To determine the relative value of different Conservation Reserve Program (CRP) plantings for breeding grassland and winter birds we measured vegetation structure, avian abundance and reproductive success, and estimated fecundity during 1993–1995 on CP1 (cool-season grass) and CP2 (warm-season grass) plantings in 16 fields in northern Missouri. CP1 fields had been planted to cool-season grasses or cool-season grass-legume mixtures and CP2 fields had been seeded with switchgrass (Panicum virgatum). Species richness, abundance and nesting success of grassland birds during the breeding season and total bird use in the winter did not differ between CPs. During the breeding season CP1 fields had higher abundances of grasshopper sparrow (Ammodramus savannarum), eastern meadowlark (Sturnella magna), Henslow's sparrow (Ammodramus henslowii) and American goldfinches (Carduelis tristis), whereas common yellowthroats (Geothlypis trichas) were more abundant in CP2 fields. Fecundity of dickcissels (Spiza americana) and nesting success and fecundity of red-winged blackbirds (Agelaius phoeniceus) were higher on CP2 than on CP1 habitat, but both CPs were likely sinks (λ < 1) for these species. Both CPs were likely source (λ > 1) habitat for grasshopper sparrows, whereas only CP1 habitat was likely a source for eastern meadowlarks and American goldfinches. In winter American goldfinches were more abundant in CP1 fields than CP2 fields. The shorter, more diverse, cool-season grass fields were equal or better habitat than taller, more vertically dense, switchgrass-dominated fields for grassland birds, including several species of high conservation concern. Single-species plantings of warm- or cool-season grasses should be avoided to increase the potential wildlife benefits of CRP and other grassland habitats.

Potential responses to human disturbance at breeding colonies of waterbirds include reproductive failure, population declines and displacement from activity areas. Several additional factors, including species interactions and environmental change, can either mask or intensify the effects of human activity. This study highlights the importance of considering these factors in concert with breeding biology when assessing the impacts of human disturbance on wildlife. We studied the effects of a Wildlife Viewing Area (WVA) at Chatfield State Recreation Area, Colorado, on a nesting colony of great blue herons (Ardea herodias) and double-crested cormorants (Phalacrocorax auritus). We stratified the colony's nest trees into near, middle and far areas relative to distances from the WVA and compared the distribution of nests, nesting and fledging success and breeding chronology among areas 2 y before and 2 y after construction of the WVA. We also evaluated whether adult nest attendance patterns and chick behavior differed relative to distance from the WVA. The number of active heron nests and nest success of herons declined during the study, but evidence that these declines were due solely to human disturbance is equivocal. These changes were most likely due to the interplay of habitat changes (loss of 14 of 31 original nest trees by windfall), acquisition of heron nests by cormorants and human disturbance. We found no evidence that cormorants were adversely affected by the WVA in distribution of nests, nesting and fledging success, breeding chronology, adult nest attendance or chick behaviors. Habitat changes and adverse weather contributed to nesting failures of cormorants.

We measured predation on artificial ground nests in 11 grassy fields ranging from 0.8–12.6 ha. We set out 12 nests in each field baited with two Japanese quail eggs at different distances (<10 m, 25 m, 50 m) from woody edges in early June, and another 12 nests baited with one quail and one zebra finch egg in each field in mid-July. Nests were monitored at 3-d intervals for 15 d. The proportion of nests in which quail eggs were depredated was 33% in June and 38% in July. Including predation on finch eggs raised the proportion of nests depredated to 78% in July, indicating that small-mouthed nest predators such as mice or shrews could account for a substantial amount of predation on nests of small passerine birds. Predation on nests was negatively related to area of field only for quail eggs in July (P = 0.04). We did not detect significant relationships between numbers of nests depredated or daily predation rates and distance to woody edges. The small size of fields in our study, which are typical of grassland patches in this region, may be below a threshold at which nests in the interior of patches experience reduced predation. Alternatively, highly fragmented natural habitats in an agricultural landscape may support high densities of predators, making edge and area effects difficult to detect.

We examined the influence of cover item type (rock, down wood, leaf litter) on capture, size (snout-vent length) and weight of two plethodontid salamander species, Desmognathus ochrophaeus Cope (Allegheny Mountain dusky salamander) and Plethodon cinereus (Green) (red-backed salamander), common in riparian forests of northwestern Pennsylvania. We wanted to determine if cover item use differed seasonally and/or by geomorphic surface within stream valleys and if cover item type and/or geomorphic surface influenced size and weight of salamanders. We also tested the hypothesis that larger salamanders are associated with larger cover items. Captures of D. ochrophaeus and P. cinereus differed significantly by season and were inversely related: captures of D. ochrophaeus peaked in the spring and midsummer; captures of P. cinereus peaked in the fall. Capture rates did not differ by geomorphic surface. Cover item use was similar for D. ochrophaeus and P. cinereus with most captures associated with rock or down wood. Size and weight of D. ochrophaeus did not differ significantly by cover item type but individuals captured on floodplains weighed significantly more than those captured on hillslopes. Size and weight of P. cinereus were significantly greater for individuals captured under rock and down wood than in leaf litter. Size and weight of both species were significantly positively correlated with cover item size only during a single season: spring for D. ochrophaeus and fall for P. cinereus, coincident with peak captures for each species.

Hydroperiod, the time a temporary pond holds water, is an important factor influencing recruitment in amphibian populations and structuring amphibian communities. We conducted an experiment to test the effect of hydroperiod on metamorphic traits of the southern leopard frog (Rana sphenocephala), a common amphibian in the southeastern United States. We reared larval R. sphenocephala in artificial ponds at a density of 32 larvae per tank (initial volume = approximately 650 liter). We dried the tanks according to natural patterns, using three different hydroperiods (60, 75 and 90 d). Experimental hydroperiods had a significant effect on the number of metamorphs and the length of the larval period, but not on overall survival (larvae metamorphs) nor size at metamorphosis. Our findings confirm a pattern observed in field studies and are similar to results of experimental investigations of closely related ranid frogs. Our results demonstrate that relatively small differences in hydroperiod length (i.e., as little as 15 d) may have large effects on juvenile recruitment in R. sphenocephala.

Beavers (Castor canadensis) create numerous scattered wetlands in the southeastern United States that alter the composition of aquatic and adjacent terrestrial communities. However, the influence of beaver ponds on communities of amphibians and reptiles (herpetofauna) is poorly known. We used drift fences, coverboards and aquatic traps to compare herpetofaunal communities from unimpounded streams (n = 3) with those of beaver ponds (n = 3) in the Piedmont of South Carolina during 1998 and 1999. We also characterized differences in environmental and upland habitat attributes between beaver ponds and unimpounded streams. There were no significant differences in overall herpetofaunal abundance between unimpounded streams and beaver ponds, although significantly more salamanders were captured at unimpounded streams and significantly more anurans, lizards and turtles were captured at beaver ponds. Estimates of amphibian and reptile species overlap were high for beaver ponds and unimpounded streams. However, the richness (S), diversity (H′) and evenness (J′) of amphibians were significantly higher at unimpounded streams than at beaver ponds. In contrast, the abundance, richness and diversity of reptiles were significantly higher at beaver impoundments. Differences in amphibian and reptile community attributes between beaver ponds were related to the lotic or lentic habitat requirements of individual species and the effects of beaver impoundments on surrounding terrestrial habitats. Our study indicates that natural disturbances resulting from beaver-created wetlands increase regional abundance and diversity of herpetofauna.

Intraseasonal nesting patters were determined for loggerhead sea turtle (Caretta caretta) nesting on Bald Head Island, North Carolina, one of the northernmost nesting concentrations along the east coast of North America. One hundred and twenty-seven nesting females were tagged with Inconel flipper tags and photographed, accounting for 83% of the 318 nests produced during the 1991 and 1992 nesting seasons. No female nested in both years. Nest production was similar both years. Most females (55.9%) nested once, 5.5% nested twice, 15.0% nested three times, 7.9% nested four times, 12.6% nested five times and 3.1% nested six times. The internesting interval averaged 13.7 d (range 10–18) in 1991 and 14.2 d (range 11–18) in 1992, and it was significantly influenced by seawater temperature. Intraseasonal nest site fidelity was signifcant; most multiple-nesting females (67.9%) confined their nesting activities within 4.8 km. Loggerheads in the northernmost part of their nesting range in the western North Atlantic exhibit intraseasonal nesting patterns similar to those of populations breeding elsewhere.

I studied variation in clutch size and ovum size of the snubnose darter Etheostoma simoterum from three sites in Bluewater Creek in southern Tennessee which I compared to published data for a population from Brush Creek in northern Tennessee. Females 29–44 mm standard length (SL) from Bluewater Creek (n = 54) produced 25–96 (x̄ = 63) eggs per clutch. SL-adjusted clutch sizes varied among the localities within Bluewater Creek with the largest mean clutch size being 1.14 times greater than the smallest. Mean diameter of unovulated ripe eggs in individual females ranged from 1.15–1.34 mm (x̄ = 1.24 mm diam) and mean egg mass ranged from 251–380 (x̄ = 312) μg. Both metrics showed significant correlations with SL. Females from Brush Creek produced significantly larger SL-adjusted clutches than females from Bluewater Creek, but egg diameters did not differ significantly between populations. Comparison of the two populations shows interesting geographic variation in clutch size that may result from racial or ecotypic differences. An alternative explanation is that food availability caused proximal nongenetic variation in clutch size.

We released bobcats (Lynx rufus) on Cumberland Island, Georgia during 1988 and 1989 as part of a cooperative effort to restore an extirpated predator to the island. We collected data on prey use and prey abundance three times a year during 2 y following the initial bobcat releases. We tested four hypotheses concerning bobcat prey selection: (1) use of a prey species was closely associated with its abundance (functional relationships), (2) the number of prey species included in bobcat diets increased as the abundance of principal prey species decreased (diet optimization), (3) the number of species included in bobcat diets and diet diversity increased as population density increased (interference) and (4) increases in the proportion of males resulted in increases in the use of large prey and decreases in the use of small prey. Bobcats' use of marsh rabbits (Sylvilagus palustris) and cotton rats (Sigmodon hispidus) was positively correlated with each species' abundance as predicted by a hypothesis of functional relationships. The number of prey species in bobcat diets and diet diversity was negatively correlated with the abundance of marsh rabbits as predicted by a hypothesis of diet optimization. Changes in diet diversity in different regions of the island between years were not correlated with changes in bobcat density within each region, suggesting interference was not occuring at the bobcat densities observed in our study. Changes in the use of Eastern grey squirrels (Sciurus carolinensis) between years were negatively correlated with increases in the proportions of male bobcats in different regions of the island. Our results suggest that diet optimization and functional responses are both useful models for describing bobcat-prey relationships. The diet optimization model had greater power to explain use of prey species that were not used in all seasons.

I determined cause-specific mortality of yellow-bellied marmots (Marmota flaviventris) to assess the importance of predation as a mortality factor. Contrary to earlier findings based on visual observation, almost all (98%) mortality during the summer active season was caused by predation. Coyotes (Canis latrans) were the most important predators, followed by badgers (Taxidea taxus), American martens (Martes americana), black bears (Ursus americanus) and raptors, probably golden eagles (Aguila chrysaetos). Predation on marmots is cryptic; none of the predation events were observed and, in most cases, the marmot was consumed or removed so quickly and completely that searching for carcass remains would have been fruitless.

Food habits of swift foxes (Vulpes velox) occupying two distinct landscapes (dominated by cropland versus rangeland) in western Kansas were determined by analysis of scats collected in 1993 and 1996. Frequencies of occurrence of prey items in scats were compared between cropland and rangeland areas by season. Overall, the most frequently occurring foods of swift foxes were mammals (92% of all scats) and arthropods (87%), followed by birds (24%), carrion (23%), plants (15%) and reptiles (4%). No differences were detected between landscapes for occurrence of mammals, arthropods or carrion in any season (P ≥ 0.100). Plants, specifically commercial sunflower seeds, were consumed more frequently in cropland than in rangeland in spring (P = 0.004) and fall (P = 0.001). Birds were more common in the swift fox diet in cropland than in rangeland during the fall (P = 0.008), whereas reptiles occurred more frequently in the diet in rangeland than in cropland during spring (P = 0.042). Variation in the diet of the swift fox between areas was most likely due to its opportunistic foraging behavior, resulting in a diet that closely links prey use with availability.

We radio-tracked 4 solitary subadult (2 male and 2 females) and 3 adult female evening bats (Nycticeius humeralis) to 7 and 14 roost trees, respectively, during the summer of 1997 on the Upper Coastal Plain of South Carolina. Roosts for both age-classes were located in cavities and under exfoliating bark in live trees and dead snags. Based on nightly exit counts, adult female evening bats roosted communally. Some roosts were presumed to be active maternity colonies. Five of six subadult roosts and one adult female roost were located under exfoliating bark on dead snags in beaver (Castor canadensis) ponds. Thirteen of 14 adult female evening bat roosts were located in cavities in live longleaf pines (Pinus palustris) in stands of relatively mature, park-like, pine plantations. One subadult roost was located under exfoliating bark in a live longleaf pine. No evening bats roosted in the more densely canopied, nearby bottomland hardwood stands, mixed pine-hardwood stands or loblolly pine (P. taeda) stands. Roost fidelity by individual evening bats was short (2.3 roost-days per bat) and did not differ between subadults and adult females. Although average daily maximum roost temperature was higher in subadult roosts than in adult female roosts, internal average temperature, daily minimum temperature and daily temperature range did not differ significantly between the two roost categories. Overstory height surrounding roost trees, canopy density and basal area of pine in the surrounding stand were greater at roosts within longleaf pine stands than those in beaver ponds. Conversely, stand basal area, overstory tree species richness, abundance of Spanish moss (Tillandsia usneoides) and basal area of hardwood trees were greater at roost sites within beaver ponds than those same measures in longleaf pine stands.

To investigate natural roost-site selection by eastern red bats (Lasiurus borealis) in an urban area we equipped 12 bats with radio-transmitters and located them at 75 individual roost sites in central Illinois from 16 July to 30 September 1996. Bats used a diversity of roosts including mature trees, leaf litter, dense grass and the shingles of houses, but the majority of roosts (89%) were in the foliage or on the trunks of large deciduous trees (>45 cm dbh). Sweetgum (Liquidambar syraciflua) and oaks (Quercus spp.) were commonly used. Most roosts were located in foliage, >5 m above the ground, within 1.5 m of the edge of the crown and with few branches beneath to obstruct flight paths. Red bats showed fidelity to roost sites within a small geographic area, but not to particular roosts. Individuals rarely used the same roost on consecutive days, but 82% of roosts used on successive days were within 100 m of each other. In the cornbelt region of the Midwest, where forests have been extensively cleared and fragmented, large urban trees provide important roosting habitat for red bats.

We estimated seasonal and annual survival and cause-specific mortality of 132 cottontail rabbits (Sylvilagus floridanus) in east-central Mississippi during 2 y. Daily survival did not differ between years for genders, but differed significantly between breeding (99.22%) and nonbreeding seasons (99.76%). Mammalian (54.8%) and avian (26.1%) cause-specific mortalities were major mortality agents. Mammalian predation was similar between genders during the breeding season (male: 51.8%; female: 59.4%). Males experienced greater avian predation during the breeding season (male: 33.7%; female: 18.8%). Mammalian predation (male: 12.8%; female: 25.1%) and harvest (male: 25.6%; female: 12.5%) did not differ between genders for the nonbreeding season. Daily mammalian (breeding: 0.45%; nonbreeding: 0.10%) and avian (breeding: 0.22%; nonbreeding: 0.03%) predation differed between seasons. Cottontails preyed upon had greater movement rates the week before death than those that survived the same week. Cottontails that experienced avian predation had larger core areas up to day of death than those that did not experience avian predation whereas home-range size did not differ. Home-range and core-area sizes did not differ between rabbits that experienced mammalian predation and those that did not. From four previous survival studies and ours, cottontails exhibited latitudinal variation in survival (possibly due to climatic factors, mammalian or predator population levels). Southern (Mississippi and South Carolina) populations exhibited survival less than northern (Wisconsin and Illinois) populations in the breeding season and greater survival during the nonbreeding season than northern populations.

The influence of soil waterlogging in determining plant species distributions among interdunal swales and between dunes and swales on Cape Cod was investigated employing field and greenhouse experiments. Most species exhibited significant responses to experimental treatments that closely matched their distributions in the field. In general dune species such as Ammophila breviligulata and Solidago sempervirens grew poorly under waterlogged conditions. When transplanted to the dry sand dune all species grew poorly or died. Swale species exhibited a range of responses to waterlogging conditions. Growth of Juncus canadensis, Scirpus pungens and Vaccinium macrocarpon increased with waterlogging. Cyperus dentatus, Juncus pelocarpus and Solidago tenuifolia grew equally well across all waterlogging conditions and are present in most swales. Juncus greenei and Myrica pensylvanica grew best in drier swales and their distributions appear to be restricted to the perimeters of swales.

Soil waterlogging is an important determinant of species distributions among swales and between swales and dunes. However, the effect of biotic factors such as competition, seed dispersal, soil seed banks and seedling recruitment needs to be examined as well to show how swale communities change over time.

We evaluated effects of small disturbances that kill parts of individual plants on plant survival by measuring tiller survival for the perennial bunchgrass, Bouteloua gracilis (H.B.K.) Lag ex Griffiths (blue grama). The importance of soil texture, grazing by cattle, disturbance type and severity were evaluated. Two disturbance types (covering or removing tillers) and three disturbance severities (50, 75 and 90% tiller mortality) were used to represent effects of natural disturbances in shortgrass communities (cattle fecal pats, nest sites of Western harvester ants, burrows of small animals).

Tiller survival was not affected by soil texture or grazing intensity, but was affected by disturbance type and severity. Plants that were covered showed a 33% increase in tiller survival for all levels of disturbance severity from August (1991) to June (1992). No net change in tiller number was observed for removed or reference plants. Different responses between disturbance types were likely due to increases in root:shoot ratios of covered plants that increased tiller production as a result of increased soil water acquisition. The number of tillers produced was small, but statistically significant (average = 20 tillers/plant), which shows that B. gracilis plants do not produce independent tillers, but consist of integrated physiological units (IPUs). The lack of plant mortality, even with 90% tiller mortality, indicates that small disturbances must kill entire plants before gaps in resource space are produced to initiate gap dynamics that result in the recovery of an individual B. gracilis plant. Because recovery through seedling establishment by B. gracilis occurs infrequently, the ability of this species to survive after partial plant mortality is important to its continued dominance of shortgrass steppe communities in the presence of these small but frequent disturbances.

The availability and distribution of essential resources such as water and nitrogen are changing on a global scale and it is important to determine the effects of such alterations on dominant and subdominant species in ecosystems. We examined the effects of long-term additions of water and nitrogen on leaf-level ecophysiological characteristics of the tallgrass prairie codominants Andropogon gerardii (big bluestem) and Sorghastrum nutans (Indian grass). Both species responded to irrigation with increased water potentials and to fertilization with decreased water potentials. Only in S. nutans, however, were net photosynthesis and stomatal conductance greater in response to both water and nitrogen addition. In contrast, A. gerardii, which currently is more abundant than S. nutans, was relatively unresponsive to resource manipulations. If net photosynthesis is positively related to plant productivity, as has been shown for tallgrass prairie species, the differential photosynthetic responses of these two grasses suggests that their patterns of abundance may change in the future if resource availability is altered. These results are also inconsistent with the historical assumption that these two grasses are “ecological equivalents.”

The 70% decline of the Federally threatened eastern prairie fringed orchid, Platanthera leucophaea (Nuttall) Lindley, has prompted concern for its recovery through artificial propagation. We describe a technique to germinate seeds and cultivate seedlings of P. leucophaea in vitro using cold treatments (=stratification) and mycorrhizal fungi (=symbiotic seed germination). Five fungal isolates were recovered from mature P. leucophaea plants in Illinois and Michigan and were identified as members of the anamorphic genus Ceratorhiza Moore. Stratified seeds inoculated with mycorrhizal fungi germinated within 25 d of sowing. Leaf-bearing seedlings were obtained by chilling young seedlings (protocorms) for 107 d. Our successful culture of leaf-bearing seedlings with a presumed mycotrophic capability may make it possible for this threatened orchid to be propagated in soil ex vitro, followed by reintroduction into suitable habitats.

The 1998 emergence of 17-y periodical cicadas (Magicicada cassini) on Konza Prairie Research Natural Area (KPRNA), Kansas, was quantified using emergence trap transects and counts of emergence holes. Emergence density, biomass (emergence production) and associated nitrogen flux were estimated for the entire 100 ha gallery forest of Kings Creek, the major drainage network on KPRNA. Emergence commenced on 22 May 1998 and lasted for 24 d, with 87% of the individuals emerging within the first 9 d. Males dominated early during the emergence, and the sex ratio for the entire population was estimated at 54:46 male:female. Average emergence abundance and biomass estimated from trap transects located in low areas where cicadas were most abundant were 152/m² and 34.9 g ash-free dry mass (AFDM)/m², respectively. Based on emergence hole counts, average density and biomass for the 59 ha of gallery forest where cicadas emerged were 27.2 individuals/m² and 6.3 g AFDM/m², and emergence hole densities >100/m² were evident in low areas of the drainage. Emergence density generally decreased with increasing elevation in the catchment. Belowground to aboveground N flux associated with M. cassini emergence in high density areas was ∼3 g N/m², and the average for the entire emergence area was 0.63 g N/m². The total number of individuals that emerged from the Kings Creek riparian forest was estimated at 19.6 million, which represents 4.6 metric tons AFDM and ∼0.5 metric tons N. This linear, fragmented, gallery forest of the Flint Hills supports a high density of M. cassini, and an emergence event constitutes a significant belowground to aboveground flux of energy and nutrients. Thus, the periodical cicada may be an exception to the notion that insects generally do not represent important resource pools at the ecosystem level.

The effect of the matrix within which habitat fragmentation occurs was studied with respect to gypsy moth (Lymantria dispar L.) egg-mass number on the E. S. George Reserve in Livingston Co., Michigan, in October 1999. Overgrown old fields were assumed to be relatively high quality matrix compared to a large herbaceous swamp. Oak forest fragments were located in each matrix type as well as in a reference matrix of extensive oak forest. The mean number of egg masses in stands of black oaks (Quercus velutina) was determined for fragments in high and low quality matrices. A significant difference in gypsy moth egg-mass number between areas of high vs. low quality matrix was detected.

The larval period of the three-lined salamander was studied over an elevational gradient in the Chattooga River watershed of Georgia, North Carolina and South Carolina. Hatching, in late winter and early spring, occurred earlier at lower elevations of the watershed. Most larvae in populations at lower elevations metamorphosed in the first year, 5–6 mo after hatching, whereas those in higher elevation populations overwintered and metamorphosed early in their second year after a 14–15 mo larval period. Contrary to expectation, we found no definitive evidence that this difference stemmed from variation in first-year larval growth. Moreover, the larval habitats (bogs, pools along sluggish streams) showed no variation in temperature associated with elevation. However, mid to late summer drying occurred in both lower elevation habitats, but not in either higher site. We propose that the observed variation in larval period reflected the norm of reaction of Eurycea guttolineata to the degree of permanency of larval habitats.

The effects of leg autotomy on running speed and foraging ability were examined in two species of wolf spider (Schizocosa ocreata and Varacosa terricola). Both species run slower following leg autotomy. However, foraging ability was not affected in either species. Since spiders with one leg autotomized are slower than spiders with all legs intact, they may be more susceptible to predation.

Of 138 wild mink (Mustela vison) from eastern Minnesota, 27% contained Dioctophyma renale, primarily in the right kidney. No significant difference between prevalence in adult male and immature male mink was found, nor between the prevalence in males vs. female mink. Thirteen worms were found in one male mink, representing the highest documented infection intensity of a single wild mink.

In December 1999 a female hoary bat (Lasiurus cinereus) was found in Ontario. The bat, which was kept in captivity from then on, showed signs of distress resembling abortion in late April 2000. Autopsy revealed the presence of two foetuses in the uterus, a male at the latest stage of development and a smaller dead one. The latter stopped development at an earlier stage and was most likely being resorbed, probably due to deficiencies in the female's diet. Mating leading to this pregnancy must have occurred in autumn, likely before the onset of migration.

We examined seasonal patterns in lepidopteran prey of Rafinesque's big-eared bat (Corynorhinus rafinesquii) by collecting culled moth wings from March to September 1998 from the floor of the entrance room to a sandstone cave used as a roosting site by a maternity colony. We identified 22 species and 6 families (Arctiidae, Geometridae, Megalopygidae, Noctuidae, Notodontidae and Sphingidae) in the diet of C. rafinesquii from 135 identifiable wings recovered from the cave floor. Twelve wings comprising >2 species of moths of the family Arctiidae were recovered, indicating that C. rafinesquii periodically preys on arctiids. Abundance of culled wings was highest from 24 May to 25 July. Mean wingspan of moth prey did not vary across seasons. We observed C. rafinesquii feeding at the entrance to the roost by gleaning moths from the walls and ceiling of the cave and by capturing moths directly in flight, confirming that this species does use more than one foraging strategy in capturing moth prey.