Estimating body size of extinct mammals presents problems when size can be estimated only by extrapolation. I examined the influence of phylogenetic, biomechanical, and statistical assumptions on body size estimates for 2 species of fossil castorids, the Pleistocene “giant” beaver Castoroides and the fossorial Miocene beaver Palaeocastor. Prior descriptions of Castoroides as “black-bear sized” were greatly exaggerated; new analyses estimated body masses at 60–100 kg. Estimates for Palaeocastor were similar to previously published estimates (0.8–1.2 kg). Biologically realistic size estimates were based on femur length and on interspecific data covering a wide range of body mass; skull length measurements or extrapolations from an ontogenetic single-species data set resulted in excessively large body mass estimates. Body mass must be estimated with due attention to the choice of both morphological trait and reference taxa.

Geometric descriptors were used to describe variation in cranial shape within and among populations of the echimyid rodent Thrichomys apereoides from localities in northeastern, central, and southeastern Brazil. Variation in shape based on dorsal, ventral, and lateral views of the skull, as inferred from contribution of uniform components and partial warps, was attributable to localized rather than global changes in morphology of the skull. Each view of the skull showed characteristic variability in shape among populations and provided different perspectives on patterns of geographic variation. The lateral view of the skull provided the most information content in terms of populations that share similarity in cranial shape and continuity over geographic space.

Geographic variation in cranial shape was examined in 20 samples of the echimyid rodent Thrichomys apereoides from northeastern, central, and southeastern Brazil. Canonical variate analysis and a dendrogram based on geometric shape descriptors revealed 2 groups of populations diagnosable by features of cranial shape and geographic continuity. The observed morphological discontinuity between the 2 groups of populations corresponds with a gap in sampling of T. apereoides, as represented by the museum series used. A positive, significant relationship between morphological and geographic distances was found among populations within each of the 2 groups but not between populations in the 2 groups. The indirect, statistical approach suggests that discontinuity in cranial shape in T. apereoides results from historical processes of differentiation rather than from an artifact caused by a gap in the available samples.

We analyzed phenotypic variation in white-eared Didelphis and its relevance to the current taxonomy of these opossums. On the basis of morphologic and morphometric data we show that the white-eared Didelphis can be split into 3 groups of populations referred to as albiventris, pernigra, and imperfecta. Morphologically, pernigra possesses a large, peg-like 3rd upper premolar (P3), whereas albiventris and imperfecta possess a small, non–peg-like P3. Also, albiventris has 88% of its specimens in the gray dorsal color phase, whereas the other 2 groups show similar percentages of specimens in the gray and black color phases. Morphometrically, pernigra is significantly larger (P < 0.05; analysis of variance) than the other 2 groups in most traits. Squared Mahalanobis distances among the 3 groups were found to be high, and discriminant analysis was successful in separating groups (99% of cases correctly classified). The congruence in variation of independent traits indicates genetic differentiation among populations. In addition, the sharp discontinuity observed between the parapatric albiventris and pernigra in Bolivia suggests that limited or no gene flow occurs between them. We conclude that the 3 groups analyzed here merit taxonomic recognition at the species level.

A sample of 244 ringed seal (Pusa hispida) skulls from 7 localities was examined morphometrically to describe the pattern of geographic variation and to assess subspecies-level taxonomy. Analyses of covariance revealed significant differences among specimens from Lake Saimaa, Lake Ladoga, and other areas. Saimaa seals are robust and short in the skull portion related to feeding, whereas Ladoga seals have narrower skulls and smaller bulla. On scatter plots of canonical discriminant scores obtained using size-free scores, Ladoga and Saimaa specimens were almost separated, but other specimens overlapped and could not be distinguished. Saimaa seals were the most distant, followed by Ladoga seals in the unweighted pair-group method using arithmetic average clustering and neighbor-joining phenograms. These results showed that although ringed seals from Lake Ladoga and Saimaa were considerably differentiated, specimens from other localities were not distinguishable from each other, suggesting similar selection pressure or extensive gene flow especially in the Arctic basin. Further recognition of subspecies for the Arctic seals was not supported.

Quill-withdrawal experiments confirm a hypothesis proposing that North American porcupines (Erethizon dorsatum) give up a quill more easily if the quill's root is first thrust back sharply into the porcupine's skin. The hypothesis was tested in 8 porcupines. When quills are impacted in a way that mimics contact with a predator, the tension required to remove an impacted quill from the porcupine's skin was reduced from 90.3 g ± 16.8 SD by 38.1%. The phenomenon was observed only in quills that had been erected by porcupines. This tension-reducing mechanism allows a porcupine to separate more readily from an attacking predator and adds to the effectiveness of porcupine quills as smart weapons.

The systematic relationship between Myotis lucifugus carissima and Myotis occultus has been the subject of multiple studies. Although several studies, including a recent allozyme study, concluded that M. occultus is a subspecies of M. lucifugus, this view has not been universally accepted. We reexamined the relationships of these 2 taxa by sequencing the mitochondrial cytochrome-b (cyt-b) and cytochrome oxidase II (COII) genes of specimens from the allozyme study. The results suggest that M. occultus represents an evolutionarily distinct monophyletic lineage and provide support to arguments that M. occultus is a distinct species.

Phylogenetic relationships of 11 species of cotton rats (Genus Sigmodon) were assessed using nucleotide variation in the mitochondrial cytochrome-b gene. Results of the phylogenetic analyses do not support the conventional composition of the Sigmodon hispidus and S. fulviventer species groups. Instead, 3 species groups are recognized: hispidus (S. alleni, S. arizonae, S. hirsutus, S. hispidus, S. mascotensis, S. ochrognathus, and S. toltecus), fulviventer (S. leucotis, S. fulviventer, and S. peruanus), and alstoni (S. alstoni). Sequence divergence values ranged from 8.7% to 21.2%, indicating that divergence events within the genus occurred approximately 2.5–5.6 × 10⁶ years ago. Taxa comprising the alstoni and fulviventer groups appear to have evolved early, whereas members of the hispidus group evolved more recently. Molecular data reflect an early origin (7 × 10⁶ years ago) and support the hypothesis that cotton rats may occupy a basal position among South American sigmodontines.

The taxonomic status of the genus Oxymycterus in Uruguay is assessed through morphologic, morphometric, and molecular comparisons with other species in the genus. Our analyses uncover the existence of a previously undescribed species, described herein, and restrict O. nasutus to SE Uruguay and SE Brazil. Uruguayan species are morphologically similar but differ substantially in sequences of the mitochondrial cytochrome-b gene. In contrast, O. rufus was found to be morphologically distinct from O. nasutus and the new species but genetically similar to the new species. Phylogenetic analyses lack resolution at the base of the genus but suggest well-supported relationships between O. amazonicus and O. delator, O. quaestor and O. judex, and O. rufus and the new species.

Bullimus is 1 of 15 genera of murid rodents endemic to the oceanic portion of the Philippines (i.e., excluding Palawan and other islands on the Sunda Shelf). Two species have been recognized previously: B. bagobus, widespread on Mindanao and on islands that were connected to it during Pleistocene periods of lower sea level and B. luzonicus, occurring only on Luzon. Recent surveys have revealed a 3rd species endemic to the small island of Camiguin, located just north of Mindanao and isolated from that large island by deep water. Multivariate analyses support the recognition of these 3 species. The new species of Bullimus from Camiguin is distinguishable from congeners by its smaller size, soft and uniformly dark pelage, inflated braincase, shorter palate and incisive foramina, more divergent molar toothrows, and other cranial and dental features. The patterns of variation and island distribution in Bullimus are consistent with our understanding of regional geography; the 3 species occur on separate islands or island groups that remained isolated during periods of lower sea level. The discovery of a species endemic to Camiguin reflects the fact that this small island is close enough to Mindanao to have received some nonvolant mammals by overwater dispersal, but its isolation has been sufficient to promote speciation.

Australian brush-tailed rock-wallabies (Petrogale penicillata) became naturalized on Oahu, Hawaii, after the accidental liberation of 2 animals (of unknown wild provenance) in 1916. Subsequently, Australian P. penicillata has experienced widespread local population extinctions as part of an ongoing decline. Thus, the P. penicillata population in Hawaii could represent an important conservation resource. In addition, there have been suggestions that the rock-wallaby population in Hawaii is distinct from Australian P. penicillata. In this study a multifaceted molecular genetic analysis (chromosomes, allozymes, mitochondrial DNA [mtDNA] control region) was undertaken to compare Hawaiian Petrogale with P. pencillata from throughout the species' natural range in southeastern Australia. Hawaiian Petrogale were found to possess a 2n = 22, all acrocentric, karyotype characteristic of P. penicillata. The allozyme (41 loci) and mtDNA profiles of the Hawaiian animals indicate that they were originally derived from northern P. penicillata populations, most likely in southeast Queensland. These results confirm that the Petrogale in Hawaii represent typical P. pencillata and are not consistent with suggestions that the Hawaiian population represents a novel taxon.

In this paper we test the hypothesis that bats of the Andean highlands show distinctive metabolic responses compared with bats from lowland forests. We compared existing literature with new information on 3 bat species having the following food habits: a nectarivore (Anoura latidens), a frugivore (Sturnira erythromos), and an insectivore (Tadarida brasiliensis). Basal metabolic rate, as determined by oxygen consumption, thermal conductance, and body temperature were measured at ambient temperatures of 10–38°C. Some distinctive metabolic responses of these bat species, although varying with respect to food guild, allow us to separate them from counterpart species that are typically found in lowland forests. A. latidens is characterized by higher basal metabolic rate; however, thermal conductance and lower critical temperature values do not show an adaptation to cool environments, as expected. S. erythromos also increases its basal metabolic rate, but it maintains thermal conductance as expected, which implies a very important displacement of thermoneutral zone to lower temperatures. At temperatures below lower critical temperature, in addition to an endothermic response, S. erythromos sometimes expresses a hypothermic response or facultative torpor, independent of sex and body mass. T. brasiliensis has a lower basal metabolic rate and thermal conductance and also has its thermoneutral zone range displaced toward lower temperatures. Likewise, this species enters obligate torpor when ambient temperatures are below 22°C.

Oxygen consumption, rectal temperature, and plasma concentrations of glucocorticoids, free fatty acids, and glucose were studied in 5 species of rodents after cold exposure that exceeded their homeostatic capability (7°C for 15 min). Adrenocortical responses and metabolic responses (oxygen consumption) to cold were large in 3 solitary, terrestrial species, dwarf hamsters (Phodopus sungorus, P. campbelli, and P. roborovskii) and were minimal in a colonial fossorial species, mole voles (Ellobius talpinus). Bank voles (Clethrionomys glareolus) showed intermediate responses. There were no interspecific differences in mobilization of main bioenergetic substrates (glucose and free fatty acids). Results are interpreted in terms of evolutionary changes of stress: reactivity was positively correlated with the rate of social stimuli and negatively correlated with physically stressful stimuli in the natural environment of the species.

To investigate seasonal tactics of energy management, core body temperature (T_b) was monitored over 1 year in eastern gray squirrels (Sciurus carolinensis) maintained on a constant diet in an outdoor enclosure. Each fortnight, we recorded hourly temperatures via radio-telemetry on 2 separate days. We also measured T_b during 10-day periods of food deprivation in winter and summer. In comparison with summer and autumn values, T_b was 1–4°C lower at all times of the daily cycle throughout winter and spring. Night and morning T_b was reduced during food deprivation, and reductions during the winter deprivation (1–3°C) generally exceeded reductions during the summer deprivation (0–2°C). Maximum sustained T_b and activity levels occurred during late afternoon or early evening hours year-round, however, and T_b at this time of day during deprivations was indistinguishable from T_b under ad-lib feeding conditions in the same season. Seasonal change in T_b, including reduction in response to food deprivation, contributes to metabolic strategies employed by eastern gray squirrels, which resemble those of other tree squirrels. Further research on mechanisms of temperature adjustment, including photoperiodism, hormones, and shifting dietary preferences, would help to illuminate the evolution of metabolic strategies among Sciuridae and other mammals.

This study examined environmental and physiological mediators of responses to hantavirus in wild-caught Norway rats. Rats were trapped in Baltimore, Maryland from 1996 to 2000, and prevalence of Seoul virus in target tissues, antibody responses against Seoul virus, concentrations of steroid hormones, and masses of reproductive organs were measured. Approximately 50% of live-trapped rats were infected with Seoul virus. No seasonal or photoperiodic changes in prevalence of Seoul virus infection or in antibody responses against Seoul virus were observed. Animals trapped during low temperatures had higher antibody responses than animals trapped during mild temperatures. Adults were more likely to have Seoul virus in kidneys and lungs and have higher antibody responses than juveniles. Older adult rats also were more likely to be infected and have higher antibody responses than younger adults. Rates of Seoul virus infection were similar between sexes; females, however, had higher anti–Seoul virus IgG responses than males. Although age- and season-related changes in concentrations of steroid hormones were observed, concentrations were not related to variation in infection by Seoul virus. Finally, the masses of reproductive organs increased with age, were responsive to environmental changes, and may be related to variation in Seoul virus infection. Thus, physiological and behavioral changes associated with sexual maturation may be the best predictors of Seoul virus infection in Norway rats, suggesting that metabolic trade-offs among growth, reproduction, and immune function occur at the onset of puberty.

The Mexican free-tailed bat (Tadarida brasiliensis mexicana) is primarily a nonhibernating, migratory, spring-breeding subspecies. Our study considers the anatomical structure, function, and seasonal cyclicity of testes and accessory sex glands. Unlike temperate, hibernating, sperm-storing vespertilionid bats, Tadarida does not store sperm, and its spermatozoa differ from spermatozoa of species with prolonged sperm storage. Recrudescence of the male reproductive tract occurs in late winter and early spring. Mating behavior of males and function of secondary sex glands are in synchrony with an activation of the male reproductive tract (spermatogenesis). These events, which take place in the spring, during and immediately following northward migration, lead to insemination of estrogenic, receptive females.

We assessed reproductive patterns of coyotes (Canis latrans) from a 12-year data set involving 24 pairs of captive animals acquired from Latah County, Idaho. None of the females had placental scars (fetal implantation sites) at 1 year of age, but over 80% had placental scars by 2 years of age. The fraction with placental scars remained at 80–90% through age 9 years and then declined to <40% by age 12 years. Similarly, mean number of placental scars per female rose rapidly through 2–3 years of age, remained stable until age 8 years, and then progressively declined to a very low level by age 12 years. Numbers of pups observed were typically smaller than numbers of placental scars, with fewer placental scars represented by pups among younger females than older females.

The estuarine dolphin (Sotalia guianensis) is common along most of the Brazilian coast, but little is known about its reproduction. We analyzed the gonads of 50 individuals incidentally caught in fishing nets on the coast of Paraná, southern Brazil, between 1997 and 1999. Testes of adult males were ≤31.9 cm in length and ≤11.6 cm in width and ≤3.3% of the total body mass. Relative size of testes suggests a multimale mating system with sperm competition. Females ovulate in both ovaries, although the left ovary matures earlier than the right. Males reached sexual maturity at 7 years and at body lengths estimated at 170–175 cm. Females matured at 5–8 years of age and at body lengths of 164–169 cm. The reproductive cycle was estimated at 2 years, with no marked seasonality in ovulation or timing of birth. Gestation was about 12 months, fetal growth rate was 9 cm/month, and length at birth was estimated at 92.2 cm. Females older than 25 years had senescent ovaries.

Postnatal development of vocalizations and audiograms was investigated in 19 juvenile Carollia perspicillata from the day of birth (age 0 days) to 7 weeks. In both juvenile isolation calls and adult social calls the 1st harmonic (fundamental) contained the most energy. Two types of isolation calls could be separated in young bats: “single notes” with 1 pulse and “double notes” with 2 pulses. Young bats 1st emitted echolocation pulses when they started to fly at age 17–22 days. Duration of these pulses was on average 1.8 ± 0.6 ms (mean ± SD; n = 11) and decreased during development to an adult-like duration of 0.9 ± 0.1 ms (n = 15) at 7 weeks. Frequency ranges of harmonics of the calls did not change significantly with age, but energy distribution of the harmonics was altered. Higher harmonics became successively emphasized. Newborn C. perspicillata hears 68% of the adult frequency range. During postnatal development the auditory range expanded upward. The upper limit of hearing, based on characteristic frequencies of multineuron clusters within the auditory midbrain (inferior colliculus), shifted from 77 kHz at birth to 110 kHz at day 50. Neural thresholds decreased most for frequencies of 60–110 kHz, which are most important for echolocation. However, from birth on, bats were most sensitive to low frequencies of 10–30 kHz.

Male Saccopteryx bilineata produces complex audible vocalizations during the day while defending roost territories within the buttress cavities of trees. To determine the amount of variability in the vocal repertoire and how this variation is apportioned among geographic locations and individuals within colonies, we recorded daytime vocalizations of male S. bilineata at 4 colonies on the island of Trinidad. Using visual discrimination, we identified 21 simple syllables and 62 composite syllables produced by males on the island. Using syllable counts for each colony and a nonlinear regression model, we found that 1 colony had an estimated repertoire size much larger than the others. Using 1st syllable of a common call type, we found several acoustic characteristics that showed significant individual variation, and 1 of those features also exhibited differences between colonies that were located in close proximity and colonies that were geographically separated. The vocal repertoire of male S. bilineata is variable, with considerable potential for information transfer.

I compared number of ultrasonic calls infant prairie voles (Microtus ochrogaster) and montane voles (M. montanus) emit when isolated. Typically, murid infants produce ultrasonic calls when they cannot thermoregulate adequately. Infant prairie voles called much more than did infant montane voles (average of 954 versus 17 calls/20 min at ages 0–10 days, P < 0.001) and more than other murids. Prairie voles also differed from montane voles in calling at high rates from day of birth, which is not typical of murids. The 2 species developed thermoregulatory competency at the same ages, so differences in calling could not be explained by differences in ontogeny. I suggest that rates of calling are related to social organization for, unlike montane voles and most murid species, prairie voles display monogamous mating behavior and a high level of biparental care. High calling rates could be adaptive for infant prairie voles, as parents are near almost all the time and can retrieve them; in contrast, montane vole mothers are away from the nest much of the time, increasing the risk of calling infants being overheard by a predator. The high rate of calling from day of birth in prairie voles may relate to prairie voles being nipple-clingers and so being at higher risk than montane voles of being dropped outside the nest when very young.

The trill vocalization is given by eastern chipmunks (Tamias striatus) when being pursued by a predator. To examine the adaptive significance of this vocalization, we assessed the effects of sex, age, and distance from the burrow on the occurrence of the call. Chipmunks were released either 10 or 100 m from their burrow and chased by an experimenter to simulate a terrestrial predator. We found that adult females trill more often when released 10 m from their burrow (67% of 24 chases) than when released 100 m from their burrow (14% of 21 chases). Adult males, on the other hand, showed no significant difference in their probability of trilling based on the distance from their burrows (45% of 20 chases at 10 m and 55% of 20 chases at 100 m). Adult females trilled more than adult males when released 10 m from their burrow (67% compared with 45%) but less than males when released 100 m from their burrow (14% compared with 55%). Juvenile females only trilled when released 10 m from their burrows (45% of 11 chases), and juvenile males did not trill at all. We attributed sex, age, and location differences in giving the trill vocalization to the effects of kin selection on antipredator behavior.

Gliding behavior of Japanese giant flying squirrels, Petaurista leucogenys, was studied at Nara Park, Japan. We observed 150 glides. We were able to calculate glide ratios on 57 glides and airspeeds on 29 glides. Glide ratios (distance/[altitude lost]) averaged 1.87, and a glide ratio of 3–3.5 seems to represent an upper performance limit for P. leucogenys. Airspeeds (4.39–9.47 m/s) were substantially lower than reported in other studies, and glide angles were higher (17.74–34.99°). Aspect ratios of the animals in mid-glide averaged 1.42.

This study tested some predictions of the local-resource-competition hypothesis and local-resource-enhancement hypothesis of sex-biased maternal investment in a population of pigtail macaques (Macaca nemestrina leonina). Offspring sex ratios in high- and low-ranking matrilines were biased in opposite directions as predicted. Sex ratio in the middle-ranking matriline was female biased similar to the high-ranking matrilines but became increasingly male biased with maternal age, similar to the low-ranking matrilines. Interbirth intervals following surviving daughters were longer than those following sons, irrespective of rank, suggesting that daughters were more costly to rear than sons. Infant mortality was not significantly different for sons and daughters. The findings of this study are consistent with some, but not all, of the predictions of local-resource-competition and local-resource-enhancement hypotheses and suggest that middle-ranking mothers have specific investment strategies that change in relation to their age.

We studied similarities and differences in the qualities of individual odors across colonies and species of African eusocial mole rats using habituation techniques. Giant mole rats (Cryptomys mechowi) treated anogenital odors of 2 conspecifics from 1 colony as similar to each other and different from the odor of a conspecific from another colony, providing evidence for kinship odors. Subjects also discriminated between the subtle differences in the individual odors of familiar colony members. In another set of tests, mole rats from 3 species (C. mechowi, C. anselli, and C. kafuensis) treated anogenital odors of genetically closer heterospecifics as more similar to odors of conspecifics than to odors of less closely related heterospecifics. Thus, odor similarities paralleled genetic similarities as determined by the analysis of allozymic and karyotypic distances. The results demonstrate that similarities in individual odors covary with genetic similarities within colonies and across species. These similarities could provide a basis for different types of social recognition based on self-referent matching.

We examined diet composition, choice of foraging areas, and distribution of food (vegetation) of a high-elevation (3,800 m) population of yellow-bellied marmots (Marmota flaviventris) in the White Mountains of California. Marmots overwhelmingly ingested forbs over graminoids (grasses, sedges, and rushes) and particularly ingested clover (Trifolium andersonii) despite its rarity. Although suitable marmot habitat is continuous at this site, marmot food plants were distributed in clumps. Areas in which marmots concentrated their foraging contained a disproportionate amount of clover compared with its abundance in the study site. Possible explanations for the choice of forbs over graminoids included the higher water content of forbs. Clumped food distribution has implications for foraging strategies and social behavior.

We banded 29 adult females, 2 juveniles, and 1 adult male from a maternity colony of Indiana bats (Myotis sodalis) during 1995–1998. Four banded bats were later found hibernating in caves that were separated by ≤325 km, indicating that all members of a summer colony do not hibernate or mate in the same location. We recaptured 41% of adult females (12 bats) near the initial banding site in later years, and because of this strong interannual fidelity, we recommend that resource managers reevaluate policies that allow removal of roost trees during winter. Eleven of these 12 females were radiotracked initially, but all were reproductively active and had normal body masses in subsequent years, suggesting negligible, long-term effects of the radiotracking process.

We studied occupation and activity at a hibernaculum and a maternity cave used by the endangered Ozark big-eared bat (Corynorhinus townsendii ingens). Big-eared bats remained active throughout winter, with many individuals leaving and entering the hibernaculum during the first 2 h after sunset; however, no fresh guano was found on cave floors during winter. Big-eared bats used caves with the coldest internal temperatures in the region as hibernacula and also moved among caves throughout winter. The breakup of hibernating colonies in spring and the subsequent buildup of maternity colonies by adult females were gradual. Nightly activity patterns at the maternity cave changed throughout summer relative to parturition and maturation of young. Females returned to the maternity cave more frequently when their young were totally dependent on them for nourishment. As the young began foraging, females returned only once during the night, and eventually adults did not return to the cave until morning. Ozark big-eared bats emerged shortly after sunset, regardless of brightness at twilight or rainfall, but they emerged at a later time when the young could not fly.

We radiotracked 22 jackals, 11 Canis mesomelas and 11 Canis adustus, in Hwange, Zimbabwe, to test the hypotheses that habitat use would differ and that the larger C. adustus would displace the smaller C. mesomelas. C. mesomelas preferentially used grassland. C. adustus used woodland and scrub. Habitat use by C. adustus differed from allopatric populations in which this species uses grassland, the likely favored habitat for jackals. C. mesomelas was shown to aggressively displace C. adustus from grassland. Aggressive displacement of a larger species by a smaller species is an unusual and probably unique behavior in carnivores.

We studied the natural history and landscape-use patterns of the guigna, Oncifelis guigna, a small forest cat, on Isla Grande de Chiloé, Chile. We put radiocollars on 7 adult (5 males, 2 females) guignas to assess factors that affect persistence in a highly fragmented, human-dominated landscape and a contiguous forest with fewer human impacts. Conservation of guignas outside protected areas requires forested areas connected by corridors, containing free-ranging domestic fowl, and dispelling local myths that guignas are vampires.

We used trap associations, nest-site preferences, a food selection experiment, and measurements of nightly activity to describe resource partitioning by syntopic Peromyscus californicus and P. boylii in Monterey County, California. P. californicus and P. boylii were active over the same range of time during the night. P. boylii was caught frequently near Quercus agrifolia, nested under canopies with high densities of Q. agrifolia, and consumed more Q. agrifolia acorns than P. californicus. Relative to P. boylii, P. californicus was a habitat generalist that did not discriminate among canopy plants for nest sites and had a broad diet. At high densities of P. californicus, the 2 species were negatively numerically associated. Both interspecific interactions and habitat partitioning (canopy level), via the specialization of P. boylii on Q. agrifolia, appear to facilitate the coexistence of P. boylii and P. californicus.