Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Molecular ecology wikipedia , lookup
Overexploitation wikipedia , lookup
Island restoration wikipedia , lookup
Biodiversity action plan wikipedia , lookup
Occupancy–abundance relationship wikipedia , lookup
Storage effect wikipedia , lookup
Habitat conservation wikipedia , lookup
Ecology of Banksia wikipedia , lookup
Ecological fitting wikipedia , lookup
Latitudinal gradients in species diversity wikipedia , lookup
Competition and the Structure of Granivore Communities Diane W. Davidson, James H. Brown, and Richard S. Inouye Historically, studies of competitive in t e r a c t i o n s in e c o s y s t e m s h a v e b e e n based on one of two approaches. A com parative approach has dominated inves tigations of competition between taxo n o m i c a l l y r e l a t e d s p e c i e s that s h a r e similar morphological, physiological, and behavioral traits. Notwithstanding their general likenesses, these species of ten differ from each other in one or a few characteristics that reduce overlap in the use of resources. Ecologists have inter preted the differences in the body sizes, feeding structures, microhabitat special izations, and foraging behaviors of re lated and coexisting species as evidence for t h e i m p o r t a n c e o f c o m p e t i t i o n in structuring natural communities. The theoretical underpinnings o f this com parative approach to the study of com petition predict "limiting similarities** in characteristics relevant to allocation of scarce resources (May and MacArthur 1972, and reviewed in Schoener 1974). A second approach to studying com petition employs experimental manipula tions of resources and/or competitive re gimes. Because experimental methods do not depend on similarities of form and function between presumed competitors, this approach is more general and has proven especially useful in analyzing in teractions between phylogenetically un related species. Such interactions are particularly common in space-limited systems; for example, organisms as dis tant taxonomically as plants and sessile a n i m a l s c o m p e t e for s p a c e o n r o c k y ocean shores (Lubchenko and Menge 1978). Like competition for space, com petition for food resources is not limited to closely related species. As investiga tions have expanded to include distantly r e l a t e d t a x a , t h e i n a d e q u a c i e s o f the Davidson is an assistant professor of biology, Uni versity of Utah, Salt Lake City, UT 84112; Brown is a professor and Inouye is a graduate student, both in the Department of Ecology and Evolutionary Biol ogy, University of Arizona, Tucson, AZ 85721. © 1980 American Institute of Biological Sciences. All rights reserved. April 1980 comparative approach have b e c o m e in creasingly apparent, and experimenta tion has gradually assumed a more prom i n e n t role in c o m m u n i t y - e c o l o g i c a l research. Here w e trace the development of our investigations of granivory in desert eco systems, illustrating the synthesis of the comparative and experimental approach es and noting the essential contributions of both. In the process, w e also call at tention to several major difficulties inher ent to experimentation on this scale and describe some relatively unconventional experiments designed to circumvent some of these problems. DESERT GRANIVORES A N D THEIR RESOURCES Associations of seed consumers in arid habitats of the southwestern United States have proven very useful systems for analyzing competitive interactions in e c o s y s t e m s . Local granivore assem blages frequently contain representa tives of several kinds of specialized seedeaters, including birds, rodents, ants, and other insects. Our studies have em phasized ants and rodents, t w o groups of seed harvesters that are abundant and permanent residents of local habitats and have relatively generalized diets. As many as 14 c o m m o n species of seed-eat ing ants and rodents may coexist and share seed resources in local habitats (see Figure 1). Seeds of both annual and perennial p l a n t s are c o n s u m e d b y d e s e r t granivores, but those of annuals predominate in the diets of ants and rodents. The seeds of annuals may regularly consti tute more than 95% of the total standing crop of seeds by number and more than 80% by b i o m a s s ( N e l s o n a n d C h e w 1977). Rapid, facultative development of annuals in response to available moisture effectively couples production of green vegetation and seeds to the infrequent and unpredictable precipitation that lim its productivity in arid regions. Geo graphic gradients in the magnitude and predictability of precipitation are reflect ed in patterns of seed production. East ward and northward from the Mojave Desert of southern California, both the mean annual precipitation and the mean amount of predictable precipitation in crease, and there are corresponding in creases in the productivity and standing crop biomass of both vegetation and seeds. THE COMPARATIVE APPROACH The distribution and coexistence of ants and rodents over the southwestern deserts afforded an excellent opportunity to gain rapid insight into processes deter mining the diversity and structure of granivore communities. We began by censusing rodents (Brown 1973, 1975) and ants (Davidson 1977a, b, Brown and Davidson 1977) over a gradient of precip itation and seed productivity. Along a transect eastward from south ern California into southern Arizona and southwestern N e w M e x i c o , patterns in the species diversity of ants and rodents are strikingly alike. Within each fauna, species richness increases along the gra dient of resource productivity. The most diverse local assemblages of granivores occupy the extreme southeastern corner of Arizona, where rainfall and seed pro duction are most bountiful and predict able. Estimates of ant and rodent numbers and biomass also vary directly with pro ductivity. Within each of the t w o major taxa, the number of c o m m o n species that occur together in a local habitat is signifi cantly fewer (by approximately five spe cies for both rodents and ants) than the number of species w h o s e geographic dis tributions make them potential colonists of these sites. These patterns constituted our first indirect evidence that coexis tence of granivores might be limited by competition for food. Further indication of the importance of resource competition emerged from 233 analyses of spatial pattern in granivore communities over the same environmen tal gradient. Ant species characterized by small foraging territories and limited worker mobility decrease in abundance as mean annual precipitation and pro ductivity decline (Davidson 1977a). For these stationary, ground-nesting species, availability of seed resources depends on access to space. Colonies with limited foraging areas are probably unable to ac quire sufficient food for maintenance and reproduction where seed densities are very low. Within individual localities, uniform intraspecific spacing of ant colo nies (Bernstein 1975, Holldobler 1976) and of mounded burrow systems of some rodents (Schroder and Geluso 1975) re flect the desirability of excluding conspecifics from the feeding grounds. Indirect evidence for resource c o m petition also c o m e s from the composi tion of local granivore a s s o c i a t i o n s . Where food resources limit consumer populations, differential use of resources should be a requirement for coexistence. Interspecific morphological and behav ioral specializations characterize co-oc curring species of both ants and rodents. The most striking patterns involve inter specific differences in body sizes (see Figure 1). Coexisting associations of har vester ants often differ conspicuously in worker body size, although species of similar body size can coexist if they dif fer in colony foraging m o d e (Davidson 1977a). Experiments and observations suggest that column-foraging ants have relatively p o p u l o u s c o l o n i e s that spe cialize temporally and spatially on ener getically rich c o n c e n t r a t i o n s of s e e d , whereas species w h o s e workers forage solitarily gather primarily dispersed seeds (Davidson 1977b). Harvester species that are alike in both worker body size and colony forag ing behavior tend not to coexist locally within homogeneous habitats, e v e n at lo calities that fall within the distributional ranges of both species. The interspecific territorial defense that occurs between s o m e such s p e c i e s pairs (Holldobler 1974, Whitford et al. 1976) suggests that these species are strong competitors and serve as ecological replacements for one another (Davidson 1977a). The importance of worker body size to coexistence of harvester ants is perhaps best exemplified by the pattern of size specialization in relationship to faunistic diversity (Davidson 1978a). Dominating the depauperate communities of the rela tively arid and unproductive Mojave and C o l o r a d o d e s e r t s is Veromessor per- 234 Fig. 1 . Seed-eating ants and rodents inhabiting the Chiricahua Bajada study site near Portal, Arizona. Within each group note the obvious body size differences among coexist ing species. From left to right, the ants are (top row): Aphaenogaster (Novomessor) cocfcerelli (I = individual forager); Pogonomyrmex rugosus (C = column forager); P. desertorum ( I ) ; a n d ( b o t t o m r o w ) : Pheidole desertorum (I, worker and soldier castes); Solenopsis xyloni (C, minor media, and major); Ph. xerophila (C, worker and soldier) and Ph. sitarches (C, worker and soldier). Rodents in the leflhand column include only spe cialized seed-eaters. From top to bottom, these are Perognathus flavus, P. penicillatus, Dipodomys merriami, D. ordi, and D. spectabilis. In the righthand column are included two additional species whose more generalized diets also contain many seeds. From top to bottom, these are Perognathus flavus, Reithrodontomys megalotus, P. penicillatus, Peromyscus maniculatus, Dipodomys merriami, D. ordi, and D. spectabilis. gandei, whose colonies display a contin uous polymorphism in worker body size. Where this species is the only c o m m o n harvester ant, the polymorphism is most extreme, and workers range from 3.5 to 8.5 mm in body length. The pattern of w o r k e r - s i z e p o l y m o r p h i s m in V. pergandei responds in a remarkedly precise way to changes in the competitive envi ronment (Davidson 1978a). In more pro ductive habitats at the eastern edge of its distribution, this ant occurs with larger and smaller ant species, and its colonies have relatively monomorphic workers of an intermediate body size. In the still more diverse and species-rich Chihuahuan Desert of southeastern Arizona, V. pergandei is replaced by more special ized monomorphic species. Patterns in the organization of commu- Bioscience Vol. 30 No. 4 nities of harvester ants suggest that dif ferences in worker body sizes function ally differentiate the ecological roles of coexisting species. T w o kinds of empiri cal data support this interpretation (Da v i d s o n 1977a, 1978a). In e x p e r i m e n t s where ants were permitted to select from four sizes of quality-standardized seeds, worker size was strongly and positively correlated with an index of seed size both in interspecific comparisons and within colonies of V. pergandei. Similar relationships hold for ants collecting na tive seeds. Mechanistic upper limits and energetic lower limits appear to determine the siz es of seeds that can be exploited profit ably by worker ants of a given body size (Davidson 1978b). H o w e v e r , body size is a complex character and may influence ecological function in many w a y s . Re cent multivariate analyses relating har vester ant diets to the pattern of resource a v a i l a b i l i t y s u g g e s t ( t h o u g h d o not prove) that large body size also enhances both worker mobility and the capacity for ants to feed selectively on rare and patchily distributed but energetically su perior resources (Davidson, unpublished analysis). Similarly, body size patterns provide circumstantial evidence for competitive organization of rodent communities. In sand dune habitats spanning a latitudinal gradient of increasing precipitation, pro ductivity, and species richness, the body sizes of coexisting rodents differ more than would be anticipated on the basis of random co-occurrence of those species w h o s e geographic ranges allow them potential a c c e s s to these sites (Brown 1973). Structurally and functionally con vergent rodent communities characterize geographically distinct deserts of similar productivity (Brown 1975). Different species with similar body size and locomotory specialization represent eco logical counterparts inhabiting geograph ically disjunct deserts. Because rodents are much larger in relationship to their resources than are ants, simple differences in the sizes of seeds foraged are less likely to provide a f u n c t i o n a l b a s i s for b o d y s i z e dis placement. For seed types differing in size and dispersion, the foraging effi ciencies of rodents are apt to be deter mined by a number of morphological and behavioral characteristics that correlate with body size. For example, the greater energetic expenditures of larger bipedal rodents such as kangaroo rats (genus Dipodomys) may constrain them to for age on large seeds or clumps of smaller April 1980 seeds, whereas smaller scansorial ro dents can profitably use small scattered seeds (Brown 1975, Brown and Lieberman 1 9 7 3 , H u t t o 1978, P r i c e 1978a, Reichman and Oberstein 1977). Tendencies for ecological separation in foraging microhabitats are correlated with locomotor specializations in desert rodents. While quadrupedal pocket mice (genus Perognathus) tend to feed pre dominantly in or under vegetation, bi pedal kangaroo rats (genus Dipodomys) and kangaroo mice (Microdipodops) are w e l l a d a p t e d t o l o c a t e s e e d s in the open. Their saltatorial locomotion per mits rapid coverage of the territory sepa rating resource patches and may also fa cilitate escape from predators (Brown 1975, Brown and Lieberman 1973, Lemen and Rosenzweig 1978, Price 1978b, Rosenzweig 1973, Wondolleck 1978). That such microhabitat separation m a y b e m a i n t a i n e d in part b y inter specific competition is evidenced by shifts in r e s o u r c e u t i l i z a t i o n a c c o m panying changes in the competitive envi ronment. On isolated sand dune habitats in the eastern Great Basin, where Dipo domys ordi occurs with a few other ro dent species, this species forages in a wider range of microhabitats, broadens the seed diet, and attains greater popu lation densities (Brown 1973) than where it coexists with diverse species of ro dents on less isolated dunes. 1 Thus, studies based on the compara tive approach have revealed remarkably similar patterns of community diversity and structure in t w o major taxa of generalist granivores. We can infer that com petition has played a prominent role in the assembly of local communities of both harvester ants and seed-eating ro dents. Although circumstantial, the evi dence for competition is strong because it has been derived from several inde pendent lines of argument. Patterns of e c o l o g i c a l separation within the two g r o u p s h a v e p r o b a b l y a r i s e n b y dis tinctive routes that reflect basic biologi cal differences between the t w o taxa. Despite such differences, the form of re source allocation is similar within ant and rodent communities, probably be cause of the similarities in the availabili ty of seed resources to both groups. Extensive biological differences be tween ants and rodents do not exclude the potential for competitive interactions between them, but they do preclude our 'Unpublished data from J. H. Brown, coauthor, and Eric Larsen, Department of Ecology and Evolution ary Biology, University of Arizona, Tucson. use of the comparative approach to seek indirect evidence of such competition. Broad overlap in the sizes (Brown and Davidson 1977) and species (Table 1) of seeds used by the two taxa suggests the potential for competition. But though dietary overlap is a necessary condition for c o m p e t i t i o n , it is not a sufficient criterion. Tentative biogeographic evidence also hints at the possibility of competition between harvester ants and seed-eating rodents ( B r o w n and D a v i d s o n 1977). Whereas both faunas b e c o m e increas ingly abundant and species-rich along a longitudinal gradient of enhanced precip itation and production, harvester ants decline in diversity and biomass over a similar but latitudinal gradient. Cooler mean temperatures at high latitudes in the Great Basin Desert may confront ectothermic ants with abbreviated forag ing seasons and actually reduce the level of seed productivity effectively harvestable by them. Rodents appear to com pensate for the missing ants by becoming e v e n more populous and diverse (Brown and Davidson 1977). But the possibility of niche expansion and density compen sation in the rodent fauna must be viewed cautiously, since translation of precipita tion to productivity need not be identical along the two environmental gradients where the seasonality of precipitation is notably different. Cross-community com parisons fail to provide really convincing data on the extent and consequences of competition between ants and rodents. EXPERIMENTAL APPROACHES An early experiment was designed as a simple and direct test of the hypothesis that desert ants and rodents compete for limiting seed resources. T o confirm this hypothesis, it was necessary to demon strate that each major taxon grows more abundant when the other group is re moved. Choosing a broad area of homo geneous Sonoran Desert scrub habitat, we established two replicates each of four experimental treatments (Brown and Da vidson 1977). From two circular plots 18 m in radius, w e excluded rodents alone with hardwarecloth fencing, partially buried to discourage tunneling from out side. By poisoning, w e eliminated ants from two similar but unfenced plots. We removed both ants and rodents from a third set of exclosures, and left one final pair of plots unmanipulated as controls. After only two years, we observed re ciprocal density compensations in the 235 TABLE 1. Items composing > 5 % of the diets of ant and rodent granivores at the Chiricahua Bajada study site.* Rodentsf Seeds Amaranthus palmeri Astragalus sp. Boerhaavia coulteri Bouteloua sp. Chorizanthe sp. Cryptantha sp. Descurainia pinnata Ephedra trifurca Eriogonum abertianum Eriogonum trichopes Euphorbia micromera Euphorbia serpyllifolia Kallstroemia grandiflora Larrea tridentata Lepidium medium Panicum arizonicum Prosopis juliflora Tidestromia lanuginosa Unidentified grass Unidentified seed X 1 2 Antsf 3 4 5 6 7 8 + 9 10 + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + *Based on seeds robbed from laden ants or found in rodent cheek pouches (or D.s. mounds) during spring, fall, and summer over the years 1 9 7 2 - 7 3 for rodents and 1 9 7 4 - 7 8 for ants. Because ant and rodent diets were sampled under different resource regimes, the overlap in 2 5 % of resource categories used by the two groups probably represents an underestimate of actual overlap. f K E Y : Rodents: 1 = Dipodomys spectabilis, 2 = Dipodomys merriami, 3 = Perognathus pennicillatus, 4 = Perognathus flavus. Ants: 5 = Aphaenogaster cockereHi, 6 = Pogonomyrmex rugosus, 7 = Pogonomyrmex desertorum, 8 = Pheidole desertorum, 9 = Pheidole xerophila, 10 = Pheidole sitarches. two faunas. In comparision with control had qualitatively different effects on the species may indirectly benefit ants by fa plots, ant colonies reached 7 1 % higher plant community (Inouye et al. 1980). cilitating growth and seed production by densities on plots from which rodents The predominant ants on the plots were small-seeded plants. The reciprocal den had been excluded. Rodent populations small column-foraging species of Phei sity compensations suggest that, at least increased 20% numerically and 29% in dole that specialized on the tiny but nu in the short term, competition for shared biomass in the absence of ants. Simulta merically dominant seeds of Filago cali- resources is the dominant interaction. neous monitoring of seed levels on one fornicus (Family Compositae). Predation E v e n if resources do not interact, the po of the r e p l i c a t e d s e t s of t r e a t m e n t s by the ant community as a whole may tential importance of indirect pathways (Reichman 1980, Brown et al. 1979) sug commonly fall most heavily on abundant can complicate studies of competition in gested that these population responses seed types (Davidson 1980), promoting c o m m u n i t i e s o f t h r e e or m o r e c o m were mediated through exploitative re diversity within the surviving seed pool. petitors (Levine 1976). Circumstantial source competition. Seed abundance, as In contrast, the rodents (principally evidence (Davidson 1980) indicates that measured either by numbers or biomass, Dipodomys rugosus, SL large and ag merriami, Perognathus am- Pogonomyrmex was reduced to approximately the same plus, and P. penicillatus) foraged selec gressive harvester ant, may indirectly fa level on plots where either ants or ro tively on large-seeded species such as cilitate coexisting ant species by dif dents or both were present, while seeds Er odium cicutarium, E. texanum, and ferentially interfering with foraging and accumulated to approximately four times Lotus humistratus, an in that tend to dominate nest-founding by P. desertorum, this level on the plot from which both annual plant biomass. With their greater termediate-sized ant w h o s e resource re faunas had been removed. s t a r t i n g r e s e r v e s , t h e s e s p e c i e s m a y quirements strongly overlap those of both Recent analyses of plant responses on g r o w r a p i d l y t o o u t c o m p e t e s m a l l e r larger and smaller species. these experimental plots (Inouye et al. seeded plants. On plots containing ro To decipher further the mechanisms of 1980) h a v e d o c u m e n t e d c o m p e t i t i o n dents, differential harvesting of Erodium interaction among granivores and their within the producer trophic level as well, and Lotus species permitted Euphorbia resources, we have repeated and ex and underscore the potential complexity polycarpa to reach significantly higher panded our experiments in southeastern of interactions in this sytem. Density-de densities. Presumably, the smaller seeds Arizona on the eastern bajada of the pendent germination and/or mortality of this species (approximately .18 mg in Chiricahua M o u n t a i n s . H e r e , an e x among annual plants is suggested by the contrast to 1.5 mg for the larger-seeded ceptionally rich association of granivores fact that, while seed densities quadru species) were less attractive to rodent includes s e v e n c o m m o n harvester ants pled in the absence of ant and rodent pre- granivores. and five-to-seven relatively abundant dation, plant densities merely doubled Competition among producers greatly species of seed-eating rodents (see Fig (Brown et al. 1979). Watering and thin enhances the potential complexity of in ure 1). In so diverse an assemblage of ning experiments undertaken on the same t e r a c t i o n s b e t w e e n a n t s and r o d e n t s . species, the number of potential direct plots r e v e a l e d significant c o m p e t i t i v e Not only may these groups compete for and indirect interactions is very great, inhibition of germination, growth rates, shared resources, but they may interact and firm hypotheses about particular in biomass, and fecundity in annual plants indirectly through their use of non-over terspecific interactions are difficult to (Inouye et al. 1980). lapping but competing resources. For ex pose and test. Seed predation by ants and rodents ample, rodent predation on large-seeded Our strategy has been instead to ma- 236 Bioscience Vol. 30 No. 4 nipulate broad parameters which prelim inary observations and experiments sug gest may have demonstrable effects on community organization. In a number of treatments, w e are supplementing natu ral resources with milo s e e d s , applied ei ther continuously (V12 of the total annual subsidy supplied each month) or in a single annual pulse coincident with sum mer seed set by ephemeral plants. Early observations indicate interspecific dif ferences within each of the two major taxa in the proficiencies with which spe cies use aggregated and dispersed re sources, and preliminary experiments (Brown et al. 1975) suggest that ants and rodents differ functionally with respect to this same trait. Greater mobility and a capacity to gather large quantities of seed into external, fur-lined cheek pouches enable heteromyid rodents to be far more successful than ants at deplet ing dense concentrations of experimen tally supplied seed. Rodents tend to ig nore low-density seed subsidies, perhaps because the higher metabolic costs of endothermy prevent them from foraging profitably at low seed densities. Ants are apparently equally adept at using lowand high-density seed supplements. Both enhanced searching efficiencies and re duced foraging costs, associated with eusociality and ectothermy, respectively, probably contribute to the ability of ants to f e e d e c o n o m i c a l l y f r o m d i s p e r s e d seed distributions. The demonstrated importance of size differentiation in coexisting granivores s u g g e s t s the p o s s i b i l i t y o f p e r t u r b i n g community organization by altering the spectrum of seed sizes. Among the newly established continuous seed addition plots, treatments are differentiated with respect to the particle sizes (small, large, and heterogeneous) of experimentally supplied milo. In addition to repeating our whole faunal removal experiments, w e have es tablished replicated plots from which in dividual dominant ant or rodent species have been eliminated. These latter ex periments should contribute to our un derstanding of e c o s y s t e m structure. Re cently, ecologists have hypothesized that high diversity may be maintained in com plex e c o s y s t e m s if interactions center in small associations of closely interacting species that interact relatively weakly with other such groups. Provided such s u b s t r u c t u r e e x i s t s in o u r s y s t e m of granivores and resources, we might an ticipate a priori that organization should follow taxonomic lines. Single-species removal experiments should lend insight April 1980 to the problem of whether or not inter actions are significantly stronger within than between major taxa. Biogeographic comparisons of granivore communities hint at the possible role that "guild structure" may play in the maintenance of diversity. Over a lati tudinal transect where precipitation in creases but temperatures change very little, greater diversity is achieved by al most parallel increases in species rich ness in ant and rodent faunas, rather than by adding members of only one tax onomic association or the other (Brown and Davidson 1977). A b o v e a certain threshold of seed productivity, diversity is enhanced by addition of a third (avian) guild of granivores. The finding that die tary similarity among coexisting rodent species e x c e e d s that between cooccurring rodents and ants (Davidson, unpub lished data) also lends support to the sug g e s t i o n of t a x o n o m i c a l l y based guild structure. T w o years following initiation of these experiments, our results are still prelimi nary; only qualitative statements are jus tified at this time. First, granivore popu lations have responded more gradually to our perturbations than in past experi ments. At this time, w e do not know w h e t h e r the s y s t e m ' s r e s i s t a n c e to change represents enhanced stability as sociated with a more complex and/or c o e v o l v e d interaction structure or sim ply the obviation of competitive stress by the recent bounty of natural seed pro duction. Competition need act only in termittently to be instrumental in regu lating c o m m u n i t y structure, and prolonged maintenance of experimental treatments may be necessary to demon strate its effects. Considerable time may also be re quired before we can distinguish tempo rary from equilibrial responses to our manipulations. The most immediate re sponses to seed additions and species re movals may occur in species that are op portunists rather than in the specialized species that eventually will prove to be the superior competitors for available re sources. Opportunism is promoted by the ability to colonize through immigra tion and/or reproduction. Where kanga roo rat species (genus Dipodomys) have been excluded from experimental plots, pocket mice (genus Perognathus) have quickly immigrated to take advantage of surplus resources. Although harvester ants (Pogonomyrmex rugosus) have been observed to capitalize on Erodium seeds accumulating on rodent-free plots, this expansion in resource use has not yet b e e n translated through reproduction into higher colony densities. On Chiricahua Bajada plots where ant densities have risen, Pheidole desertorum has contributed most significantly to the increase. This true opportunist can colonize by overground colony move ment and grow from small single-colony units to fill expansive nests with many disjunct units. N e w l y established queens of the species are similarly well-suited for taking advantage of underutilized for aging territory; their first worker broods are larger than those of coexisting ant species, and worker development is sig nificantly more rapid. Finally, short-term and long-term re sponses may differ for yet another rea son. Responses involving very indirect routes of interaction may be expressed much later than those mediated through relatively short and direct pathways. CONCLUSIONS T w o fundamentally different method o l o g i e s h a v e m a d e u n i q u e and o f t e n complementary contributions to our un derstanding of granivory in desert eco systems. Biogeographic comparisons of species numbers and characteristics in independently assembled communities quickly provided an overview of the eco logical processes underlying community organization. While lending useful per spective on interactions between closely related groups of c o m p e t i t o r s , crosscommunity comparisons have proven less fruitful in identifying patterns of as sociation between taxonomically distant competitors. Experiments provide an al ternative approach that we have em ployed successfully to demonstrate com petitive interactions between ants and r o d e n t s as w e l l a s a m o n g primary producers. At least two factors complicate the de s i g n i n g of a d d i t i o n a l e x p e r i m e n t s to p h r a s e and t e s t e x p l i c i t h y p o t h e s e s about interactions among these distantly related taxa. It is unclear which of the many differences between ant and rodent granivores favor their coexistence in des ert e c o s y s t e m s . Functional differenti ation of the two groups may well depend on numerous minor distinctions in the use of seed resources. Secondly, poten tial interactions among granivores through distinct but competing resources, as well as through those that are shared, not only complicate hypothesis-formulation but may prolong the time period over which the system responds to a given perturbation. 237 Our ongoing experiments are uncon ventional in the sense that their immedi ate goal is description rather than hypoth esis testing. B y manipulating resource parameters whose natural variation figures p r o m i n e n t l y in r e s o u r c e s u b d i v i s i o n among closely related competitors, w e hope to alter community structure to a degree that will enable us to recognize major patterns in the organization of consumer-resource interactions. Brown, J. H. 1973. Species diversity of seedeating desert rodents in sand dune habitats. Ecology 54: 775-787. 1975. Geographical ecology of desert rodents. Pages 315-341 in M. L. Cody and J. M. Diamond, eds. Ecology and Evolution of Communities. Harvard University Press, Cambridge. Brown, J. H. and D. W. Davidson. 1977. Competition between seed-eating rodents and ants in desert ecosystems. Science 196: 880-882. Brown, J. H., D. W. Davidson, and O. J. Reichman. 1979. An experimental study of competition between seed-eating desert ro ACKNOWLEDGMENTS dents and ants. Am. Zool. 19: 1129-1143. We thank the many colleagues, s t u • Brown, J. H., J. J. Grover, D. W. Davidson, and G. A. Lieberman. 1975. A preliminary dents, and assistants w h o have contrib study of seed predation in desert and mon uted in major ways to our research on tane habitats. Ecology 56: 987-992. g r a n i v o r e s . S p e c i a l t h a n k s g o to A . Brown, J. H., and G. A. Lieberman. 1973. Brown, G. Byers, J. Munger, and K. Resource utilization and coexistence of Zemanek. D. Alstad provided helpful seed-eating rodents in sand dune habitats. criticism of an early draft of the manu Ecology 54: 788-797. script. The investigations summarized Davidson, D. W. 1977a. Species diversity and here have been supported by the Desert community organization in desert seed-eat Biome, US/IBP, and by grants from the ing ants. Ecology 58: 711-724. National Science Foundation. 1977b. Foraging ecology and commu nity organization in desert seed-eating ants. Ecology 58: 725-737. 1978a. Size variability in the worker REFERENCES CITED caste of a social insect {Veromessor perBernstein, R. A. 1975. Foraging strategies of gandi Mayr) as a function of the com ants in response to variable food density. petitive environment. Am. Nat. 112: 523¬ Ecology 56: 213-219. 532. Reprints from BioScience may be ordered directly from our editorial offices. M i n i m u m order: 50 copies. To receive our brochure, "How to Order BioScience Reprints," w h i c h includes a s c h e d u l e of prices, please c o m p l e t e the c o u p o n b e l o w and mail to: BioScience Reprints American Institute of Biological S c i e n c e s 1401 W i l s o n Blvd. Arlington, VA 2 2 2 0 9 P l e a s e s e n d a c o p y of t h e " H o w to Order BioScience brochure to: Reprints' Name. Institution Street City 238 -State. -Zip. 1978b. Experimental tests of the opti mal diet in two social insects. Behav. Ecol. Sociobiol. 4: 35-41. _. 1980. Some consequences of diffuse competition in a desert ant community. Am. Nat., in press. Holldobler, B. 1974. Home range orientation and territoriality in harvester ants. Proc. Natl. Acad. Sci. USA 71(8): 3271-3277. 1976. Recruitment behavior, home range orientation and territoriality in har vester ants, Pogonomyrmex. Behav. Ecol. Sociobiol. 1: 3-44. Hutto, R. 1978. A mechanism for resource al location among sympatric heteromyid ro dent species. Oecologia 33: 115-126. Inouye R. S., G. Byers, and J. H. Brown. 1980. Effects of predation and competition on survivorship, fecundity and community structure of desert annuals. Ecology, in press. Lemen, C. A., and M. L. Rosenzweig. 1978. Microhabitat selection in two species of heteromyid rodents. Oecologia 33: 127-135. Levine, S. 1976. Competitive interactions in ecosystems. Am. Nat. 110: 903-910. Lubchenko, J., and B. A. Menge. 1978. Com munity development and persistence in a low rocky intertidal zone. Ecol. Monogr. 48: 67-94. May, R. M., and R. H. Mac Arthur. 1972. Niche overlap as a function of environmen tal variability. Proc. Natl. Acad. Sci. USA 69: 1109-1113. • Nelson, J. F., and R. M. Chew. 1977. Factors affecting seed reserves in the soil of a Mojave Desert ecosystem, Rock Valley, Nye County, Nevada. Am. Mid. Nat. 97: 300¬ 320. Price, M. V. 1978a. Seed dispersion prefer ences of coexisting desert rodent species. J. Mammal. 59: 624-626. 1978b. The role of microhabitat in structuring desert rodent communities. Ecology 59: 910-921. Reichman, O. J. 1980. Desert granivore forag ing and its impact on seed densities and dis tributions. Ecology, in press. Reichman, O. J., and D. Oberstein. 1977. Se lection of seed distribution types by Dipo domys merriami and Perognathus amplus. Ecology 58: 636-643. t Rosenzweig, M. L. 1973. Habitat selection experiments with a pair of coexisting heter omyid rodent species. Ecology 54: 111-117. Schroder, G. D., and K. H. Geluso. 1975. Spatial distribution of Dipodomys spectabilis mounds. J. Mammal. 56: 363-368. Schoener, T. 1974. Resource partitioning in ecological communities. Science 185: 27¬ 38. Whitford, W. G., P. Johnson, and J. Ramirez. 1976. Comparative ecology of the harvester ants Pogonomyrmex barbatus (F. Smith) and Pogonomyrmex rugosus (Emery). Ins. Soc. 23: 117-132. Wondolleck, J. T. 1978. Forage-area separa tion and overlap in heteromyid rodents. J. Mammal. 59: 510-518. BioScience Vol. 30 N o . 4