Experimental removal of flower heads of Taraxacum officinale (Common Dandelion) tested whether the presence of these showy yellow flowers affected seed production of Delphinium nuttallianum in natural populations where both species are visited by native bumble bees. Seed production per flower and per plant was independent of dandelion flower presence in the experimental plots. There was no indication of competition for pollination between the two plant species.

We investigated whether Japanese barberry in forest understory responded to N availability through physiological adjustments at the leaf level or growth components at the plant level. Barberry increased foliar N, photosynthesis at ambient and saturating light and specific leaf area in proportion to N availability. Barberry relative production rate increased with foliar N, leaf biomass, specific leaf area and total canopy N mass. Increases in production and N uptake were proportional and, thus, neither photosynthetic nitrogen-use efficiency nor production per unit of total canopy N differed with treatment or across the range of N status.

Coastal wetlands of the Laurentian Great Lakes provide many valuable ecological functions and, currently, there are extensive efforts to preserve and manage remaining habitats. We examined plant recruitment in a Lake Erie coastal marsh on bare mudflats exposed during 2 y of below average water levels. In 2000 mudflat vegetation was sampled in shallow, medium and deep water transects and compared to germinable seeds found in sediments from 0–15 cm below the surface. Effects of sediment disturbance and herbivory on plant recruitment were tested in 1 m² quadrats in 2000 and vegetation was sampled in 2001 to assess how well the disturbance/herbivory experiment predicted plant communities that became established on the mudflats. In 2000, 18 species were found on mudflats and 19 species were reared from the seedbanks. Seed densities (∼3000 seedlings/m²) were comparable to other freshwater wetlands and did not differ by sediment depth. Taxa richness was highest in sediments 0–5 cm below the surface. Most plants on the mudflats were found in the shallow transect, but seed abundance and richness were highest in deep transects. Sediment disturbance decreased taxa richness and herbivory decreased both plant cover and richness; however, responses varied among plant taxa. Our results indicate that abundant seedbanks exist in this wetland, but plant recruitment will be restricted to shallow areas due to turbidity. Communities in 2001 were more diverse (40 species) than predicted from vegetation in 2000 probably because propagules were introduced from nearby habitats. Communities in 2001 were most similar to ungrazed and undisturbed quadrats in 2000 because these had the most species.

The invasive vine Celastrus orbiculatus Thunb. (Oriental bittersweet) dominates gap and edge environments, but may also colonize undisturbed forest. We compared survival and growth of C. orbiculatus seedlings in field plots under 2%, 28% and 100% sun. From transplanting through the first autumn, survival and growth did not differ among treatments. In the second growing season, survival at 2% sun was 76%, compared to 96% in 28% sun. Growth and biomass were greater in the 100% and 28% sun treatments than 2% sun. The ratio of leaf to total biomass (LBR) decreased with shade, but leaf mass per leaf area (LMA) decreased proportionally more, so that the leaf area per unit biomass (LAR) increased in the shade. Photosynthesis, stomatal conductance and the ratio of photosynthesis to conductance (A/g) decreased in the shade. The ability of C. orbiculatus to survive under deep shade despite its slow growth implies that intact forests are vulnerable to invasion and that established understory populations should be controlled before harvesting or thinning the forest.

Over the last 200 y the tallgrass prairie of the Midwestern United States has experienced widespread conversion to agricultural production. Today, the few remaining tracts of unplowed grassland persist as small isolated patches within a landscape of row-crop agriculture. The small size and isolation of these prairie remnants raises concerns over their long-term sustainability. In this study I examined changes in the upland plant community that have occurred since state acquisition of one of Iowa's oldest prairie preserves. I found that 50 y of management has succeeded in reducing the frequency of exotic species and, thus, improved the overall integrity of the native plant community. However, during this same period dramatic changes in the frequency of many native species have also occurred. A general increase by mesic and late flowering species and a decrease by xeric natives was observed. Changes coincided with a documented shift in management from mid-summer haying prior to state ownership to spring burning, suggesting that burning may have exerted selective forces altering the composition of the native community over the past 50 y. My results emphasize the need to merge our current understanding of the processes that help sustain diversity into implemented management practices that will prolong the diversity of our remaining small isolated prairie preserves.

Most old fields in the Queets Valley of Olympic National Park, USA, remain dominated by exotic herbs 60 y after abandonment although the fields are surrounded by temperate rain forest. However, areas of some fields have been invaded by Picea sitchensis, one of three dominant forest species (with Alnus rubra and Tsuga heterophylla). This provided an opportunity to examine local variation of factors (competition, facilitation, cervid herbivory, soils) that influence tree colonization within a set of old fields, an approach rare in previous studies. Picea sapling invasion of field edges was negatively correlated with percent cover of Agrostis gigantea and positively correlated with Anthoxanthum odoratum. Potential indicators of competition (sod thickness, thatch thickness, percent ground cover) were correlated with Agrostis cover. Picea edge invasion was also correlated with soil organic matter. In experiments, seedlings of Picea and Tsuga emerged as readily in Agrostis as in Anthoxanthum or Pteridium aquilinum, but suffered higher mortality in Agrostis. Experimental seedling establishment was low and required reduction of competing vegetation. In experiments with transplanted seedlings, cervid herbivory suppressed growth of Tsuga and Alnus, but not Picea. Growth of Picea seedlings was facilitated by Pteridium. Differential tree colonization of the fields resulted from: (1) differential competition by invaded cover types against seedling establishment, (2) selective herbivory on tree species of established seedlings and (3) facilitation by fern cover of seedlings of an unbrowsed species.

Large rivers of the United States have been altered by construction and maintenance of navigation channels, which has resulted in habitat loss and degradation. Using 7 y of Long Term Resource Monitoring Program data collected from the unimpounded upper Mississippi River, we investigated Ohio and Glass Shrimp abundance collected from four physical habitats of the unimpounded upper Mississippi River: main channel border, main channel border with wing dike, open side channel and closed side channel. Our objective was to assess associations between Ohio and Glass Shrimp abundance, environmental measurements and the four habitats to better understand the ecology of these species in a channelized river system. Ohio Shrimp were most abundant in the open side channels, while Glass Shrimp were most abundant in the main channel border wing dike habitat. Thirty-two percent of the variance in Glass Shrimp abundance was explained by year 1995, year 1998, water temperature, depth of gear deployment, Secchi disk transparency and river elevation. Approximately 8% of variation in Ohio Shrimp abundance was explained by Secchi disk transparency. Catch-per-unit-effort (CPUE) was greatest in 1998 for Glass Shrimp but lowest in 1997. Conversely, CPUE was greatest in 1996 for Ohio Shrimp and lowest in 2000. Both species exhibited inter-annual variability in CPUE. Long-term impacts of river modifications on aquatic invertebrates have not been well documented in many large river systems and warrants further study. The findings from this study provide ecological information on Glass and Ohio Shrimp in a channelized river system.

Size limited brood capacity is common among species with hard exoskeletons or shells. In these species, brood size is limited by the physical capacity to hold offspring. Here we present evidence that brood size is limited by physical constraints in Sphaerium striatinum, a small brooding bivalve. Sphaerium striatinum is a sequential brooder and produces offspring throughout the year. Offspring are brooded in marsupial sacs located on the inner demibranch. In an unconstrained brooder one would predict that brood size would increase as a function of adult length cubed, a volumetric relationship. In S. striatinum, brood size increases as less than a function of adult length squared. We demonstrate that brood size is limited by two general constraints: marsupial sacs and the retention of extra-marsupial offspring. The number of marsupial sacs increases as less than a function of adult length squared. This relationship may be a result of physiological process such as feeding and respiration. Offspring size at independence is a crucial factor in determining offspring survivorship. The retention of extra-marsupial offspring promotes growth inside a safe environment and increases survivorship upon independence. However, the exponent relating brood size to adult length is significantly less for adults that contain extra-marsupial offspring than compared to adults that do not contain extra-marsupial offspring. Although the evolution of brooding in S. striatinum has resulted in severe constraints on brood size, the benefits of brooding outweigh the cost of limited brood capacity. We discuss our results in relation to brooding strategies and size limited brood capacity in other brooding bivalves.

Infestation by introduced zebra mussels has extirpated native unionids in many Great Lake habitats. Shallow areas in coastal wetlands are intermittently dewatered by seiches and seasonal water level changes, and we examined how water level fluctuations and sediment characteristics affected interactions among unionids and zebra mussels in a Lake Erie coastal marsh. In 2001 we sampled unionid distributions and measured zebra mussel colonization on PVC plates at 1 cm, 18 cm and >35 cm water depths. We found a diverse unionid community (15 species) with many juvenile unionids. Unionid densities (0.01 unionids/m²) were comparable to other coastal wetlands, but are lower than reported in offshore areas before zebra mussels were introduced. Zebra mussels colonized plates at >3000 individuals/m² in some locations. Although >60% of unionids had byssal threads on their shells, >75% of unionids had no attached zebra mussels. Therefore, zebra mussels are colonizing unionids, but are not surviving. Unionid numbers and zebra mussel colonization were low in shallow (1–35 cm) water depths, indicating that water level fluctuations limited their distributions. Only two species of unionids were collected in 1–17 cm deep areas, and areas that became mudflats in September had almost no unionids. Numbers of zebra mussels and unionids were not correlated with organic content or silt/clay content of the sediments. Habitat characteristics shared by this wetland and other coastal wetlands that are important refuges of unionids include: a hydrological connection with the lake, areas deep enough for unionids to survive low water levels and soft sediments that allow unionid burrowing.

We simulated populations of threeridge mussels (Amblema plicata, Say, 1817) using empirically derived survival rates. Survival rates were derived from a mark-recapture study of mussels conducted in the Mississippi and Otter Tail rivers, Minnesota. The software package VORTEX was used to develop a base line population model which changed little or not at all through time. Models based on the initial base line data were constructed to simulate behaviors of A. plicata populations under various levels of mortality attributed to zebra mussel (Dreissena polymorpha, Pallas 1771) colonizations. Models were also developed that simulated varying levels of commercial harvesting of A. plicata. Sensitivity analysis revealed that A. plicata populations are most sensitive to alterations in adult survival. Adult survival is the life history stage most often impacted by both D. polymorpha colonizations and commercial harvesting. Population models simulating D. polymorpha mediated mortality revealed that population extinction was likely to occur within 50 y if survival rates remain at the levels recently measured. Models that simulated a commercial harvest of only 5% of the adult A. plicata population predicted that populations will decline by almost 50% in only 40 y. Our recommendation is that a concerted effort should be made to estimate the population size of potentially harvested populations of freshwater mussels. These estimates would allow for a more accurate estimate of what percentage of the adult mussels are actually being harvested. The population estimates would also allow for setting of quotas or limiting the number of permits for harvesters in the face of increased D. polymorpha infestations, thereby ensuring that harvested populations of mussels are not being over exploited. These modeling efforts will be useful for resource managers attempting to establish a sustainable harvest of commercially valuable mussel species which may be competing with a nonindigenous species.

In certain arthropod groups, including spiders, males seeking copulations may expose themselves to cannibalism from females that are larger and stronger than they are. Although old males are commonly believed to be particularly vulnerable to sexual cannibalism, virtually no data exist to back that supposition. Female crab spiders Misumena vatia (Thomisidae) regularly attacked prospective mates experimentally presented in pairs whose individuals differed in age. They usually attacked the older male; all but one of the seven males cannibalized in these encounters was the older member of the pair. These attacks paralleled a decline in numbers of field-observed males, probably a consequence of their increasing vulnerability. The females' responses appeared to result from differences in male behavior or condition, since virgin females, wild-captured adult (and probably previously mated) females and penultimate females all attacked males with similar frequency.

Veery (Catharus fuscescens) nest sites were compared to unused sites in a Middle-Atlantic Piedmont forest to determine if nest placement was random or biased with respect to forest structure and alien vegetation. Thus far, available data shows alien plants have a detrimental or neutral effect on the ecology of forest birds; however, empirical data regarding the proximate influence of invasive alien shrubs on avian nest placement in North American forests is lacking. Nest sites were distributed non-randomly in relation to vegetation density and were characterized by dense foliage below 1.5 m with sparse overstory at 2.5 to 3 m. Sites occurred within moist forest in floodplains and on south- and east-facing slopes. All nest sites contained alien shrubs, and alien vegetation supported 84% of nests. Shrub diversity did not differ between nest sites and unused sites yet more alien shrub species were found at nest sites. The density of native shrub-layer foliage did not differ between the two treatments; however, the density of alien shrub foliage was greater at nest sites. In this forest there was no relationship between the density of alien vegetation and the density of native vegetation. These data suggest that alien shrubs have replaced native shrub species and exerted a largely additive effect on foliage density providing the proximate cues for nest placement. The high success rate (70%) of nests within sites used in analyses suggests that the alien shrubs providing these cues are not substantially elevating nest failure rates. Thus, some temperate-breeding Neotropical forest-interior birds may react positively to a change in forest structure resulting from the invasion of alien shrubs. Ecological release resulting from the increase in available nest sites created by alien shrubs may explain the recent regional spread of the Veery. Region-specific studies are needed to determine the forest breeding birds that are affected, either positively or negatively, by the altered spatial heterogeneity created by alien shrubs.

Declines in loggerhead shrike populations have been attributed to pesticide use and habitat loss on the breeding grounds and factors outside the breeding range. To determine the role of breeding habitat limitation, Brooks and Temple (1990) designed a habitat suitability index for shrikes based on data from Minnesota. This paper describes an application of their model to a site in Illinois. Like Brooks and Temple, I found that breeding habitat does not appear to limit shrike populations and shrikes seem to be making settlement choices based on discernable habitat criteria. I suggest changes to the model for adaptation to Illinois shrike populations, including an adjustment of the cutoff for “suitable” habitat, an adjustment of the conversions of variables leading to the calculation of the index (V4 to SI4), the use of GIS to measure variables (usable foraging habitat) and the addition of variables (length of fence) used in the model.

Allegheny woodrats (Neotoma magister) currently receive protected status throughout their range due to population declines. Threats associated with habitat fragmentation (e.g., introduced predators, disease, loss of connectivity among sub-populations and habitat loss) may explain why Allegheny woodrats are no longer found in many areas where they existed just 25 y ago. In southern West Virginia, surface coal mining is a major cause of forest fragmentation. Furthermore, mountaintop mining, the prevalent method in the region, results in a loss of rock outcrops and cliffs within forested areas, typical habitat of the Allegheny woodrat. To determine the extent that Allegheny woodrats make use of reclaimed mine land, particularly rock drainages built during reclamation, we sampled 24 drainage channels on reclaimed surface mines in southern West Virginia, collected habitat data at each site and used logistic regression to identify habitat variables related to Allegheny woodrat presence. During 187 trap nights, 13 adult, 2 subadult and 8 juvenile Allegheny woodrats were captured at 13 of the 24 sites. Percent of rock as a groundcover and density of stems >15 cm diameter-at-breast-height (DBH) were related to Allegheny woodrat presence and were significantly greater at sites where Allegheny woodrats were present than absent. Sites where Allegheny woodrats were present differed substantially from other described habitats in West Virginia, though they may simulate boulder piles that occur naturally. Our findings suggest the need for additional research to examine the dynamics between Allegheny woodrat populations inhabiting rock outcrops in forests adjacent to mines and populations inhabiting constructed drainage channels on reclaimed mines. However, if Allegheny woodrats can use human-created habitat, our results will be useful to surface mine reclamation and to other mitigation efforts where rocky habitats are lost or disturbed.

Little is known about nest tree use of the endangered Virginia northern flying squirrel (Glaucomys sabrinus fuscus). Because nesting sites could be a limiting factor, it is important to understand the denning ecology to further manage and protect this subspecies. We compared characteristics of nest trees used by Virginia northern flying squirrels with randomly selected trees during summer and fall of 2000 and 2001. We tracked 13 Virginia northern flying squirrels to 59 nest trees. The squirrels used an average of 3.6 nest trees/month, switching trees frequently. Sixty-nine percent of the nests were in cavities and 31% were leaf nests. Yellow birch (Betula alleghaniensis) and Fraser magnolia (Magnolia fraseri) were selected for nest trees more than expected based on availability. A large portion of nest trees were in larger and taller trees than trees in the surrounding area. There also was a significant number of trees located next to or near skidder and hiking trails. Although a large number of nest trees were similar across sites, there was variation in the characteristics of nest trees used, suggesting that Virginia northern flying squirrels may not be as specialized in nest tree selection as indicated by previous studies.

We examined the abundance and placement of leaf nests by fox squirrels in six urban woodlots in central Indiana ranging in size from 1.06 to 8.28 ha. Four of the woodlots were disturbed, or subject to extensive human impact, whereas the remaining two were nature preserves. We counted all leaf nests present in each woodlot and recorded nest tree characteristics. We then conducted a quantitative vegetation analysis of trees present and estimated percentages of herbaceous and shrub cover along a minimum of two 100 m transects at each site. Fox squirrels showed a preference to build nests in certain species of trees. However, preference for nest tree species was not consistent across sites. Fox squirrels preferred to build nests in large trees with vines in the canopy at all sites. Characteristics of nests and nest trees did not differ among sites, but nest density was greater in the disturbed sites compared to the nature preserve sites. The nature preserve sites differed from the disturbed sites only with regard to the amount of shrub and herbaceous cover; shrub cover was greater and herbaceous cover was less at the disturbed sites. Results of this study suggest that fox squirrels are flexible with regard to nest tree species used and that the choice of a nest tree is dependent, in part, on tree size and the presence of vines. Further, a higher density of leaf nests in disturbed woodlots suggests that habitat disturbance and fragmentation due to urbanization may not have detrimental effects on the abundance and persistence of fox squirrels.

Badger (Taxidea taxus) activity was recorded in colonies of yellow-bellied marmots (Marmota flaviventris) during behavioral observations and trapping. Badgers were observed seven times in a marmot colony and extensive digging at marmot burrows was recorded five times in 40 y. When four badgers occupied a burrow at the edge of a marmot colony, the behavior of marmots and badgers was observed for 7 d and the marmots for an additional 6 d after the badgers departed. Badgers hunted at a significantly higher rate in the afternoon than in the morning. The adult female usually hunted alone, failed to capture any adult marmots, but did kill young marmots. When badgers were present, marmots had more frequent alarm calls, lower rates of foraging and higher rates of vigilance than when badgers were absent. I estimated that badgers killed 67 of 1423 individual marmots. The risk of being killed by a badger was highest for young, intermediate for yearlings and lowest for adults. Badgers probably have little impact on marmot demography except for localized, short-term increases in mortality, but may significantly affect the fitness of individual marmots.

We surveyed small mammal assemblages at 20 high-elevation wetlands in West Virginia and Maryland and examined relationships among mammal capture rates, richness and evenness and landscape features at multiple spatial scales. In 24,693 trap nights we captured 1451 individuals of 12 species. Small mammal species richness increased with wetland size and was negatively correlated with trail density. Generalists, such as meadow voles (Microtus pennsylvanicus) and shrews (Sorex cinereus, Blarina brevicauda), dominated larger, more open wetlands, whereas southern red-backed voles (Clethrionomys gapperi) were more prevalent at smaller sites surrounded by mixed coniferous-deciduous forest stands. Furthermore, meadow voles were captured more often at sites with higher road density and lower trail density. Southern bog lemmings (Synaptomys cooperi) were captured at less than half the sites, all of which were surrounded by a high proportion of deciduous forest. Although significant relationships were found, landscape features explained <20% of total variation at any spatial scale. Other factors, such as land use history or competition, likely have exerted a greater influence in small mammal abundance and distribution at these sites.

Little is known about the small mammalian fauna of Prince Edward Island, particularly shrews. Although historical data indicate the presence of five shrew species on PEI (Sorex cinereus, S. fumeus, S. hoyi, S. palustris and Blarina brevicauda), recent studies have failed to capture S. fumeus, S. hoyi and S. palustris on the Island. During a study designed to examine the abundance of amphibians in 11 forest fragments located in the central region of Prince Edward Island, we collected 344 shrews in pitfall traps. Sorex cinereus was the most abundant species captured, and was present in all forest fragments. Blarina brevicauda was captured in all but one forest fragment. Sorex fumeus was found in only one fragment. Sorex hoyi and S. palustris were not captured during this study. No significant correlations were found between physical characteristics of forest remnants (i.e., area, perimeter, ratio area:perimeter) and the total abundance, species richness or species diversity of shrews. However, shrew captures were correlated with various biotic characteristics including ground temperature, the presence of canopy cover, number of stumps and number of logs. We conclude that both B. brevicauda and S. cinereus are common and widespread-distributed on Prince Edward Island. In conjunction with other recent studies, our findings indicate that S. hoyi is extremely rare on the Island and that it is possible that the species is extirpated from Prince Edward Island. We stress the need for a more comprehensive study specifically designed to determine the abundance, distribution and conservation status of shrew species on Prince Edward Island.

White-footed mice, Peromyscus leucopus, were removed from a riverine peninsula to assess the effects of a river functioning as a barrier to movement. During an 8-mo period in 2001, 101 mice were live trapped and released across a river from the 14.2-ha peninsula. No small mammals were found to emigrate from the peninsula; 6 of the mice removed, however, exhibited homing behavior and immigrated across a fifth-order river and returned to or near the original site of capture. More studies at this temporal and spatial scale need to be conducted to demonstrate how natural barriers, particularly riverine peninsulas, could be utilized as experimental mesocosms in the study of small-mammal population dynamics at the landscape scale.

Black-tailed and white-sided jackrabbits (Lepus californicus and L. callotis) occur sympatrically throughout much of the Chihuahuan Desert; black-tailed jackrabbits are widely distributed, whereas white-sided jackrabbits are endemic to the Mexican Plateau. I measured abundance, distribution and habitat associations of each species in western Chihuahua, Mexico, during the winters of 1998 and 1999. No difference was observed in mean density; however, black-tailed jackrabbits were more abundant in semi-desert grasslands, whereas white-sided jackrabbits were most abundant in high elevation, plains grasslands. The occurrence of the two species was negatively correlated, a result of different habitat preferences or competition. Black-tailed jackrabbits were solitary, whereas white-sided jackrabbits were usually observed in pairs (74 and 58% of the time for 1998 and 1999, respectively). Dry conditions and heavier grazing pressure present in 1999 accentuated differences in characteristics of sites used and not used by jackrabbit species. The presence of black-tailed jackrabbits was positively associated with vegetation density in 1998 and positively associated with shrub density in 1999. White-sided jackrabbit presence was positively associated blue grama grasslands in 1999. Remaining open patches of desert grassland are important for the persistence of white-sided jackrabbit populations.