The canopy and understory vegetation of 19 aspen (Populus tremuloides) stands near Crested Butte, Colorado, were characterized in 1964 and were resampled in 1994 to assess the successional status of aspen communities. After 30 yr, aspen basal area was unchanged although the total number of trees decreased by 23%. There were 42% more aspen over 18 cm dbh. Understory species remained relatively unchanged; Thalictrum fendleri, Lathyrus leucanthus, Carex geyeri and Ligusticum porteri, the four most dominant species, as measured by cover, in 1964 were also abundant in 1994. Conifer basal area and the total number of trees remained constant over 30 yr, though the number of stands with conifers increased from six to 10. These results suggest that the aspen communities in the Crested Butte area are persistent, and while Picea engelmannii and Abies lasiocarpa may gradually replace some aspen stands, particularly on N-facing slopes, this replacement may take centuries.

To recover direct evidence of surface fires before European settlement, we sectioned fire-scarred logging-era stumps and trees in 39 small, physically isolated sand patches along the Great Lakes coast of northern Michigan and northern Wisconsin. While much information was lost to postharvest fire and stump deterioration, 147 fire-free intervals revealed in cross-sections from 29 coastal sand patches document numerous close interval surface fires before 1910; only one post-1910 fire was documented. Cross-sections from the 10 patches with records spanning >150 yr suggest local fire occurrence rates before 1910 ca. 10 times the present rate of lightning-caused fire. Since fire spread between or into coastal sand patches is rare, and seasonal use of the patches by Native people before 1910 is well documented, both historically and ethnographically, ignition by humans probably accounts for more than half of the pre-1910 fires recorded in cross-sections.

To document how species richness and diversity (H′) recover from severe large-scale disturbance, we report temporal patterns of species composition and diversity following grass-to-forest succession from a long-term experiment in the Coweeta Basin, western North Carolina. The original experiment—clear-cutting, 5 yr of grass cover followed by a herbicide treatment, and abandonment in a Southern Appalachian mixed deciduous forest—represents the most severe human disturbance in the Coweeta Basin. For several years after cessation of management, Robinia pseudoacacia quickly sprouted from roots and exceeded the growth rates of other species. Liriodendron tulipifera increased in density and basal area because of its prolific seedling establishment and rapid growth rate. Regeneration of large seeded species was mixed—sparse for Quercus rubra and Q. coccinea and nonexistent for Q. prinus and Q. velutina. In the overstory, density-based H′ increased from 1958, before grass conversion, to 15 yr and 28 yr following disturbance. In contrast, basal area-based H′ had significantly declined at 15 yr, then increased at 28 yr. The initial decline in basal area-based H′ was attributed to a decline in evennness of species distribution (J′) rather than to a change in species richness. The severe disturbance increased the abundance of early successional woody species and of herbaceous genera that tolerate open habitats, such as Erichtites, Phytolacca, and Erigeron. Shade-tolerant understory ferns and herbs such as Polystichum acrostichoides, Dennstaedtia punctilobula, Galium latifolium and Viola cucullata gradually became more abundant. The 28-yr-old forest of WS6 had much lower species richness than the adjacent reference watersheds, but more than threefold higher density.

Radial growth (indices of tree-ring widths) of Eastern cottonwood (Populus deltoides Marsh.), a flood-tolerant species, and black walnut (Juglans nigra L)., a flood-intolerant species, were compared with climate and streamflow data by stepwise multiple regression for a riparian forest along the Olentangy River in central Ohio to investigate: (1) the influence of these variables on riparian primary productivity; and (2) whether these species can be used as predictors for the reconstruction of streamflow. Black walnut showed a significant growth response to streamflow. There was no relationship between growth of Eastern cottonwood and streamflow. Growth relationships with climate (mean monthly air temperature, total monthly precipitation) and Palmer Drought Severity Index (PDSI) showed black walnut growth on an upland riparian site was more sensitive to temperature and PDSI than on a lower-elevation site. Growth of black walnut on the low-elevation riparian site was directly and positively influenced by increased streamflow in midwinter and in late summer when subsurface soil moisture recharge from streamflow may be a critical moisture source. Growth of Eastern cottonwood on the low-elevation site was not related to streamflow, but was reduced by excess summer precipitation.

Concentrations of heavy metals in old-field plant species were compared among untreated reference subplots, unlimed sludge-treated subplots and limed sludge-treated subplots during the 4th yr of ecosystem recovery following long-term (11-yr) nutrient enrichment. Although results varied on a metal-specific and species-specific basis, liming was overall successful in reducing metal concentrations in old-field plant species. For example, concentrations of Cd in leaves and roots and concentrations of Pb and Zn in roots of Ambrosia trifida were significantly lower in plants collected from limed compared to unlimed sludge-treated subplots. Liming significantly reduced concentrations of Cd in leaves and roots of Solidago canadensis, and in the leaves, roots, and seeds of Setaria faberii to levels found in untreated reference subplots. Liming also significantly reduced concentrations of Cu in the roots of Solidago canadensis, Pb in the roots and seeds of Setaria faberii, and Zn in the roots and leaves of S. faberii to reference levels. Changes in plant species composition during secondary succession may also indirectly reduce metal uptake since the annual Ambrosia trifida accumulated Cd and Zn in leaves, roots and seeds at levels two to three times greater than did the perennial Solidago canadensis.

Buffalograss (Buchloe dactyloides) has been described as an excellent fit for Janzen's “Foliage is the Fruit” (FF) hypothesis, which suggests that large grazing animals ingest and later disperse seeds of some herbs when consuming their foliage. We tested this hypothesis by feeding buffalograss burs and legume seeds to ruminally fistulated beef steers. Our objectives were to determine (1) rumen residence times of buffalograss burs, free buffalograss caryopses and legume seeds; (2) total tract residence times for the three types of propagules; (3) percentage propagule survival after passage and (4) germinability of fed and unfed burs and caryopses. Bur survival (3%) and germination percentage were lower than previously reported, but this can be explained by differences in feeding and germination procedures. Significant numbers of free caryopses recovered from the feces also germinated. Previously undescribed hairs on the burs' awn-like projections delayed passage through the animals and assured that many burs were broken during rumination. We concur that buffalograss is an excellent fit for the FF hypothesis, but our results and observations suggest that the buffalograss/grazing-animal interaction is more complex than it initially appeared.

Foraging behavior of the southern flying squirrel was studied under conditions of declining photoperiod and temperature. Experiments tested predictions derived from a theoretical model suggesting hickory (Carya spp.) nuts as the optimal food for flying squirrels at temperatures below −10 C. Above this temperature flying squirrels could meet their daily energy needs more quickly on a diet of acorns. During a declining photoperiod and mild temperatures flying squirrels stored hickory nuts whereas acorns were usually eaten at the site of discovery. Test animals foraging at low temperatures significantly increased the number of hickory nuts consumed and decreased the number of acorns consumed. By eating acorns in the autumn and storing hickory nuts for winter consumption flying squirrels make use of a dynamic foraging strategy that allows optimal use of both autumn and winter food resources.

During 1991–1993 we radio-collared and monitored 50 porcupines (Erethizon dorsatum) on two study sites in central Massachusetts to quantify habitat use. During summer, porcupines were found resting mostly on tree limbs (≥ 80% of locations) in daylight hours. During winter, they rested mostly in dens (e.g., rock caves, cavities in trees) (≥70%) in daylight hours. Where talus was abundant, porcupines rested in dens consisting of rock caves (90%). Where talus was absent, they rested in cavities in trees (50%) more than rock caves (27%). During summer, their use of oaks (Quercus spp.) (60 and 50% of daytime locations in each of two study areas) for feeding and resting was more than expected, based on the abundance of oaks in the two areas (corresponding abundance = 8 and 18%, respectively). During winter, their use of hemlocks (Tsuga canadensis) for feeding and resting when not in dens (93 and 71% of daytime and nighttime locations, in each area, respectively) was more than expected (6% in each area). The diameter at breast height (dbh) of red oaks (Quercus rubra) (x̄ = 52 cm) used by porcupines was larger than a random sample in the area (x̄ = 27 cm). The dbh of hemlocks used for resting and feeding (x̄ = 45 cm) was larger than the dbh of hemlocks used for feeding only (x̄ = 33 cm) and was also larger than a random sample (x̄ = 35 cm). Porcupine selection of tree species varied between study areas according to availability of tree species. Their selection of den types also varied between areas according to availability. This study shows that porcupines are adaptable during the summer but are more restricted in their habitat use during the winter. The loss of hemlock may increase mortality in this population due to winter stress and predation.

Dietary preferences of Indiana bats were determined by analyzing 382 fecal pellets collected beneath roost trees in southern Michigan, over parts of 3 yr. Although terrestrial insects (Lepidoptera and Coleoptera) usually dominated the diet of Indiana bats in more southern states, those in Michigan consumed mostly insects associated with aquatic environments. Indiana bats in Michigan ate primarily Trichoptera (55.1% of volume) and Diptera (25.5%), followed by Lepidoptera (14.2%) and Coleoptera (1.4%). Consumption of Diptera was highest during lactation (48.2%), whereas consumption of Lepidoptera was least during this time (7.7%). Although most insectivorous bats do not prey on mosquitoes (Culicidae), these insects were a consistent component of the diet of Indiana bats and were eaten most heavily during pregnancy (6.6%).

On 4 October 1995, Hurricane Opal made landfall to the E of one of our long-term study sites at Bon Secour National Wildlife Refuge in Baldwin County, Alabama. The primary dunes were destroyed and seawater covered much of the site for several days after the storm. This natural event provided a unique opportunity to study changes in the dynamics of a population of Alabama beach mice (Peromyscus polionotus ammobates) following a catastrophic alteration to a primary component of their habitat mosaic. Mouse numbers at their posthurricane peak were 30% lower than those observed prehurricane. However, the probability of mice surviving between the bimonthly trapping period which included the hurricane was not significantly different from the average between period estimates. Overall, posthurricane survival was not significantly lower than prehurricane values. However, we observed a potential delayed effect in the form of unusually low survival during the summer of 1996 which coincided with a significant reduction in body mass. Approximately 47% of recaptures on the scrub/transition habitat in December 1995 were of mice originally marked in the beach habitat. This trend continued through April 1996 suggesting that the scrub/transition areas may provide a refuge after tropical storms. While Hurricane Opal was destructive, the immediate demographic changes were not significantly different from those driven by annual climatic cycles. The hurricane may have compromised the ability of the population to survive the difficult summer season.

To assess the impact of zebra and quagga mussel (Dreissena spp.) infestation on unionids, unionids (Bivalvia: Unionidae) were sampled in the Detroit River in 1982–1983, before mussels invaded the river, and in 1992 and 1994, after mussels invaded the river. Live unionids at four stations along the southeastern shore accounted for 97% (20 species) of all shells collected in 1982–1983, whereas live unionids accounted for only 10% (13 species) in 1992. A similar decline in live unionids occurred at nine stations along the northwestern shore, except the decline occurred over the three sampling periods: in 1982–83, 84% (22 species) were live; in 1992, 65% (26 species) were live; and, in 1994, only 3% (13 species) were live. The difference in time to near-total mortality of unionids along the southeastern and northwestern shores is attributed to differences in the time of invasion and abundance of zebra mussel veligers in distinct water masses emanating from Lake St. Clair located immediately upstream of the Detroit River. Although individuals of all species of all unionid subfamilies declined between 1982 and 1992/1994, members of the subfamilies Anodontinae and Lampsilinae declined more than Ambleminae. Between 1986 and 1992/1994, five Anodontinae, three Lampsilinae and 0 Ambleminae species have been extirpated from the river due to dreissenid mussel infestation. Numbers of individuals of commonly found species declined more than numbers of individuals of uncommonly found species. However, the number of uncommon species declined 47% (17 to 9) along both the southeastern and northwestern shores, whereas common species remained the same (3 species) along the southeastern shore and declined only 40% (5 to 3 species) along the northwestern shore. This study, and others, suggest that high mortality of unionids can occur between 4 and 6 yr after initial invasion by dreissenids or up to 8 yr depending on water current patterns. Infestation-induced mortality of unionids in the Detroit River is similar to that observed at a few locations in other rivers, but is higher over a larger area than that measured in other rivers to date, probably because the Detroit River was the first to be colonized by dreissenid mussels in North America.

Clear-cut logging of boreal riparian forests in northwestern Ontario, Canada shifted the composition of allochthonous leaf litter from largely coniferous to largely deciduous broadleaf. Because of rapid regrowth after tree harvesting, long-term changes in the composition rather than in the amount of allochthonous litterfall has been suggested as being the more serious environmental consequence of riparian deforestation. Macroinvertebrates within the littoral zones of four oligotrophic Canadian Shield lakes actively colonized leaf litter for both food and habitat. Food-related, early-stage (1 wk to 1 mo) colonization abundances were never significantly greater in coniferous (black spruce—Picea mariana, jack pine—Pinus banksiana, and eastern white cedar—Thuja occidentalis) litter than in deciduous (trembling aspen—Populus tremuloides, white birch—Betula papyrifera, and speckled alder—Alnus rugosa) litter. Late-stage (10 mo to 1 yr) colonization abundances were unrelated to differences in the breakdown rates of coniferous and deciduous litter. Therefore, shifts in tree species composition following riparian logging of boreal forests should not be detrimental to the long-term abundance of littoral macroinvertebrates colonizing terrestrial litter.

We describe massive autumn migrations of dragonflies (Odonata) which occurred at Chicago, Illinois (14 September 1978), Cape May, New Jersey (11 September 1992), and Crescent Beach, Florida (3–5 September 1993). Estimated numbers of migrant dragonflies involved in these flights were approximately 1.2 million, >400,000, and 200,000, respectively. We also document other recent observations of large swarm migrations of dragonflies in eastern North America, review previous reports of this phenomenon, and compare these events to the flights at Chicago, Cape May and Crescent Beach. Records of large dragonfly migrations show several distinct patterns: (1) all reports fell between late July and mid-October, with a peak in September; (2) most of the large flights occurred along topographic leading lines such as coastlines and lakeshores; (3) massive swarm migrations generally followed the passage of synoptic-scale cold fronts; and (4) the common green darner (Anax junius) was the principal species involved in the majority of these flights. Striking parallels between the patterns of seasonal timing, geographical distribution, and meteorological context of dragonfly migrations and those of birds suggest that similar causal factors are involved.

This study documents the distribution of the Argentine ant (Linepithema humile), a well known invader of urban and agricultural ecosystems, in a riparian corridor near San Luis Obispo, California. Sampling of the ant community was done using pitfall traps placed along transects that crossed three vegetation types found at the study site: (1) riparian woodland dominated by coast live oak (Quercus agrifolia), poison oak (Toxicodendron diversilobum), and willow (Salix spp.) (2) Hard chaparral dominated by scrub oak (Q. dumosa) and silk-tassel bush (Garrya sp.) (3) Coastal scrub dominated by black sage (Salvia mellifera), California sagebrush (Artemisia californica), and Yucca (Yucca whipplei). Linepithema humile has become well established throughout the lower reaches of the study site except in the chaparral vegetation. Ten species of native ants were found in the study area, but there was little to no correlation or cooccurrence between L. humile and native ant species. Total species diversity in areas where L. humile occurred was significantly less than in areas where L. humile was absent. These results suggest that invasion by L. humile is restricted by habitat type, and has significant negative impacts on native ant diversity.

We studied patterns of association within and between two species of Pachypsylla gall insects (P. celtidismamma and P. sp. A) on hackberry (Celtis occidentalis L.) trees in eastern Iowa. Pachypsylla is a recently radiating clade of jumping plant lice (Homoptera: Psyllidae). Exploitative competition has been invoked to account for niche diversification in Pachypsylla as new lineages switched from leaves to feeding on different plant parts. Our results suggest that competition did not drive diversification, but could have reinforced it. Focal P. celtidismamma gallmakers grew larger when they shared a leaf with more conspecific gallmakers, which is consistent with intraspecific facilitation. Facilitation may result because gallmaker aggregations act as physiological sinks for photosynthate originating elsewhere in the plant. Pachypsylla celtidismamma gallmakers were smaller when they shared their gall with more Pachypsylla sp. A, which live as inquilines inside P. celtidismamma galls; this is consistent with the occurrence of interspecific competition. At spatial scales larger than a single leaf, we detected no significant relationships between larval densities and performance.

Checklist and “Pollard Walk” butterfly survey methods were contemporaneously applied to seven public sites in North Dakota during the summer of 1995. Results were compared for effect of method and site on total number of butterflies and total number of species detected per hour. Checklist searching produced significantly more butterfly detections per hour than Pollard Walks at all sites. Number of species detected per hour did not differ significantly either among sites or between methods. Many species were detected by only one method, and at most sites generalist and invader species were more likely to be observed during checklist searches than during Pollard Walks. Results indicate that checklist surveys are a more efficient means for initial determination of a species list for a site, whereas for long-term monitoring the Pollard Walk is more practical and statistically manageable. Pollard Walk transects are thus recommended once a prairie butterfly fauna has been defined for a site by checklist surveys.

We studied the relationship between breeding birds and seral stages of green ash (Fraxinus pennsylvanica) woodlands in central South Dakota between 1990 and 1992. Stands of early seral green ash undergoing primary succession had few small trees with western snowberry (Symphoricarpos occidentalis) in the understory. Some early seral green ash stands resulted from retrogression and had large trees with grass understory. Late seral green ash stands were represented by greater overstory cover consisting of green ash and chokecherry (Prunus virginiana) shrubs and small trees in the understory. Sixty-five of 81 bird species that were observed in these woodlands required woodland vegetation as part of their habitat. Tree- and shrub-nesting birds were the most common and were associated with late seral stages. Ground-nesting birds had mixed relationships between early and late seral stages. Cavity-nesting birds used snags and dead tree branches which occurred in all seral stages. Correlations of birds with vegetation measurements suggested habitat features birds may have selected for in these woodlands.

We compared age- and sex-class structure between two populations of eastern chipmunks (Tamias striatus) occupying study plots within uncut (reference) and managed (fragmented) forest sectors in central Pennsylvania from 1992–1995. In addition, we compared reproductive condition, weight and length of time chipmunks remained on study plots between the two populations. There was no difference (P > 0.05) in age class, sex class or reproductive condition between the populations occupying each sector. Chipmunks remained on plots in both uncut and managed forest sectors for an average of 2.2 and 2.3 mo, respectively. Weights of individual chipmunks did not differ (P > 0.05) between the two populations and averaged 81.5 g. Chipmunks in both populations, however, weighed significantly (P ≤ 0.05) less in 1993 following a poor acorn crop in 1992 than in other years. Although eastern chipmunks primarily inhabit mature eastern deciduous forests, fragmentation had no apparent influence on the demography of populations of this sciurid.

We determined maintenance nitrogen requirements of the adult prairie vole (Microtus ochrogaster) using a nitrogen balance trial approach. Prairie voles required 401 mg kg^−0.75 day⁻¹ of nitrogen to meet maintenance requirements and met requirements on diets containing approximately 4.4% crude protein. Dry matter intake did not vary with percentage of nitrogen in the diet, suggesting that animals on nitrogen-insufficient diets do not display appreciable compensatory intake. Body mass change was only moderately associated with nitrogen balance, indicating that fluctuations in body mass during the trial were likely associated with factors other than nitrogen intake.

Larvae of the tortricid moths Cydia lunulana and C. nigricana feed inside developing pods of Vicia spp. (Fabaceae: Papilionoideae) in North Yorkshire, England. In the common vetch Vicia sativa, larvae commonly consume or at least damage every single seed in the pod in which they develop. Seeds that are completely consumed cannot perpetuate the vetch species, but what about partially damaged seeds? Naturally damaged seeds were obtained from field-collected pods, and sorted into five categories: (1) no damage, (2) <10% damage (volume missing), (3) 10–25% damage, (4) 25–50% damage, and (5) 50–75% damage. Seeds were soaked and planted individually to monitor germination. While scarification of the seed coat of perfect seeds is beneficial to imbibition and germination, the percent of germination decreased with each successive damage category. That a substantial proportion of damaged seeds do germinate suggests that seed damage is not always a death sentence; but growing conditions in North Yorkshire are likely to favor the survival of seeds with seed coat intact to prevent germination in the autumn rains and plant death (before reproduction) in the winter frost.

We examined the responsiveness of adult gray rat snakes to avian prey in enclosures that simulated natural habitat conditions. Subjects searched for arboreal nest contents in the presence or absence of a bird model flown so as to simulate avian provisioning behavior. In the absence of the model, gray rat snake latency to capture prey was over 92 % greater than when searching for nest contents with the model. Snakes responded to model flight by turning their heads in the direction of its movement. We conclude that gray rat snakes perceive relatively distant visual cues from prey, and that nest provisioning (activity localized around a nest) may facilitate snakes correctly choosing to ascend trees that contain active nests and ignoring those that do not.