autumn, soon after falling to the ground, while RO acorns remain dormant for the winter (Fox, 1982; Young and Young, 1992). These properties affect the foraging and caching behavior of animals.

The eastern gray squirrel (Sciurus carolinensis) is the most intensively studied oak dispersal agent, partly because of its close association with oaks, and the ease with which it can be observed: it often habituates to people on university and college campuses. Smallwood and Peters (1986) predicted that gray squirrels preferentially store acorns from the RO group, while eating WO acorns immediately. This prediction has been tested and confirmed over a number of years, across a wide geographic range (Smallwood et al., 1998; Hadj-Chikh et al., 1996). At least four other mammal species preferentially store acorns from the RO group (PDS, unpublished data; Ivan and Swihart, 2000; Steele et al., 2001). Smallwood and Peters based this prediction on the hypothesis that WO acorns were more perishable than RO acorns, due to the early germination of WO acorns. Perishability has since been shown to influence the caching behavior of other animals (e.g., the eastern woodrat, Neotoma floridana; Riechman, 1988; Post and Reichman, 1991).

The autumn germination of WO acorns is different from the typical spring germination of dormant seeds. They do not send up a green shoot in the fall; instead, a thick taproot grows down several cm into the soil. WO acorns transport much of the biomass of the acorn into the taproot (Fox, 1982). Since squirrels find this taproot unpalatable (Smith and Follmer, 1972), early germination renders WO acorns less suitable for storage than RO acorns. We hypothesize that the primary explanation for the ultimate advantage to the squirrels for preferentially storing RO acorns is that early germination renders WO acorns more perishable.

Gray squirrels do store WO acorns, albeit far less often than RO acorns. When they do, they usually excise the embryo of the acorn (Fox, 1982). Embryo excision consists of a neat, very small whole cut from the tip of the acorn, where the embryo is located, and prevents germination. Squirrels do not excise the embryos from RO acorns in the fall, except in unusual circumstances-see below. Embryo excision seems obviously related to the early germination schedule of WO acorns. In contrast, there are several hypotheses to explain the ultimate advantage for animals to preferentially store RO acorns. For example, Jacobs (1992) suggested that handling times may be a more important in determining the caching preferences of squirrels. Hadj-- Chikh et al. (1996) tested that hypothesis against our germination-perishability hypothesis, finding strong support for the latter and none for the former.

Other hypotheses concern tannin concentrations and insect infestations. Squirrels may store RO acorns to avoid consuming the high tannin concentration in RO acorns, as storing them in the ground may allow tannin concentrations to decline. Smallwood and Peters (1986; table 6 therein) presented circumstantial evidence that squirrels are not affected by the higher tannin concentrations in RO acorns, but this evidence is contradicted by the direct physiological tests of Chung-MacCoubrey et al. (1997). They find that squirrels are negatively affected by the high concentrations of tannins in RO acorns. Thus, squirrels may prefer to cache RO acorns to allow the tannins to decline-either by abiotic factors (water from the soil leaching out the tannins) or physiological processes in the acorn. Finally, it is also possible that the advantage to storing RO acorns comes from avoiding loss to insects. It is presumed that RO acorns are less infested with weevils, due to their higher levels of tannin (Weckerly et al., 1989a). Under this hypothesis, tannin and germination schedules would merely be correlated with insect infestation rates, but would not themselves be ultimate causal factors. Of course, these hypotheses are not necessarily mutually exclusive.

Below, we test the assumptions of the tannin-avoidance hypothesis and insect-- avoidance hypothesis. We cached RO and WO acorns in the ground, retrieving samples through the winter to test the assumption that tannin concentrations decline during storage, and to compare losses to insect infestations. We also examine the influence of germination schedules on squirrel caching preferences directly, by manipulating the timing of germination in RO acorns.

MATERIALS AND METHODS

The first experiment was designed to test the assumptions of the tannin-avoidance hypothesis and insect-avoidance hypothesis: specifically, that tannins decline during storage, and that cached WO acorns lose more to insect infestation that cached RO acorns. We planted acorns in two garden plots in the fall of 1988. Plot 1 was located in a suburban back yard in west-central Ohio, USA, 10 km south of Dayton OH. The plot had been used to grow vegetables and fruits during the previous several summers, and lay fallow during the winter months. The soil was mostly bare, tilled, and aerated. We used Q. rubra for RO acorns, and Q. alba for WO acorns, both purchased from Sheffield's Seed Company (www.sheffields.com). These acorns had been collected from sources in Ohio and Iowa. On 25 October, we buried 66 RO acorns and 80 WO acorns in a regular array, in a plot of 10 m^sup 2^. Rows were 20 cm apart, and acorns were planted at 20 cm intervals. RO and WO acorns were alternated within and between rows, so that most acorns were surrounded by 4 acorns of the other species. Acorns were buried approximately 2 cm deep to mimic squirrel caching habits (Barnett, 1977). A strip of hardware cloth over the plot prevented pilferage by squirrels and other vertebrates.

To mimic the embryo-excision behavior of squirrels (Fox, 1982), a small knife was used to remove embryos from half (40) of the WO acorns before planting them. The remaining 40 were buried intact. The intent was to test the hypothesis that embryo excision makes WO acorn more susceptible to rot. All RO acorns were buried intact. Every acorn (RO and WO) was examined. Acorns showing exit holes from weevil larvae, discolorations, or apparently underweight were discarded, because squirrels are known to avoid caching infested acorns (Steele et al., 1996).

Plot 2 was planted on 2 November in another suburban back yard, in northeastern Ohio, approximately 20 km south of Akron, OH. All of the acorns used in plot 2 were collected from private grounds in eastern Kansas. We used Q. rubra again as the RO, and Q. macrocarpa (the Bur oak) as the WO species. Q. macrocarpa was chosen because of its unusual properties: it is a WO species, yet produces acorns with high tannin levels. The examination, selection, and spacing of acorns was similar to that of plot 1, but all acorns were buried intact. We cached 60 RO and 60 WO acorns in plot 2.

We retrieved samples of 10 to 30 acorns of each species from each plot at irregular intervals (as weather allowed) from 7 December 1988 to 17 April 1989. Each acorn was cleaned, weighed, and examined for insect larvae exit holes, splits in the shell, and protruding radicles. Each acorn was deshelled, chopped into 1/8 pieces, and visually inspected. One of us (PDS) scored each acorn on a scale of 0-10 for insect damage to the cotyledon (0 = no visible damage, 5 = 50%, 10 = 100% of the cotyledon eaten by insect larvae). Acorns were similarly scored for fungal damage and rot, but this scoring was apparently less reliable (see Discussion). On the day the acorns were buried, an initial sample of acorns was set aside for the same analyses.

After removing the frass (feces remaining in tunnels eaten by the insect larvae), one of us (SHF) assayed the remaining cotyledon for tannin concentrations. Hydrolyzable and condensed tannins were assayed separately. Condensed tannins were measured using a modified acidified vanillin method (Broadhurst and Jones, 1978), while hydrolyzable tannins were analyzed with a modified iodate technique (Bate-- Smith, 1977). See Faeth (1986) for details of the tannin assay procedures. Although these assay procedures may not give an accurate absolute measure of the physiological or ecological effects of tannins (Martin and Martin, 1982; Hanley et al., 1992), they are repeatable. Therefore, they can be used as a relative measure of tannin levels for within-experiment comparisons, and to see if tannin levels change over time (Faeth, 1986).

The second experiment was designed to directly test the effects of impending germination on squirrel caching behavior by manipulating germination schedules of RO acorns. In October of 1995, RO acorns (Q. rubra) were collected from the grounds of the campus of Swarthmore College in Pennsylvania. These acorns were kept in cold storage (0 to 4 deg C) for a year, which delayed their germination. Normally, they would have germinated in the spring of 1996, but by keeping them in continuous cold storage, we delayed germination until the fall of 1996. This allowed us to present RO acorns to squirrels that were at the same stage of germination as normal WO acorns. In November of 1996 these old RO acorns were presented to free ranging gray squirrels in Kirby Park, a 48.5 ha urban park in Wilkes-Barre, northeast Pennsylvania. This park includes open lawns, mature oaks and maples, and a large population of gray squirrels, habituated to humans.

Some of the old RO acorns had radicles protruding from the tip, others did not yet show signs of root emergence. We conducted two sets of trials. In one set, old red oak acorns with protruding radicles were paired with the current year's dormant RO acorns. In a second set, old red oak acorns with no visible root were paired with the current year's RO acorns. Each pair of acorns were presented sequentially to free-- ranging squirrels in Kirby park, alternating the order of presentation, following the methods of Hadj-Chikh et al. (1996). These trials are also used in Steele et al. (2001b) to investigate the proximate cues that influence squirrel caching preferences.

RESULTS

Experiment I. Properties of cached acorns

There were no significant differences or apparent trends between the intact WO acorns and those with their embryos excised, in terms of tannin concentrations or insect damage. There were no significant differences or apparent trends with respect to fungal or bacterial rot between intact and excised WO acorns, but the assessment of rot was much more subjective. We attempted to culture fungi from these acorns, and there was no correlation between the presence of fungi in our cultures and our visual assessment of the presence or absence of fungi in the acorns. Most of the WO acorn cotyledons appeared to be healthy, living tissue, whether the embryo was present or had been removed. In all remaining analyses, intact and embryo-excised WO acorns are pooled.

The RO acorns show no evidence of a decline in tannins-in either hydrolyzable or condensed tannins, in either plot (Fig. 1). In fact, simple regression analyses reveal a small but significant increase in tannins with the number of days spent buried, for hydrolyzable tannins in plot 1, and condensed tannins in plot 2 (Table 2). In contrast, WO acorns show a trend of tannins declining over the winter (Fig. 2). This trend is significant for condensed tannins in Q. alba acorns in plot one, and both hydrolyzable and condensed tannins in Q. macrocarpa acorns in plot 2 (Table 2). The bur oak (Q. macrocarpa) acorns in plot two had high concentrations of tannins, similar to those of RO acorns.

In plot 1, insect damage was higher in WO than in RO acorns (Fig. 3). By the end of the experiment, an average of 25% of the cotyledon of Q. alba acorns had been reduced to frass, while there was no apparent damage in any of the final sample of RO acorns. Few larvae were recovered, but the exit holes and frass-filled tunnels in the cotyledons were characteristic of those made by larvae of curculionid beetles. We compared the amount of insect damage to RO and WO acorns, using a t-test for unequal variances (Sokal and Rohlf, 1981). WO acorns suffered significantly more insect damage (t = 8.77, df = 26, P

Experiment II. Response of squirrels to germinating RO acorns

Squirrels cached significantly more of the new, dormant RO acorns, and ate most of the old, germinating RO acorns (Fig. 4). This was true when the old RO acorn had visible radicles protruding, and when the acorns did not yet have visible roots. In a few of the cases where old RO acorns were cached, we were able to recover those acorns, and found that the embryos had been excised. In other words, germinating RO acorns were treated like WO acorns.

DISCUSSION

Squirrel caching preferences

Squirrels typically excise the embryos of WO acorns before burying them, but usually do not do the same to RO acorns (though squirrels and other animals often consume part of the other end of acorns: Steele et al., 1993). We predicted costs to embryo excision: namely, that acorns with the embryos excised would be more susceptible to rot. We were surprised by the lack of evidence to support this prediction. However, the lack of evidence may be due to the difficulty in assessing the costs. There was no correlation between our visual scoring of the presence of fungal infestation in an acorn, and the success of culturing fungi from that acorn. Mismatches occurred in both directions; i.e., fungi was successfully cultured from acorns with no visible infestation, and vice versa. This uncertainty in assessing the degree of rot leaves us unable to draw firm conclusions as to the costs of embryo excision. However, the costs appear to be small. These results are consistent with those of Steele et al. (2001) who found that WO acorns with excised embryos appeared to resist rot as well as whole WO or RO acorns.

The tannin-avoidance hypothesis states that the advantage to preferentially caching RO acorns comes from avoiding the consumption of tannins. This hypothesis assumes that tannin concentrations in RO acorns decline through the winter. Our results do not support this assumption. If anything, tannin concentrations in RO acorns appear to increase slightly through the winter. Our results are consistent with those of Dixon et al. (1997). They conducted similar experiments in South Dakota under much colder conditions, and used different methods to assay tannin concentations, yet they also found no decline of tannins in RO acorns. Dixon et al. (1997) studied two of the same species we studied: the RO Q. rubra and the WO Q. macrocarpa. Interestingly, they did not see a decline in the tannin concentrations of Q. macrocarpa. While the decline in tannins we observed in WO acorns is interesting, it does not support the tannin-avoidance hypothesis to explain squirrel preference for caching RO acorns. Koenig and Faeth (1998) examined tannin levels through the winter in different RO and WO species from western North America, cached in a different fashion. They also found no evidence of a decline in tannin levels.

One might hypothesize that the benefit from caching RO acorns comes from delaying their consumption of tannins, but this hypothesis predicts that squirrels should cache Q. macrocarpa acorns. We find that gray squirrels prefer to eat rather than store this WO acorn, despite the fact that it has tannin levels comparable to RO acorns. Steele et al. (2001b) experimentally manipulated tannin levels in acorns; squirrels show no preference for caching high-tannin experimental acorns.

We found significantly higher losses of cotyledon to insect infestation in Q. alba acorns, even though we selected acorns that did not yet show outward signs of insect infestation. Tannins are well known as a deterrent to insects (Schultz, 1989), so the finding that low tannin acorns lost more cotyledon to insects is consistent with expectations. Since most WO species produce acorns with low concentrations of tannin, the preference for caching RO acorns may benefit squirrels by reducing losses to insects. However, not all WO species are low in tannins or high in insect infestation (Fleck and Layne, 1990; Servello and Kirkpatrick, 1989). Squirrels eat rather than cache the WO Q. prinus (Hadj-Chikh et al., 1996), despite its high tannin levels. In this study, we found low rates of insect infestation and high tannin concentrations in the WO Q. macrocarpa. We conclude that the squirrels' preference for storing RO acorns instead of WO acorn may result in a reduction in losses to insects in some cases, but not in others. The insect-avoidance hypothesis is not deemed sufficient to explain the consistency of the preference.

Squirrels preferred to eat rather than cache old, germinating RO acorns, even when there was no visible sign of impending germination. When they did cache the old RO acorns, they excised the embryo on at least some of the acorns. Squirrels treated germinating RO acorns as WO acorns. This indicates that squirrels can detect germination schedule, and use this as the cue influencing their caching preferences. Two other results support this interpretation. First, the RO Q. crassifolia in Mexico is unusual in that its acorns have a very short or no dormancy period. The Mexican gray squirrel (Sciurus areogaster) excises the embryos from germinating Q. crassifolia acorns (Steele et al., 2001a). Second, we have observed gray squirrels digging up RO acorns in the spring, only to excise the embryos and recache them (Steele et al., 2001a; PDS, personal observation). Taken together, these results and observations are strong evidence that it is the differences in germination schedule that most strongly affect squirrel caching preferences.

Experiment 2 serves both our investigation of ultimate advantages and an investigation of proximate cues (Steele et al., 2001b). The relationship between proximate cues and ultimate advantage can be complex, and the factors that serve as proximate cues need not always be the same factors that provide the ultimate advantage (Drickamer, 1998). Therefore, the fact that squirrels respond to germination cues in determining which acorns to cache does not prove that the ultimate advantage to their caching preferences concerns germination per se. Despite the findings in experiment 1, germination could merely the correlated character by which they recognize acorns with other characters, such as levels of tannin and/or insect infestation (e.g., they may use dormancy to recognize high tannin acorns, and cache them merely to delay consuming tannins as long as possible). However, we think it unlikely that squirrels would use such a subtle cue as germination as a surrogate for tannin or insect infestation. Smallwood and Peters (1986) demonstrated that squirrels can detect differences in tannin concentrations comparable to the differences between RO and WO acorns, and Steele et al. (1996) showed that squirrels detect infestation, and preferentially cache uninfected acorns. Given that squirrels can detect these other parameters directly, we think it unlikely that they would use germination schedule as a substitute. Thus, we conclude that squirrels respond to germination cues because germination itself matters. Steele et al. (2001b) investigate the proximate basis of this behavior, and conclude that there is a specific cue in the shell.

While this study has focused on the eastern gray squirrel, other mammals that may serve as dispersal agents for oaks (i.e., those that cache acorns in shallow scatterhoards) seem to exhibit similar preferences for caching intact RO acorns (Ivan and Swihart, 2000; Steele et al., 2001a; unpublished data, PDS). In eastern deciduous forests, blue jays (Cyanocita cristata) are also important dispersal agents for oaks (Scarlet and Smith, 1991). While they are limited to oak species that produce relatively small acorns, there is circumstantial evidence that they too preferentially cache acorns from the RO group (Scarlet and Smith, 1991; Smallwood et al., 1998; Steele and Smallwood, 2001). We suspect that the ultimate basis for this general phenomena of differential dispersal of RO acorns resides with their winter dormancy, making them less perishable as a stored food.

Implications for the ecology and evolution of oaks

Since WO acorns are usually killed by the known dispersal agents for oaks, one would think that they are most likely to survive and become seedlings if they are not discovered by animals. Thus, they should often establish where they fell-near the parent tree. RO acorns are carried off and cached; therefore, RO seedlings should be more widely dispersed from their parent trees than WO seedlings. However, seed dispersal does not always translate directly into seedling dispersal patterns, let alone adult plant distributions. Many other factors may intervene to distort or completely eliminate the influences of seed dispersal patterns on subsequent distributions (Howe, 1986; Schupp and Fuentes, 1995; Nathan and Muller-Landau, 2000). In this oak system, high seed/seedling mortality near the parent tree could yield a seedling dispersal pattern that shows no differences between RO and WO species.

Efforts to test this prediction have begun. It is difficult to estimate seed dispersal distances from the patterns of seedlings found, especially within a patch of adult trees (Nathan and Muller-Landau, 2000). This is because seedlings that disperse further from their parent tree often end up closer to a non-parent tree, making it difficult to assign seedlings to their parents. A number of statistical methods have been developed. Smallwood et al. (1998) employed the methods of Ribbens et al. (1994), in a hardwood forest in northern Virginia. There were four species of oaks in their study: two RO (Q. rubra, Q. velutina) and two WO (Q. alba, Q. prinus). Their best estimates for the average dispersal distances for the WO species were near 5 m from the parent tree. RO species were estimated to have much longer average dispersal distances (~14 m for Q. rubra, and ~40 m for Q. velutina). More extensive efforts to test this hypothesis over a wider geographic area are under way. We are using DNA fingerprinting to determine parent-offspring relationships, which allows us to measure dispersal distances directly.

Several ecological and evolutionary questions are raised by the differential dispersal of RO and WO acorns by animals. Since WO acorns seem most likely to escape predation if they are not found by the animals, WO seedlings may be adapted to thrive in the conditions found near parent trees. Since RO seedlings are typically dispersed from their parent tree by animals, RO species should be more likely to have seeds and seedlings arrive first into open areas-if their seedlings can survive conditions found there. Thus RO seedlings should be adapted to a wider range of environmental conditions, while WO seedlings should be adapted to thrive in the relatively shaded conditions found next to their seed parent. It suggests that RO species should be able to extend their range faster than WO species, although more needs to be known about blue jay foraging behaviors to strengthen that prediction. If these predictions are supported, they suggest the existence of counterbalancing advantages that allow WO species to persist in the face of comparatively low dispersal. We are currently testing these predictions.

AcKNOWLEDGMENTS

PDS thanks the Department of Biology at the University of Richmond, and the Department of Ecology and Evolutionary Biology at the University of Arizona for support. MAS thanks the Department of Biology of Wilkes University, and SHF thanks the department of Zoology at Arizona State University. We thank Kyle Hammon for help with the tannin assays and the microbial assays. We thank Wendy Marussich, two anonymous reviewers and the editor of this journal for their comments on the manuscript. Finally we acknowledge the support of NSF grant # DBI-9978807 to MAS and PDS during the preparation of this manuscript.

1 From the Symposium An Integrative Approach to the Study of Terrestrial Plant-Animal Interactions presented at the Annual Meeting of the Society for Comparative and Integrative Biology, 5-8 January 2000, at Atlanta, Georgia.

REFERENCES

Barnett, R. J. 1977. The effect of burial by squirrels on germination and survival of oaks and hickory nuts. Am. Mid]. Nat. 98:319-330.

Bate-Smith, E. C. 1977. Astringent tannins of Acer species. Phytochem. 16:1421-1426.

Broadhurst, R. B. and W. T Jones. 1978. Analysis of condensed tannins using acidified vanillin. J. Sci. Food. Ag. 29:788-794.

Chung-MacCoubrey, A. L., A. E. Hagerman, and R. L. Kirkpatrick. 1997. Effects of tannins on digestion and detoxification activity in gray squirrels (Sciurus carolinensis). Physiol. Zool. 70:270-277.

Connell, J. H. 1971. On the role of natural enemies in preventing competitive exclusion in some marine animals and in rain forest trees. In P J. den Boer and G. R. Gradwell (eds.), Proceedings of the advanced study institute in dynamics of numbers in popuations, pp. 298-310. Centre for Agricultural Publishing and Documentation, Wageningen.

Culver, D. C. and A. J. Beattie. 1978. Myrmecochory in Viola: Dynamics of seed-ant interactions in some West Virginia species. J. Ecol. 66:53-72.

Dixon, M. D., W. C. Johnson, and C. S. R. Adkisson. 1997. Effects of caching on acorn tannin levels and Blue Jay dietary performance. Condor 99: 756-764.

Drickamer, L. C. 1998. Vertebrate behavior: Integration of proximate and ultimate causation. Amer. Zool. 38:39-42.

Faeth, S. H. 1986. Indirect interactions between temporary separated herbivores mediated by the host plant. Ecology 67:479-494.

Fleck, D. C., and J. N. Layne. 1990. Variation in tannin activity of acorns of seven species of central florida usa oaks. J. Chem. Ecol. 16:2925-2934.

Fox, J. E 1982. Adaptation of gray squirrel behavior to autumn germination by white oak acorns. Evolution 36:800-809.

Hadj-Chikh, L. Z., M. A. Steele, and P. D. Smallwood. 1996. Caching decisions by grey squirrels: A test of the handling time and perishability hypotheses. Anim. Behav. 52:941-948.

Hanley, T A., C. T Robbins, A. E. Hagerman, and C. McArthur. 1992. Predicting digestible protein and digestible dry matter in tannin containing forages consumed by ruminants. Ecology 72:537-541.

Harms, K. E., S. J. Wright, 0. Calderon, A. Hernandez, and E. A. Herre. 2000. Pervasive density-dependent recruitment enhances seedling diversity in a tropical forest. Nature 404:493-495.

Herrera, C. M. 1986. Vertebrate-dispersed plants: Why don't they behave the way they should? In E. A. and T H. Fleming (eds.), Furgivores and seed dispersal, pp. 5-18. Junk Publishers, Dorderecht, Netherlands.

Howe, H. E 1986. Seed dispersal by fruit-eating birds and mammals. In D. R. Murray (ed.), Seed dispersal, pp. 123-190. Academic Press, Sydney, Australia.

Howe, H. E and M. N. Miriti. 2000. No question: Seed dispersal matters. Trends Ecol. Evol. 15:434-436. Howe, H. E and J. Smallwood. 1982. Ecology of seed

dispersal. Ann. Rev. Ecol. Syst. 13:201-228.

Ivan, J. S. and R. K. Swihart. 2000. Selection of mast by granivorous rodents of the central hardwood forest region. J. Mammal. 81:549-562.

Jacobs, L. F 1992. The effect of handling time on the decision to cache by grey squirrels. Anim. Behav. 43:522-524.

Janzen, D. H. 1970. Herbivores and the number of tree species in tropical forests. Amer. Nat. 104:501528.

Janzen, D. H. 1971. Seed predation by animals. Ann. Rev. Ecol. Syst. 2:465-492.

Jensen, R. J. 1992. Acorn size redux. J. Biogeogr. 19: 573-579.

Johnson, W. C., L. Thomas, and C. S. Adkisson. 1993. Dietary circumvention of acorn tannins by blue jays: Implications for oak demography. Oecologia 94:159-164.

Kaul, R. B. 1985. The reproductive morphology of Quercus (Fagaceae). Am. J. Bot. 72:1962-1977.

Koenig, W D. and S. H. Faeth. 1998. Effects of storage on tannin and protein content of cached acorns. Southwest. Nat. 43:170-175.

Levey, D. J. and C. W. Benkman. 1999. Fruit-seed dispersers interactions: Timely insights from a long-term perspective. Trends Ecol. Evol. 14:4143.

Loreto, F, P. Ciccioli, E. Brancaleoni, R. Valentini, M. de Lillis, 0. Csiky, and G. Seufert. 1998. A hypothesis on the evolution of isoprenoid emission by oaks based on the corelation between emission type and Quercus taxonomy. Oecologia 115:302305.

Martin, J. S. and M. M. Martin. 1982. Tannin assays in ecological studies: Lack of correlation between phenolics, proanthocyanidins, and protein-precipitating ability. Oecologia 54:205-211.

Miller, H. A. and S. H. Lamb. 1985. Oaks of North America. Naturegraph Publishers, Happy Camp, CA, USA.

Nathan, R. and H. C. Muller-Landau. 2000. Spatial patterns of seed dispersal, their determinants and consequences for recruitment. Trends Ecol. Evol. 15:278-285.

Ofcarcik, R. P. and E. E. Burns. 1971. Physical and chemical properties of selected acorns. J. Food. Sci. 36:576-578.

Packer, A. and K. Clay. 2000. Soil pathogens and spatial patterns of seedling mortality in a temperate zone tree. Nature 404:278-281.

Post, D. and 0. J. Reichman. 1991. Effects of food perishability, distance, and competitors on caching behavior by eastern woodrats. J. Mammal. 72: 513-517.

Reichman, 0. J. 1988. Caching behavior by eastern woodrats, Neotoma floridana. Anim. Behav. 36: 1525-1532.

Ribbens, E., J. A. Silander, and S. W. Pacala. 1994. Seedling recruitment in forests: Calibrating models to predict patterns of tree seedling dispersion. Ecology 75:1794-1806.

Scarlett, T L. and K. G. Smith. 1991. Acorn preference of urban blue jays cyanocitta-cristata during fall and spring in northwestern arkansas. Condor 93: 438-442.

Schultz, J. C. 1989. Tannin-insect interactions. In R. W. Hemingway and J. J. Karchesey (eds.), Chemistry and the significance of condensed tannins, pp. 412-433. Plenum Press, New York.

Schupp, E. W. and M. Fuentes. 1995. Spatial patterns of seed dispersal and the unification of plant population ecology. Ecoscience 2:267-275.

Servello, F A. and R. L. Kirkpatrick. 1989. Nutritional value of acorns for ruffed grouse. J. Wild. Mgmt. 53:26-29.

Short, H. L. 1976. Composition and squrirel use of acorns of black and white oak groups. J. Wild. Mgmt. 40:479-483.

Short, H. L. and E. A. Epps. 1976. Nutrient quality and digestibility of seeds and fruits from southern forests. J. Wild. Mgmt. 40:283-289.

Smallwood, P. D. and W. D. Peters. 1986. Grey squirrel food preferences: The effects of tannin and fat. Ecology 67:168-174.

Smallwood, P D., M. A. Steele, E. Ribbens, and W. J. McShea. 1998. Detecting the effect of seed hoarders on the distribution of seedlings of tree species: Gray squirrels (Sciurus carolinensis) and oaks (Quercus) as a model system. In M. A. Steele, J. E Merritt, and D. A. Zegers (eds.), Ecology and evolutionary biology of tree squirrels, Vol. 6. pp. 211-222. Virginia Museum of Natural History, Martinsville, Virginia.

Smith, C. C. and D. Follmer. 1972. Food preferences of squirrels. Ecology 53:83-91.

Smith, C. C. and 0. J. Reichman. 1984. The evolution of food caching by birds and mammals. Ann. Rev. Ecol. Syst. 15:329-351.

Sokal, R. R. and F. J. Rohlf. 1981. Biometry. W. H. Freeman and Co., New York.

Stapanian, M. A. and C. C, Smith. 1984. Density-dependent survival of scatterhoarded nuts: An experimental approach. Ecology 65:1387-1396.

Steele, M. A., L. Z. Hadj-Chikh, and J. Hazeltine. 1996. Caching and feeding decisions by Sciurus carolinensis: Responses to weevil-infested acorns. J. Mammal. 77:305-314.

Steele, M. A., T Knowles, K. Bridle, and E. L. Simms. 1993. Tannins and partial consumption of acorns:

Implications for dispersal of oaks by seed predators. Am. Midi. Nat. 130:229-238.

Steele, M. A. and P. D. Smallwood. 1994. What are squirrels hiding? Natural History 10:40-45. Steele, M. A. and P. D. Smallwood. 2001. Acorn dis

persal by birds and mammals. In W. McShea and J. W. Healy (eds.), Ecology and management of oaks for wildlife. John Hopkins University Press, Baltimore, MD. (In press)

Steele, M. A., G. Turner, P. D. Smallwood, J. 0. Wolff, and J. Radillo. 2001a. Cache management by small mammals: Experimental evidence for the significance of acorn embryo excision. J. Mammal. 82:35-42.

Steele, M. A., P D. Smallwood, A. Spunar, and E. Nelsen. 2001b. The proximate basis of the oak dispersal syndrome: Detection of seed dormancy by rodents. Amer. Zool. 41:000-000.

Stiles, E. W. 1989. Fruits, seeds, and dispersal agents. In W. G. Abrahamson (ed.), Plant-animal interactions, p. 87-122. McGraw Hill Publishing, New York.

Stiles, F G. 1980. Patterns of fruit presentation and seed dispersal in bird-disseminated woody plants in the eastern deciduous forests. Amer. Nat. 116: 670-688.

Vander Wall, S. B. 1990. Food hoarding in animals. The University of Chicago Press, Chicago. Weckerly, E W., D. W. Sugg, and S. R. D. 1989. Ger

mination success of acorns (Quercus), insect predation, and tannins. Can. J. For. Res. 19:811-815. Wenny, D. G. 2000. Seed dispersal of a high quality

fruit by specialized frugivores: High quality dispersal? Biotropica 32:327-337.

Wenney, D. G. and D. J. Levey. 1998. Directed seed dispersal by bellbirds in a tropical cloud forest. Proc. Nat. Acad. Sci. USA. 65:6204-6207.

Willson, M. E 1992. The ecology of seed dispersal. In M. Fenner (ed.), Seeds: the ecology of regeneration in plant communities, pp. 61-65. CAB International, Wallingford, UK.

Young, J. A. and C. G. Young. 1992. Seeds of woody plants in North America. Dioscorides Press, Portland, OR, USA.

PETER D. SMALLWOOD,2,* MICHAEL A. STEELE,^ AND STANLEY H. FAETH^^

*Department of Biology, University of Richmond, Richmond, Virginia 23173 ^Department of Biology, Wilkes University, Wilkes-Barre, Pennsylvania 18766 ^^Department of Zoology, Arizona State University, Tempe, Arizona 85287

2 E-mail: psmallwo@richmond.edu

Copyright Society for Integrative and Comparative Biology Sep 2001
Provided by ProQuest Information and Learning Company. All rights Reserved