Predators of unionoid mussels are generally identified as fishes or aquatic mammals. During a mark and recapture study of the critically endangered mussel Popenaias popeii, we discovered a nymph of the dragonfly Gomphus militaris eating the gills of a gravid mussel; larvae and gill material were found in the nymph's gut. Many (15.2%) of the other mussels captured during a quantitative survey exhibited damage consistent with that inflicted by this dragonfly. Few non-gravid mussels were damaged and gravid mussels exhibited substantially more damage in gills used for brooding larvae than in gills not typically used for brooding. This previously unreported parasitic relationship may reflect a unique cost associated with reproduction and should be considered in the development of conservation strategies for P. popeii. Our observations underscore the need for basic ecological data when monitoring endangered species.

Promethea (Callosamia promethea) and tulip tree silk moths (C. angulifera) were compared under semi-natural conditions for the presence of polyandry. Promethea were polyandrous, the first documentation for a saturniid moth, whereas tulip tree moths were monandrous. Experiments showed that polyandrous and monandrous promethea females achieved similar egg fertility, but polyandrous females laid 10% more eggs, a significant difference (P < 0.05). This difference was found in five sibling groups, whose larvae were reared on the same individual food plant. Higher fecundity for polyandrous promethea females was not related to female size, duration of copulation, egg size and number, number of days in laying period or sibling group. Egg size declined later in the laying period for all females. Small females laid more eggs earlier in the laying period than large females. A separate mark/recapture study showed that male promethea also mate multiple times (polygyny) and distinguish virgin from nonvirgin females in mating preference. The diurnally mating promethea may gain increased genetic variability and/or possibly seminal gifts from polyandry. Monandry in totally nocturnal saturniid moths may result from a time conflict between egg laying and mating, which overrides the benefits of polyandry. If polyandry increases fecundity, it is predicted to occur in other diurnally mating saturniids.

A current challenge for ecologists is identifying the relative effect of factors affecting abundance in spatially segregated populations. I investigated the effects of variables affecting dispersal (current patch size, interpatch distance and patch quality) and reproduction (previous female abundance, previous patch size and their interaction) on the local abundance of an herbivorous beetle, Tetraopes tetrophthalmus. Mark-release-recapture methods were used to monitor beetle populations in a 40 ha landscape containing numerous patches of the beetles' host plant, Asclepias syriaca. Factors affecting dispersal and reproduction were compared using statistical models, and the relative effects of each factor were evaluated by examining the joint and unique deviance explained. Over 3 y, 3952 individuals were captured 10,715 times. The models explained over 80% of the deviance in beetle abundance with significant contributions from factors affecting both dispersal and reproduction. Current patch size and number of females in a patch the previous generation had the largest effects on abundance. Immigration exceeds emigration in larger patches resulting in higher abundance in larger patches, but a large effect of current patch size was also due to confounding with factors affecting reproduction. Abundance also increased with habitat quality and connectivity through their effects on dispersal. The abundance of females in the previous generation affected current abundance directly and via interactions with resources. In a low density year (1995) there was a positive interaction between female abundance and patch size (resources) such that large patches with an abundance of females had high abundances the next generation (1996). In a higher density year (1996) abundance the next generation increased with female abundance, but showed either a negative interaction with patch size (male beetle abundance) or a negative effect of patch size in 1996 (female). For patchy populations a combination of landscape and population factors are needed to understand local abundance.

The Washington ground squirrel (Urocitellus washingtoni) is listed as threatened in Oregon and is a species of special concern in Washington. We investigated abundance and habitat selection of the species on the Boardman Bombing Range, Morrow County, Oregon during 1996–1997. Transect surveys were used to determine relative abundance among seven vegetative types, and mark-recapture methods were used to compare density of squirrels among habitats where they were present. Transect surveys indicated highest abundance in sagebrush, followed by perennial grassland then annual grassland; and no squirrels were detected in bitterbrush or low-shrub habitats. Results from live-trapping and capture-recapture methods also indicated higher densities in sagebrush habitat. Mean maximum distance moved on trapping grids was greater for males than females for adults and juveniles alike. Macro-habitat analyses revealed a selection for areas with higher silt content (Warden soils) in soils and higher vegetative cover. Micro-habitat associations indicated a selection for areas with lower clay content in soils. The continued existence of Washington ground squirrels depends upon maintenance of the remaining suitable habitat, particularly large tracts of sagebrush and perennial bunchgrass on Warden soils.

We collected data on the diet of eight species of insectivorous bats (Chiroptera, Vespertilionidae): big brown bats (Eptesicus fuscus), red bats (Lasiurus borealis), evening bats (Nycticeius humeralis), northern myotis (Myotis septentrionalis), little brown myotis (M. lucifugus), Indiana myotis (M. sodalis), southeastern myotis (M. austroriparius) and eastern pipistrelles (Perimyotis subflavus). Bats were mist netted during the summers of 1999 and 2000 at 41 forest sites throughout southern Illinois. We analyzed prey remains in fecal pellets of 305 individuals to assess diet similarity among species and relationships between bat body mass and prey diversity and hardness. Larger species included big brown bats and evening bats that ate primarily hard-bodied beetles (Coleoptera). These bats had the greatest dietary similarity index value compared with the other chiropterans in the community, and the highest hardness indices of prey consumed. Red bats, second only to E. fuscus in mean body mass, ate more soft bodied moths (Lepidoptera) and leaf hoppers (Homoptera) than beetles. Smaller bats, including three species of Myotis, consumed the greatest diversity of prey and were generally grouped together in diet similarity indices. Little brown myotis, northern myotis and Indiana myotis fed primarily on moths and beetles. Northern myotis and little brown myotis also fed extensively on spiders, suggesting significant gleaning behavior. Unlike other Myotis, the southeastern myotis had a low dietary diversity index and fed primarily on caddisflies (Trichoptera), as did eastern pipistrelles. Pipistrelles and myotines had the lowest hardness indices of prey consumed. Bats in southern Illinois exhibited landscape level (macroscale) feeding patterns consistent with the predicted relationship between body size and hardness of prey consumed, while at the local, site-specific level (microscale) they foraged with extensive overlap among similar-sized species, especially most Myotis. Regional differences in diets were minimal within the same assemblage of bat species in southern Indiana.

Over the past 20–30 y, northern and western populations of the Allegheny woodrat (Neotoma magister) have experienced large declines, whereas populations in the core of the range are assumed to be stable. We examined population trends at two study areas in northcentral West Virginia along the western ridge of the central Appalachian Mountains. Temperature and precipitation parameters along with mast production were examined to determine if these environmental variables impacted the population. Based on a 5 y dataset, our results indicate a yearly decrease in the overall population, with adult females most affected. Hard and soft mast availability related to adult female capture rates, whereas temperatures significantly affected juvenile, adult female and overall capture rates. Juvenile summer capture rates increased with warmer temperatures the preceding winter. Female summer capture rates decreased with warmer temperatures the preceding spring suggesting that effects of warming should be added as a potential threat to the Allegheny woodrat.

Beavers (Castor canadensis) are important as ecosystem engineers and are useful model organisms for testing central-place foraging theory. Much previous work has done this in controlled situations, whereas here we tested predictions in a complex natural habitat by collecting data on beaver-cut and uncut trees in sites at various distances from an isolated beaver lodge on a large reservoir. The most basic theoretical predictions are that selectivity of predators should increase with distance from the central place and that preferred prey size should increase with distance. Consistent with the first prediction, beavers were more selective for preferred tree species when foraging far from shore or at sites some distance from their lodge than when near the shore or at sites near their lodge. Consistent with both predictions, beavers were more selective for particular sizes of trees and selected larger sizes of trees as distance from the shore increased. Overall, beavers showed a preference for intermediate tree sizes, avoiding both very large and very small trees, but as distance from shore increased, beavers cut fewer trees from the smaller end of this size range and more from the upper end in such a manner as to increase both selectivity and mean size. Similarly and within this same size range, as distance from the lodge through the water increased, beavers cut larger trees at greater distances due to reduced cutting of small trees and increased cutting of larger ones. In this case, however, overall selectivity did not increase, just an increase in the size selected. Overall, then, our study shows that central-place foraging theory can predict the behavior of beavers foraging in a complex natural landscape, and the patterns observed have implications for how beavers might influence tree species composition in forests.

North American beavers alter habitat properties such as stream flow, faunal composition and plant community structure. In North America, the majority of studies on impacts by beavers have occurred in the northeast or western regions. This study examined the effect of wood-cutting and foraging by beaver on the surrounding plant community in Southeastern Georgia. To determine their impact, beaver modified habitat and nearby, but non-impacted riparian habitat, were sampled in a matched pair design. Vegetation surveys were performed, and species richness, density, relative abundance and canopy cover were calculated for every site. Richness of herbaceous vegetation was higher at distances closer to shore while richness of large woody vegetation increased with distance from shore with no difference found between beaver and control sites. The composition of herbaceous or woody vegetation did not differ significantly between beaver and control sites; species had an equal chance of occurring in either site. Canopy cover was similar at the two site types. Non-native species were rare and did not differ by site type. Beavers were in low density and had a relatively benign impact on the plant community. Density of woody plants was higher at beaver sites, suggesting that beavers may favor areas of abundant woody vegetation. Beaver impact may be less in southern regions than in northern ones. This may reflect the lower density of beavers, but impact also could be due to the year round availability of and access to vegetation.

An examination of the secondary sex characteristics of 3197 adult specimens of Sorex (Soricomorpha, Soricidae), including 50 species, processed by 560 preparators over 126 y, and housed in 18 mammal collections, revealed that the sex of 9.57% of these specimens was misidentified or not identified. If only the 20 species for which I examined ≥20 specimens are considered, the sex of 5.82% of males and 15.89% of females was misidentified or not identified. Diverse methods of determining sex are available, although their use necessarily depends on the age of animals and whether they are dead or alive. For live shrews, examination of the secondary sex characteristics, distance between the urogenital and excretory openings, and presence or absence of sex chromatin can be used to identify the sex. For dead shrews, sex also can be determined by dissection of reproductive systems, examination of pelvic bones and determination of karyotypes. Of genetic methods, the Sry gene is the best characterized for identification of the sex of live or dead shrews. Several techniques are described for determining sex of specimens of shrews, however, due diligence must be observed by application of the appropriate technique(s) within the capabilities of specimen collectors or preparators that will provide the correct results.

Northern flying squirrels (Glaucomys sabrinus) of the Black Hills National Forest (BHNF) of South Dakota represent a unique and isolated population, but little is known about the nesting ecology of this population. We radio-collared 59 northern flying squirrels and collected a daytime nest location every 2–4 wk during May through Aug., 2005 through 2007. The radio-collared northern flying squirrels used 133 different nests, including drays in live trees, cavities in live trees and cavities in snags. We examined distance between consecutively located nests and characteristics of nest trees to random and available trees within the northern flying squirrels' home ranges. The distance between consecutively located nests was farther for males than females. Sixty-eight percent of the nests used were in cavities. Snags and larger trees were selected for nest sites more than expected based on availability. This study will help managers understand an aspect of microhabitat resource use by northern flying squirrels in ponderosa pine (Pinus ponderosa) dominated habitat of BHNF and an isolated population at the southern edge of their range.

Northern flying squirrels (Glaucomys sabrinus) of the Black Hills National Forest (BHNF) of South Dakota represent a unique and isolated population, but little is known about the home range size and habitat use of this population. Fifty-nine northern flying squirrels (34 males and 25 females) were radio-collared and tracked during their active period, from dusk until dawn, by point sampling. Minimum convex polygons (MCPs) were determined from observation locations of squirrels with ≥15 radio-tracking locations (n  =  49). Males occupied larger home ranges (n  =  30; 11.23 ± 1.48 ha) than females (n  =  19; 6.91 ± 0.94 ha; P  =  0.02). Using the radiotelemetry data and GIS vegetation layers, habitat use (grass-shrub, aspen-birch, bur oak and pine) and structural stage class (1, 3A, 3B, 3C, 4A, 4B and 4C) selection were determined for all squirrels with ≥10 radio-tracking locations (n  =  54). Habitat selection was determined by comparing the proportion of radio-tracking locations (observed) within each habitat to the proportion of habitat within the MCPs (available) using techniques developed by Neu et al. (1974). Ponderosa pine (Pinus ponderosa), which dominates 83% of BHNF, was the only habitat used proportionally more than available. Within their home ranges northern flying squirrels also selected larger trees and more canopy cover, as well as more live trees >12.7 cm dbh, higher basal area of live trees and fewer snags. This study aids managers in understanding habitat use by northern flying squirrels in pine dominated habitat of BHNF and an isolated population at the southern edge of their range.

Populations at the edge of their geographic range may demonstrate different population dynamics from central populations. Endangered Mt. Graham red squirrels (Tamiasciurus hudsonicus grahamensis), endemic to southeastern Arizona, represent the southernmost red squirrel population and are found at lower densities than conspecifics in the center of the range. To determine if differences are due to conditions at the southern periphery of the range, we compared habitat characteristics, demography, body mass, space use and nesting behavior with another subspecies located at the southern edge of the range, the Mogollon red squirrel (T. h. mogollonensis). We found that mean and minimum daily temperatures were higher at Mt. Graham whereas maximum temperatures were higher in the White Mountains, male Mogollon red squirrels were heavier than male Mt. Graham red squirrels in all seasons and female Mogollon red squirrels were slightly heavier than female Mt. Graham red squirrels in spring, proportion of squirrels in reproductive condition was lower in female Mogollon red squirrels, Mogollon red squirrels had smaller home ranges, used different types of nests and traveled less distance to nest than Mt. Graham red squirrels. There were no differences in annual rainfall, seedfall, habitat characteristics or survival between mountain ranges. Localized conditions appear to account for the disparity between populations. These differences demonstrate the importance of evaluating attributes of peripheral populations for maximizing persistence and intraspecific diversity.

We investigated roost use by western long-eared bats (Myotis evotis), including roost fidelity, movement between roosts and the size of roosting home-ranges. We expected reproductive females, especially lactating females, to exhibit higher roost fidelity and to move shorter distances between roosts than males or nonreproductive females. Individuals generally switched roosts every 1 or 2 d. Despite their low roost fidelity, individuals occupied small roosting home-ranges (<2 ha), demonstrating fidelity to a roosting area. Roost fidelity, distances moved between roosts and roosting home-range sizes were not associated with reproductive class or reproductive period, possibly indicating that any energetic cost of roost switching is trivial in the energy budget of M. evotis, or that the advantages of roost switching outweigh any energetic cost.

Large ungulates are an important driver of plant community composition and structure. In eastern North America, white-tailed deer (Odocoileus virginianus) thrive in agricultural mosaics and fragmented forested landscapes, at times reaching unprecedented densities. Nevertheless, few long-term data sets are available that allow an assessment of the long-term consequences of chronic herbivory. We quantified herbaceous-layer change over a 26 y period in Cades Cove, Great Smoky Mountains National Park, USA. Cades Cove has a long and well-documented history of deer overabundance, with densities reaching 43 deer km⁻² during the late 1970s. Over the 26 y sampling interval, mean coverage of herbaceous species declined significantly (P < 0.001) in the forests bordering Cades Cove. Although most plots only lost 1–2 species during the interval, 46 herbaceous species recorded on plots during the 1970s were wholly absent in 2004 (63% of which were forest species). Additionally, the herbaceous layer has become significantly more homogeneous over time. In contrast, species richness and cover on reference plots increased by 106 and 183%, respectively, over a similar time interval. Whereas some compositional changes were associated with forest succession, proximity to the Cove's edge environments was the most informative environmental gradient, lending support to the hypothesis that deer foraging behavior results in a biotic edge effect in fragmented landscapes. Chronic herbivory may result in impoverishment and simplification of herbaceous layers in forests otherwise protected from habitat degradation and loss.

Recent attention has been paid to spatial variation in the direct and indirect effects of trophic interactions. Because abundances of predators and prey vary naturally through space, their interactions and the effects of these interactions may vary as well. We conducted a bird exclosure experiment on white oak (Quercus alba L.) using a randomized block design to assess how the direct effects of bird predation on arthropods and indirect effects of birds on plant damage and growth differ between five sites separated by 350–1000 m. Insect herbivore and arthropod predator abundances varied spatially, but were not affected by the exclosure treatment. Bird abundance also varied among sites. Herbivore community structure (herbivore feeding guilds) differed by site as well. Bird predation significantly reduced damage to oak leaves, but this effect did not vary spatially. However, the size of this effect was positively correlated with insectivorous bird abundance. Thus despite herbivore and predator communities that varied among sites, the direct and indirect effects of bird predation appeared to be constant at the local scale at which this experiment was conducted.

We characterize structural variation of forests dominated by American beech (Fagus grandifolia) on Naushon Island, Massachusetts with respect to edaphic and environmental conditions, and describe dwarf beech forests that are rare throughout the eastern U.S. Age structure reconstructions and historical references confirm that dwarf beech stands have persisted in the study area for >100 y. Dwarf beech are characterized by extremely slow radial growth rates, with some individuals growing <0.1 mm per year for >25 y, and ages up to 200 y. In the most severely stunted stands, all beech stems are <2 m tall and <8 cm basal diameter. In contrast, adjacent tall-stature forests support beech trees of similar age that are 20–30 m tall and up to 70 cm diameter (at 1.4 m). Variation in vegetation structure is strongly related to topographic position and edaphic conditions. Dwarf stands occur on morainal knobs and ridges characterized by excessively-drained sandy soils and well-developed E-horizons; soil organic horizons are absent or minimal as a result of chronic wind-removal of leaf litter. Tall-stature beech stands occur in nearby topographic depressions characterized by finer-textured soils, greater soil fertility and protection from chronic wind disturbance. Dendroecological analyses document differential tree growth and establishment responses to severe disturbances among structural types. However, individual disturbance events do not appear to contribute significantly to forest structure. Instead, dwarf growth forms apparently develop in response to harsh edaphic conditions, including chronic nutrient depletion, drought stress and wind exposure.

Variation in available resources (e.g., environment or food) can influence life history variation. Populations of the streamside salamander (Ambystoma barbouri) breed in both streams and ponds and larvae exhibit diverse life history patterns related to the environment within which they develop. Previous experiments demonstrated that environmental factors are largely responsible for differences in larval period, mass and size at metamorphosis between the two forms of A. barbouri; however, it is unclear which environmental factors influence the variation. Under controlled laboratory conditions, we tested if hydrological differences between lotic and lentic habitats contribute to life history divergence in A. barbouri. We reared laboratory born larvae from stream- and pond-collected eggs in laboratory environments with hydrological patterns similar to their natal environments. Ambystoma barbouri stream larvae metamorphosed faster and were smaller in mass and size compared to pond larvae. Although we did not test whether genetic or environmental factors influence larval life history traits, our data provide support that hydrological variation alone may contribute to the life history variation in the two forms of larval A. barbouri.

We investigated the attributes of rock crevices selected by Allegheny (Neotoma magister) and eastern woodrats (N. floridana haematoreia) in their zone of contact in the Appalachian Mountains of North Carolina. In North Carolina, N. magister and N. f. haematoreia both occur in rocky habitats above 300 m, and are listed as species of special concern. We studied 14 active sites (N. magister: n  =  9; N. f. haematoreia: n  =  5) where woodrats had been captured 1 y prior to our study and identified to species using the mitochondrial DNA D-loop analysis. At each site, we measured the attributes of 10 crevices used by woodrats and 10 corresponding random crevices located < 15 m from each used crevice. Neotoma magister and N. f. haematoreia selected crevices with larger dimensions (height, width and depth) and more internal fissures (openings >5 cm in diameter) than those available in the surrounding environments. All crevices used by N. magister (n  =  90) and N. f. haematoreia (n  =  50) were dry. Neotoma magister were more specialized than N. f. haematoreia, as they selected crevices that were south-facing. These results suggest that both N. magister and N. f. haematoreia are habitat specialists in the southern Appalachians, preferring crevices with larger dimensions and more internal fissures to enhance their protection against severe weather and predators. The preference for south-facing crevices by N. magister suggests that they may be better adapted at surviving colder climatic conditions, thus enabling them to inhabit higher elevations in the mountains. Based on these specialized habitat preferences, we suggest that suitable rock crevices may be a limiting factor to both species in the southern Appalachians. In addition, the similarity in attributes of rock crevices selected by these species suggests that habitat is not a factor that will prevent hybridization between these species where they co-occur in the mountains of North Carolina.

Post-dispersal seed predation by rodents represents a potentially important element of biotic resistance to plant invasion. Selection for five different types of seeds by granivorous rodents was studied in maple-beech forests, old fields and conifer plantations in Madison County, New York. Rodents visited dishes containing equal masses of seeds of the native Cornus amomum and Rubus idaeus, and the non-native Lonicera morrowii, Rhamnus cathartica and Rosa multiflora. Greater masses of C. amomum and R. idaeus seeds were consumed during a night of mammal visitation than of the three non-native species, and pattern of selection did not differ among habitats. Rodents encountered seed dishes sooner in forested habitats than old fields. The primary seed predators in our region, Peromyscus spp., were more common at forests and plantations than old fields. Patterns of habitat use by Peromyscus spp. may aid in resisting invasion of intact forests by invasive plants; however, selection of native over non-native seeds may facilitate differential establishment of non-native invaders.