We examined vertical, longitudinal and seasonal variation in the abundance, taxa richness and community composition of the epibenthic and hyporheic macrobenthos at Elklick Run, a first- through fourth-order stream continuum in the central Appalachian Mountains in West Virginia. We sampled the macrobenthos at three levels (0–10 cm, 10–20 cm, 20–30 cm) below the stream bed using baskets filled with natural substratum. We quantified environmental factors including interstitial organic matter, fine sediment, water velocity and dissolved oxygen. Abundance, taxa richness and the relative abundance of streamlined taxa decreased with depth into the substratum. Interactions between depth into the substratum and site and between depth and season, were caused by a winter spate that reduced the abundance and taxa richness of the benthos in the upper layer of the hyporheic habitat (0–10 cm), but only at downstream sites. There was broad overlap in community structure among depths into the hyporheos, and virtually no taxa increased in absolute abundance with depth into the substratum or were restricted to hyporheic habitat. Abundance and taxa richness varied more with depth into the substratum than among sites or among seasons. However, both epibenthic and hyporheic community structure varied much more among seasons than among sites or among depths. The hyporheic fauna resembled the epibenthic fauna less at downstream sites where there was a more specialized epibenthic fauna including more streamlined taxa, filterers and scrapers. Abundance and taxa richness of the macrobenthos was positively correlated with interstitial flow, especially at upstream sites in Fall when stream flow was lowest, water temperature was high and interstitial dissolved oxygen concentration was low. The coefficient of variation in hyporheic abundance, taxa richness and community composition decreased with increasing stream order, suggesting decreased spatial variability in interstitial habitat suitability at downstream sites.

Three species of freshwater sponges (Porifera): Spongilla lacustris (Linneaus), Eunapius fragilis (Leidy) and Ephydatia muelleri (Lieberkuhn) were identified in four southern Lake Michigan harbors: Belmont (IL), East Chicago (IN), Michigan City (IN) and St. Joseph (MI). Samples were collected from artificial substrates and are the first report of sponges in three of the harbors. Sponge cover on revetment walls, stone rip-rap and wood pier posts ranged up to 13% of the available surface area, whereas no sponges were growing on the soft silt bottom. Eunapius fragilis appeared to be the most common species. An angle-density relationship was observed for sponges growing on the stone rip-rap in Michigan City and East Chicago; as the orientation of the attachment substrate changed from horizontal to vertical, the density of sponges increased.

We used a mark-recapture method to measure survival of a native unionid mussel, Amblema plicata, at sites with low, moderate and high zebra mussel (Dreissena polymorpha) densities (mean values ± se equaled 0.4 ± 0.2, 50.5 ± 13 and 1750 ± 260 individuals/m² respectively) in the Mississippi River, Minnesota, and at one site without D. polymorpha in the Otter Tail River, Minnesota. In 1996 240 A. plicata from each site were uniquely marked. In 1997, 1998, 1999 and 2000 marked mussels were recovered; identified; survival determined; and, if alive, they were returned to the substratum. Mean annual survival differed significantly among sites (χ² = 4.08, df 3, P < 0.0001). Mean annual survival rates in the Mississippi River's low infestation population (LOW) were 99 ± 0.3%, and 89 ± 1.2% in the moderately infested population (MOD), while survival at the highly infested population (HIH) was 65 ± 7.8%. Mean annual survival for the Otter Tail River population (REF) of A. plicata was 98 ± 0.5%. Results demonstrated that A. plicata has high mean annual survival (>97%) in natural habitats that are not colonized by D. polymorpha and the survival rates decline significantly relative to increases in D. polymorpha densities.

The ranges of the Ozark sculpin, Cottus hypselurus, and the mottled sculpin, C. bairdi, in Missouri are unresolved. Current distributions have been based on a series of indexed morphological features and morphometric ratios that can result in taxonomic difficulties. Clarification of the distributions was sought using a molecular approach to defining distinctness and diversity of the species samples. Data were used to provide Randomly Amplified Polymorphic DNA comparisons between collections from Clear Creek, Pearson Creek, the Meramec River, Little Black River and Niangua River systems. Data analysis yielded a cladogram that revealed inaccuracies in currently accepted taxonomic relationships. The results show a monophyletic clade representing mottled sculpins from four of the sampled sites. Individuals from the Niangua River formed a separate clade consisting of Ozark sculpins. The results suggest a need for further investigation and a revised classification of the sculpins studied.

We examined patterns of spatial heterogeneity in the Spring River basin fish community along with environmental correlates to assess the relative importance of geographic distances and habitat differences among sites in explaining community structure. Spatial patterns of the fish community and environmental correlates, as indicated by results of Mantel tests and correlograms, were consistent with the hypothesis that environmental factors were the primary factors organizing the community at the basinwide scale. However, a linkage between mainstream and tributary communites, along with spatial autocorrelation in species composition, suggested contagious biotic processes were important in maintaining community structure, particularly at the interface between the mainstream and its tributaries. Space-constrained cluster analysis and principal coordinates analysis revealed three primary groups of sites, reflecting relatively distinct fish faunas within the Ozark Highlands, Central Plains and mainstream regions of the basin. Results of partial Mantel tests indicated that whereas environmental differences were significantly correlated to species differences, variation in community data could alternately be explained by contagious biotic processes, a factor often neglected in community analyses of stream drainage networks.

We assessed short-term variability and long-term change in the composition of the littoral fish community in Spirit Lake, Iowa. Fish were sampled in several locations at night with large beach seines during spring, summer and fall of 1995–1998. Long-term changes were inferred from comparison with a similar study conducted over 70 y earlier in Spirit Lake. We found 26 species in the littoral zone. The number of species per sample ranged from 4 to 18, averaging 11.8. The average number of species per sample was higher at stations with greater vegetation density. A distinct seasonal pattern was evident in the number of species collected per sample in most years, increasing steadily from spring to fall. Patterns of variability within our 1995–1998 study period suggest that: (1) numerous samples are necessary to adequately characterize a littoral fish community, (2) sampling should be done when vegetation and young-of-year densities are highest and (3) sampling during a single year is inadequate to reveal the full community. The number of native species has declined by approximately 25% over the last 70 y. A coincident decline in littoral vegetation and associated habitat changes during the same period are likely causes of the long-term community change.

Muskrat (Ondatra zibethicus) herbivory and lodge construction create open water areas in stands of emergent vegetation. This study examined how aquatic invertebrates are affected by changes in the physical structure of plant stands created by muskrats. We compared density, species composition and functional feeding groups of benthic macroinvertebrate communities and abiotic conditions in open water areas at muskrat lodges and in adjacent cattail stands. Although total numbers of macroinvertebrates were not affected by muskrat activities, communities at muskrat lodges were less diverse than in cattail stands. Gathering collectors such as chironomid midges that feed on fine particulate organic matter were the dominant trophic group; these were significantly more abundant at muskrat lodges than in cattail stands. Scrapers, such as gammarid scuds that feed on periphyton, were also abundant; these were significantly more abundant in cattail stands than at muskrat lodges. Water temperature, dissolved oxygen, pH, conductivity and sediment organic content were different in these habitats. These results indicate that muskrats alter abiotic conditions and invertebrate communities in wetlands, which will affect food resources for wildlife and fish that feed on aquatic invertebrates in these habitats.

A field study was conducted in Jalisco, Mexico, to quantify temporal resource switching by ants between honeydew produced by the fivespotted gama grass leafhopper, Dalbulus quinquenotatus, and nectar produced by plants with extrafloral nectaries (EFNs). Four plant taxa (Acacia pennatula, Leucaena esculenta, Lobelia laxiflora and Lysilona sp.) which produced nectar from EFNs were monitored for ants during the dry and wet seasons in ecological communities where D. quinquenotatus was tended by ants on their gama grass (Tripsacum spp.) hosts. These plants with EFNs were visited preferentially in May (the driest month) by Brachymyrmex obscurior, a common ant species that tended D. quinquenotatus in all months except May. Our results suggest temporal resource switching by ants between plants with EFNs and D. quinquenotatus. This cycle is regulated by the amount of nectar produced by plants with EFNs and honeydew produced by D. quinquenotatus. Although B. obscurior was the most common ant to visit plants with EFNs and tend D. quinquenotatus, four additional species of ants were also associated with both food resources.

Weekly reports of the abundance of the Argentine ant, Linepithema humile, in 69 households for 18 mo (1/98–7/99) in the San Francisco Bay Area in northern California were compared with weather data. Ant abundance inside homes was highest in cold rainy weather, and there was a second smaller peak of ant abundance in hot dry weather. Pesticide use in the home decreased ant abundance, from one week to the next, only when ant abundance was extremely high.

The North American porcupine (Erethizon dorsatum) has expanded its range into the Edwards Plateau of Texas, an area harboring a disjunct relict population of papershell pinyon pine (Pinus remota). We captured 39 porcupines and monitored 37 of these animals by radiotelemetry from 1997 to 1999 to assess their demographic characteristics and determine their potential impact in this unique wooded ecosystem. The adult population was biased toward females (3 females: 1 male), and annual rates of survival were greater (P < 0.01) for adult females (S = 0.91, 95% ci = 0.77–1.00) than for adult males (S = 0.42, 95% ci = 0.00–0.85). Birth rates of reproductively mature females averaged 0.82 annually during the 3-y study. We calculated a mean reproductive rate of 0.41 female young/female based on birth rates and juvenile sex ratio. The estimated population growth rate of 0.034 indicated that the study population was growing. Direct and indirect ecological effects of porcupines in this woodland ecosystem warrant investigation.

In 1995 we deployed 800 simulated nests equipped with nest mortality timers to assess temporal patterns of predation of duck nests in the Canadian prairies. Timers recorded 470 predation events (70% of 672 depredated nests). Predation was nonrandom throughout the day, with a peak occurring between 00:00 and 06:00. Coyote (Canis latrans) and red fox (Vulpes vulpes) depredated nests at similar times (mean 18:32 and 22:42, respectively), whereas mean depredation time was significantly later for striped skunks (Mephitis mephitis, 02:08). We failed to detect a difference in depredation interval for neighboring nests destroyed by striped skunk (median = 62 h 18 min, n = 15) vs. pairs of nests depredated by different species (red fox-skunk, skunk-coyote and coyote-red fox, median depredation interval = 123 h 1 min, n = 6). Simultaneous observation of radio-marked skunks revealed that skunks spent 2–25 min at simulated nests and became inactive after nest destruction in ca. 50% of cases. Based on observations of striped skunk behavior and nest-timer data, we suggest that density-dependent nest predation is caused by increased patch use rather than area-restricted search.

We analyzed macrohabitat and microhabitat associations of four soricid and five rodent species in five macrohabitats on the Coastal Plain of Virginia. There were no significant differences in total small mammal abundance among macrohabitat types based on total captures/unit effort. However, abundances of four species, Cryptotis parva, Reithrodontomys humulis, Microtus pennsylvanicus and Zapus hudsonius, were significantly higher (P < 0.05) in old fields than in four forested habitats. Canonical correspondence analysis revealed that 27% of the variation in small mammal distributions was attributable to microhabitat characteristics. Three characteristics that had a particular influence on small mammal presence in forested habitats were shrub frequency (Sorex longirostris), canopy openness (S. hoyi) and diameter of downed woody debris (S. hoyi, Blarina brevicauda, Microtus pinetorum). Correlations between small mammals and microhabitat characteristics are due to local moisture gradients and structural heterogeneity. Lack of correlations between Peromyscus leucopus and any microhabitat characteristic is due to the ability of this species to obtain requirements from a variety of sources. Preservation of microhabitat characteristics like downed woody debris and understory vegetation, and certain macrohabitats (e.g., old fields), would require minimal management effort and provide suitable habitat for a diverse small mammal fauna in fragmented landscapes.

We studied the influence of grazing by bison (Bos bison) and by cattle (B. taurus) on deer mice (Peromyscus maniculatus) in tallgrass prairie at the Konza Prairie Biological Station in 1997 and 1998. Small mammals were sampled by one 10-station trapline in each of four bison-grazed enclosures, four cattle-grazed enclosures and four ungrazed sites. Enclosures were 4.9 ha and the biomass of grazers in each was similar. All sites were burned annually. We sampled small mammals for 4 consecutive nights in spring before fire, in spring after fire and in autumn. Deer mice were the most abundant species (n = 285; 83% of all small mammals) captured in all treatments and in each trapping period. Deer mice were significantly more abundant in bison-grazed and cattle-grazed sites than in ungrazed sites in spring before fire (P < 0.01 and P < 0.05, respectively), but were similar in abundance in grazed and ungrazed sites following fire. Abundance of deer mice was significantly higher in bison-grazed sites than in cattle-grazed and ungrazed sites in autumn (P < 0.05 and P < 0.001, respectively). Bison and cattle differ in grazing and nongrazing behaviors (e.g., wallowing by bison) that result in differences in vegetation structure. It is likely that differences in deer mouse abundance between bison-grazed and cattle-grazed treatments were due to differences in vegetation structure caused by the two types of grazers.

The raccoon parasite Baylisascaris procyonis is implicated in the decline of the Allegheny woodrat (Neotoma magister). Baylisascaris procyonis is highly pathogenic in non raccoon hosts and is transmitted by eggs which are passed in raccoon feces. Granivorous species foraging for seeds in raccoon latrine areas risk exposure to the parasite. Various granivores exhibit different foraging behaviors, which may affect their probability of consuming B. procyonis eggs. To investigate the effects of species specific foraging strategies on the likelihood of B. procyonis transmission, I compared the latrine foraging behaviors of the white-footed mouse (Peromyscus leucopus) and the Allegheny woodrat during 1997 and 1998. The species use different foraging strategies, which result in differential risks of consuming embryonated B. procyonis eggs. Specifically, Allegheny woodrats carried whole feces to food caches, whereas white-footed mice primarily extracted seeds from the feces. Thus, woodrats run a greater risk of contaminating their entire cache with parasite eggs. In addition, woodrats did not collect fresh feces, but waited an average of 21 d before collection, while mice removed seeds immediately upon discovering the feces and preferentially foraged on seeds imbedded in fresh soft feces rather than on those in dried hard feces. This is relevant because B. procyonis eggs are harmless until embryonated, a process that takes approximately 2 to 4 wk. Therefore, mice reduce their risk by removing seeds from feces when the eggs are harmless, whereas woodrats increase their risk by waiting until eggs may be infective. These distinct foraging strategies help to explain the differential impact of B. procyonis on the two species.

Flowering dogwood (Cornus florida L.) fruits are an important fall food for many species of birds and mammals throughout the eastern United States. However, the rapid spread of the fungal disease dogwood anthracnose (Discula destructiva Redlin) has caused extensive mortality of dogwoods throughout the Appalachians. We examined the impacts of dogwood anthracnose on the production of inflorescences by flowering dogwoods in western North Carolina as an indication of fruit production. We also collected ripe fruits in the fall, weighed the pulps and analyzed them for protein, lipids, Ca, K, Mg and P. Only one-third of 173 live trees flowered. Large trees were significantly more likely to flower than small trees, and as disease severity increased, the probability of flowering decreased. Neither the mean number of inflorescences per tree nor mean pulp weight differed among two size classes or three disease categories of trees. However, pulps from moderately infected trees contained significantly more Ca and P than pulps from lightly infected trees. Pulps from severely infected trees also contained higher levels of Ca than pulps from lightly diseased trees. A discriminant function analysis examining all six nutritional variables separated lightly diseased pulps from the other two categories along an axis represented by Ca, P and lipids. Moderately diseased pulps were separated from severely diseased pulps by K, P and protein. The most significant impact of dogwood anthracnose on frugivores will likely be the loss of flowering dogwoods from the landscape, with the concomitant loss of actual and potential fruit production.

Acclimation to low light may interfere with the ability of woody plants to resprout. We examined the patterns of biomass allocation and resprout ability of seedlings of Paulownia tomentosa, a heliophytic r-selected species. Plants were grown in shade houses in a common garden with two replicates of three light treatments: full ambient light, artificial edge and shade. The shade treatments were similar to light patterns and levels typical of a deciduous forest understory. The artificial edge treatment provided an intermediate light regime since it received only direct morning sun and was shaded in the afternoon. The experimentally browsed treatment consisted of clipping 10 tree seedlings to ground level in each replicate light treatment four times over the growing season. In all light treatments, whole plants were harvested on each clipping date. Plants in all treatments allocated more to belowground biomass in the first weeks of the experiment and then shifted allocation to aboveground biomass. Plants grown in shade had lower relative growth rates (RGR) and higher specific leaf areas (SLA) and leaf area ratios than plants in the other light treatments. Ability to resprout was influenced by the amount of accumulated belowground biomass and, since this was lowest in the shade treatment, resprouting was reduced in low light. Increased SLA was not correlated with an increase in RGR. Also, correlation networks showed decreased integration in the shade. Our data suggest that, although resprout ability is dependent upon belowground biomass, P. tomentosa seedlings can resprout at an early age, even in low light. This ability may allow the species to become established even in areas of high herbivore density.

We monitored 93 nests of blue-gray gnatcatchers (Polioptila caerulea) from 1995–1997 in east-central Illinois. Habitat selection was assessed at three spatial scales (1 m sphere around nest, 0.005 ha plot and 0.04 ha plot). Leaf density at the 1 m scale was greater at nest sites than non-nest sites. This was the only variable differing between nest sites and paired random sites. Thus, habitat selection was not detected at the larger scales. Overall nest success was 11% (ci 95%: 6–18%); successful nests were significantly higher and farther from habitat edges than unsuccessful nests. Of 36 nests whose contents we could directly observe, 20 (56%) were parasitized by brown-headed cowbirds (Molothrus ater). Gnatcatchers abandoned 16 (80%) parasitized nests during egg-laying or the first 3 d of incubation, whereas only 1 (6%) unparasitized nests were abandoned during these stages. Daily nest success of renests was significantly lower (0.89 ± 0.017) than that of first nests (0.94 ± 0.009). The percentage of successful pairs was 36%. Habitat characteristics at a pair's second nest site were not different from its initial nesting attempt. Lack of specific habitat requirements and persistent renesting may be necessary for this species to achieve reproductive success in this study area.

Tunnel entrance distributions at six bank swallow sand-pit colonies showed consistently nonrandom, too-regular patterns, supporting the hypothesis that the distance between tunnel entrances is determined by territorial disputes at the tunnel mouths. Average nearest neighbor distances, and numbers of burrows per unit area of pit face, were both consistently greater than their random expectations, a paradox that is explained algebraically. Evidence of tunnel coalescence and communal nesting was found in a completely evacuated colony of 30 tunnels. It is argued that bank swallow colonial behaviors evolved to maximize populations on small exposed bank faces along streams and rivers, and that they are still behaving as if only small fractions of large sand pit banks are available. On small areas, regular spacings help to avoid too close tunneling and concomitant bank collapse. It is argued that a tolerance for communal nesting, once coalescence has occurred, involves less tunneling than the forced evacuations of entirely new tunnels by ousted pairs.

Data on the diet of the web-invading spider Mimetus notius were collected to determine the diet of this araneophagic spider. Mimetus notius primarily invades Theridiidae and Aaraneidae webs and is capable of preying upon at least seven families of spiders, principally in the genus Theridion (Araneae: Theridiidae). Mimetus notius invades the webs of several species on which no successful predation was observed. These data provide the first field data showing significant insectivory by a web invading mimetid, constituting 29% of the prey in this study. The presence of insects in the diet provides a potential explanation for the invasion of webs where successful predation on the resident is unlikely.

Although fluorescent powder marking is a well established and common marking technique used by small-mammal ecologists, few studies have commented on how it might affect behavior. Potential effects on behavior are particularly relevant in studies that use fluorescent powder marking as a method of quantifying mating behavior. We performed a laboratory experiment using white-footed mice (Peromyscus leucopus) to test the null hypothesis that the presence of fluorescent powder has no significant effect on mate choice. We tested this hypothesis by establishing choice between a pair of females by a male and determining if the choice changed when one of the females was powdered. The presence of fluorescent powder on females did not alter mate choice by males after a choice had been established. Our results demonstrate that using fluorescent powder on females should not alter male mating activity.

Before 1992 the leafhopper Attenuipyga vanduzeei (Osborn & Ball) had seldom been collected and little was known about its life history. The Prairie Invertebrate Biodiversity Inventory, a multi-partner project focused on the invertebrate fauna of midwestern prairies, has located four sites for the species in Wisconsin and added to the number of curated specimens. Field collection data and laboratory observations of living specimens contribute new information regarding A. vanduzeei's possible host plant(s) and provide the first details of life history and behavior. It is possible that A. vanduzeei alternates hosts as an adult. Laboratory specimens demonstrated a preference for prairie dropseed (Sporobolus heterolepis) as a resting host, yet chose sideoats grama (Bouteloua curtipendula) for oviposition. Plant and habitat associations demonstrate that this is a rare prairie-dependent species.

Fecal pellets from 115 federally endangered West Virginia northern flying squirrels (Glaucomys sabrinus fuscus) were collected in the Monongahela National Forest, West Virginia during the spring and fall in 1989–1991 and analyzed to determine the squirrel's diet. In spring the squirrel's diet consisted primarily of tree buds, lichens and hypogeous fungi. In fall hypogeous and epigeous fungi and beechnuts were the most common food items. The epigeous fungi were in the families Boletaceae, Strophariaceae and Russulaceae. Hypogeous fungi in the genus Elaphomyces were consumed by 50.8% of the squirrels in the spring and 48.2% in the fall. Five other hypogeous taxa were identified in 5.2% of the samples. All hypogeous species identified in the scats form mycorrhizal relationships with forest trees and are dependent on mycophagy for spore dispersal. West Virginia northern flying squirrels facilitate spore dispersal of these mycorrhizal fungi and may contribute to the health of tree species in high elevation red spruce (Picea rubens)/northern hardwood forests in West Virginia.

Approximately 70 plant families worldwide have ant-dispersed seeds (myrmecochory). In this putative ant-plant mutualism, ants are attracted to and disperse seeds that have a lipid-rich elaiosome. We observed yellow jackets (Vespula spp.) dispersing seeds of three elaiosome-bearing species—Trillium cuneatum, T. undulatum and T. catesbaei—in the Blue Ridge Mountains of North Carolina and South Carolina. Moreover, we estimated the mean distance yellow jackets dispersed seeds of T. cuneatum by placing intact fruits on index cards and recovering dispersed seeds on sheets placed on the ground surface. Of the seeds presented, 41% were recovered and the average dispersal distance was 1.4 m (range 0.1–2.6 m). Some yellow jackets carrying Trillium seeds flew out of sight and probably dispersed seeds farther (perhaps 20 m or more). To our knowledge, this is the first report of yellow jackets dispersing elaiosome-bearing seeds in eastern North America. Although the fate of vespid-dispersed seeds is unknown, seed dispersal by yellow jackets might benefit plants by increasing the dispersal distance of seeds and, therefore, potentially reducing density-dependent mortality and expanding species ranges.

After several years of disuse, bank slumping in a large (395.6 m circumference), roughly circular sand pit had raised the talus to a depth of 2.0 m, reaching to within 0.5 to 0.7 m of the overlying sod. No bank swallows (Riparia riparia) were nesting in this sand pit. Two truckloads of this readily-scooped talus were removed in 1995, creating separate stretches (each about 11 m) in which a shallow new talus (<0.5 m) left 2.0 m to 2.2 m of exposed, vertical bank. In the spring of 1996, bank swallows established colonies in each of these newly exposed stretches and nowhere else. A “roulette wheel” probability model is presented from which the probability that these low talus stretches were selected as colony sites by chance is 0.001.