The food habits, movement and activity patterns, and habitat use of sympatric carnivores, and wildlife harvest and utilization, were studied at the Taohong Village, Jiangxi Province, southeastern China during 1992–1996. Food habits of four species of sympatric carnivores were studied by scat analysis. Crab-eating mongooses (Herpestes urva) had the most diversified diet that included many water-edge food items. Masked palm civets (Paguma larvata) consumed a large proportion of fruits. Both small Indian civets (Viverricula indica) and hog badgers (Arctonyx collaris) fed heavily on rodents. Though there was a certain degree of difference in habitat use among these species, the rather high diet overlap suggested that the population densities of these species were reduced to a very low level by constant harvest pressure and secondary poisoning so that the diet ceased to have relevant effect on the sympatry of these species. The food habits (based on scat analysis) of the sympatric dhole (Cuon alpinus) and wolf (Canis lupus ) showed that their depredation on the endangered Sika deer ( Cervus nippon kopschi), domestic animals and small carnivores was negligible. Telemetry studies showed that small Indian civets, masked palm civets, and crab-eating mongooses did not have permanent dens but moved among their numerous daybeds. Small Indian civets used daybeds among the bushes and grass on the ground while other species used underground dens exclusively. Small Indian civets and crabeating mongooses limited their activity to the foothills at low altitude, while masked palm civets included habitat of higher altitude in their home ranges. Both small Indian civets and masked palm civets were nocturnal, but the latter had a low activity level in the daytime. Crab-eating mongooses were active in the daytime. Ferret badgers (Melogale moschata) were strictly nocturnal. Their daybeds included both natural and man-made sites. The lack of direct conflict of interest with humans allowed them to live in close proximity to human settlement. Wildlife harvest during the 1992/96 harvest seasons was studied by the method of participant observation. Wildlife harvesters comprised 1.5% of the local population, and a few professional harvesters accounted for a large proportion of the game yield. Shotguns and two kinds of traps were the most common harvest methods used. Muntjac (Muntiacus reevesi) and hares (Lepus sinensis and L. capensis) were the most important game species. In spite of the steady increase in the price of wildlife parts, the game yields experienced a gradual decline due to reduced harvest efforts. Wildlife harvest was market-oriented and played an insignificant supplemental role in the local economy. Marketing channels for both pelt and game meat have been well established and are spreading. Wildlife harvest remained largely unregulated. The prospect of wildlife harvest at Taohong is discussed and recommendations to control harvest are proposed.

Frankliniella occidentalis Pergande (Thysanoptera: Thripidae) is the most important insect pest for greenhouse flower crops. Thripinema nicklewoodi Siddiqi (Tylenchida: Allantonematidae) is an entomoparasitic nematode attacking and sterilizing the thrips. Methods to propagate and study T. nicklewoodi were developed. I observed an excretion rate of 21.4 nematodes per day by parasitized female thrips. The sex ratio of the excreted nematodes was 6:1 (female:male). After exposing 50 healthy first instars to four parasitized female thrips in a rolled bean leaf, I obtained a 75.3% mean parasitization rate in the adult stage of the thrips. In contrast to previous reports, male thrips were found to be parasitized as readily as females. Parasitism reduced the longevity of both adult female and male thrips by 26% and 61%, respectively. T. nicklewoodi when presented with various thrips life stages achieved the highest attack rate in first and second instars and prepupa. Free-living nematodes were found to escape from hosts through the anus and penetrate new host thrips through the intersegmental membranes of the thorax and abdomen. While nematode parasitization affected tospovirus propagation, it did not reduce transmission of impatiens necrotic spot virus even though parasitism reduced feeding activity of adult female thrips by 81% on leaves, 38% on pollen, and 22% on honey. However, despite lowered total feeding, probing by parasitized thrips (in honey) was not reduced, and this may explain why lowered feeding does not result in lowered virus transmission. In a study of the population dynamics on caged impatiens in greenhouse, nematode transmission persisted for seven host generations and populations of normal (i.e., not parasitized) female thrips declined by 39–79% in the nematode treatment compared to the control. However, no significant reductions were found in numbers of larval thrips between nematode treatments and the control. A higher proportion of male thrips occurred in populations with nematodes in which adult female thrips declined significantly compared to the control population. T. nicklewoodi released seven times on caged impatiens in a greenhouse did not provide preventative control of thrips, though the population growth of second instar, adult female, and male thrips was suppressed by 44, 68, and 49%, respectively.

We investigated whether group-selection provides habitat of similar quality for early-successional shrubland-nesting birds as does clearcutting. There was no difference in avian reproductive success between clearcuts and groupcuts, indicating that clearcuts and groupcuts provide habitat of similar quality for early-successional shrubland-nesting birds. We studied Chestnut-sided Warblers (Dendroica pensylvanica) nesting in 29 patches of regenerating northern hardwoods forest 0.15–0.69 ha in area to determine whether the ecology of this species is affected by patch size or shape. Chestnut-sided Warbler density was higher, and pairs initiated nesting later in smaller patches, yet patch size was unrelated to fledging success or nest predation rates and patch area. Territory density, nest initiation dates, fledging success and nest predation rates were unrelated to patch shape. We compared nest predation rates between groupcut and clearcut borders (0–5 m from edges) and forest interior areas (45–50 m from edges) using artificial shrub nests baited with Zebra Finch (Taeniopygia guttata) eggs. The probability of a nest being depredated was higher in edge areas than forest interior areas (P = 0.02) and was independent of nest concealment, nest height, or whether the nest was adjacent to a clearcut or a groupcut (P = 0.18). We compared survival rates of used natural nests baited with House Sparrow (Passer domesticus) eggs with survival rates of active bird nests at the same sites. Survival rates of artificial nests (27.7%) were significantly lower than nest survival rates of natural nests (58.6%). We suggest that lack of parental defense is a contributing factor responsible for higher predation rates on artificial nests, and is likely to be a potential confounding factor in future nest predation experiments using artificial nests. We studied patterns of plumage variation in the Chestnut-sided Warbler to determine if plumage brightness was related to reproductive performance. There were no relationships between plumage brightness and reproductive success of either male or female Chestnut-sided Warblers. Adult males and females were brighter than subadult males and females, however, adult and subadult males and females fledged as many young as adult males and females. Thus, delayed plumage maturation in the Chestnut-sided Warbler is not associated with decreased reproductive output in subadult birds. We suggest that the duller plumage of subadult Chestnut-sided Warblers is more likely a reliable indicator of subordinate status, and that delayed plumage maturation serves in this species to reduce aggression from adult birds.

The cranberry fruitworm, Acrobasis vaccinii Riley (Lepidoptera: Pyralidae), is a serious pest of cranberry (Vaccinium macrocarpon Ait.). Reported hosts also include highbush blueberry ( Vaccinium corymbosum L.), lowbush sweet blueberry (Vaccinium augustifolium Ait.), and black huckleberry [(Gaylussacia baccata (Wang.) K. Koch], all of which grow in the woodlands (called uplands) surrounding Massachusetts bogs. Since a thorough understanding of movement patterns of a pest species within and between host habitats is crucial for the implementation of a successful management program, moth movement was studied in time and space through trapping of wild and mark-released moths. The results, reported in Chapters I and II, demonstrate that the moth is very mobile and that large flights occur in both cultivated bogs and their surrounding uplands. Trap studies also showed a significant effect of trap height on the number of moths captured. Moths were active later at night in the upland, and males, in particular, were active in the trees, which suggested adoption of a “hilltopping” strategy to enhance their chances of mate finding. In field and lab studies reported in Chapter III, the periodicities of emergence, mating, and oviposition were recorded, and the data suggest that it is advantageous for females to delay mating when conditions are favorable for oviposition. Females were polyandrous, and assessment of spermatophore age in females suggested that mating may occur more frequently in the upland compared to the bog. Studies reported in Chapter IV showed that moth activity and infestation occurred earlier in both blueberry plantings and in the uplands than cranberry. The large numbers of moths captured in the uplands did not appear to originate from larvae developing in the uplands, since infestation of wild berries could not be corroborated. A study of host preference and performance in four different hosts (cultivated cranberry and blueberry, wild blueberry and huckleberry) is reported in Chapter V. Huckleberry was the least preferred for oviposition, and both of the cultivated berries were more suitable for larval development when compared to the wild hosts. The relationship between preference ranking and performance was positive. Implications of these findings are discussed.

The federal-listed bog turtle (Clemmys muhlenbergii) is the smallest and rarest freshwater turtle species in North America. I studied bog turtles in New England from 1994–1997, examining habitat use and seasonal movements, breeding ecology, demographic characteristics and population viability. I made 1,553 captures, marked 75 adults and radiotracked 50 bog turtles. Home range sizes (0.06–2.79 ha) were similar to other studies; there were no detectable differences between sexes, although males and females used different habitats among seasons. Size of female at sexual maturity (plastral length = 74 mm) was based on reproduction instead of secondary sex characteristics. Clutch size ranged from 2–6 eggs (x¯ = 3.5), with individual variation among years. Incubation (74–103 days) was longer for Massachusetts sites compared to southern nests, and both hatchlings and adults were smaller in body size compared to southern populations. I modeled a stable population with estimated survival rates of 0.32 for hatchlings, 0.97 for adult females, size and age at sexual maturity of PL = 74 mm and 12 y, respectively, and a derived juvenile survival rate of 0.83 (assuming λ = 1.00). The cooler climate and shorter breeding season may place additional constraints on northern populations of bog turtles. These environmental factors may result in slower growth rates, delayed sexual maturation, smaller adult body sizes, iteroparity, and lower nest temperatures resulting in longer incubations of fewer and smaller hatchlings than those in the south. While biologists cannot manage for climate, we can implement protection of this threatened species by identifying and protecting important habitats for hibernation and nesting, preventing hydrologic changes to the system, maintaining open canopy cover, and reducing direct human and animal impacts on adults through monitoring and active management strategies. I suggest the best strategies for bog turtle conservation involve identification and protection of habitats occupied by reproducing populations rather than captive breeding and translocation.

Mixed-species bird flocks are ideal systems for investigating fundamental questions of community ecology. Traditionally, researchers have studied how flocking species avoid competition while foraging; other behavioral interactions have received less attention. I investigated the role information exchange plays in the organization of flocks in a rain forest in Sri Lanka. In particular, I focused on the two most frequent species: the highly gregarious leaf-gleaning Orange-billed Babbler Turdoides rufescens, and the less gregarious fly-catching Greater Racket-tailed Drongo Dicrurus paradiseus. Alarm calls are an important type of information shared among species. Babblers make the most frequent and rapid calls to natural and artificial stimuli; however, their calls are often made to non-raptors. Drongo alarm calls are more reliable and sensitive to real threats. Though playback experiments I asked two questions about alarm calls and flock organization. First, why does more than one species alarm call? Hypotheses include (a) conspecific calls receive greater response, (b) information in heterospecific calls is imperfect, and (c) there are direct benefits to calling. Hypothesis (a) is unlikely because in a playback experiment at least one species responded as much to heterospecific as to conspecific alarms. A second question was whether species prefer to associate with babblers or with drongos. Heterospecific birds were attracted to drongo playback as much as to babbler playback, demonstrating that an alarm-calling species may be as important to flock organization as a highly gregarious species. Another form of communication in flocks is practiced by drongos that vocally mimic other species. Drongo mimicry is contextual: drongos include other species' alarm calls in their own alarm vocalizations and other species' songs in their own songs. I hypothesized that drongos might reform flocks through mimicry, since they are well known to benefit by catching insects other species disturb. I found that twice as many heterospecific birds were attracted to playback of drongo mimicry than when mimicry was removed. Drongo mimicry is thus an example of unidirectional and perhaps manipulative information flow in flocks. Put together, the playback experiments demonstrate that birds use other species' vocalizations to find flocks—or drongos that sound like flocks.

Brachymeria intermedia is an introduced endoparasitoid of the gypsy moth, Lymantria dispar. To gain understanding of the basic mechanisms by which B. intermedia influence population densities of the gypsy moth, the oviposition behavior of this parasitoid was studied. Both the physiological state and the informational state of the animal at a certain point in time may influence the development of the behavior. Factors of the physiological state here investigated are egg load and age of the female parasitoid. Factors of the informational state here investigated are the number of hosts previously encountered, the host species encountered and characteristics of the sites where pupae previously were encountered. Chapter 1 contains the introduction and outline of the study. Chapter 2 shows that experience is an important factor influencing the rate of acceptance of hosts by the parasitoid. The sequence of behaviors before ovipositor insertion does not change with experience. Chapter 3 shows that parasitoids, deprived of hosts from emergence, accumulate eggs up to a certain level, but have a low rate of acceptance. Early exposure to pupae increases the rate of acceptance. B. intermedia adjust their egg production to host availability. Chapter 4 shows that the rate of acceptance for an alternate host increased with age in parasitoids, deprived of hosts from emergence. Rearing the parasitoids on an alternate host and/or oviposition experience on an alternate host did not influence the rate of acceptance of the alternate host. After oviposition experience on the alternate host contaminated with kairomone, uncontaminated hosts were accepted at the same rate as gypsy moth. Chapter 4 shows that B. intermedia is able to use learned visual cues, thereby restricting their searching area to a microhabitat previously found to be profitable. Under laboratory conditions and under semi-natural conditions, parasitoids could be trained to search for pupae either on the ground or on a tree model. Subsequent training to the other microhabitat reversed this effect. The studies show that both the physiological state and the informational state of the parasitoid influence oviposition behavior. Questions are raised and discussed as to the extent of polyphagy of B. intermedia in North America.

Increases in numbers of Herring (L. argentatus) and Great Black-backed (L. marinus) gulls, and concurrent decreases in numbers of other beach-nesting birds, led to the proposed use of avicides to control gulls on Monomoy NWR, Chatham, Massachusetts. The purpose of this study was to obtain biological information on Monomoy's gull populations, and to use this information to develop and evaluate gull control strategies. Landfill use differed between species. Herring Gulls used more, and traveled farther to, landfills than did Great Black-backed Gulls. Herring Gulls used fewer landfills during the non-breeding than breeding season. Numbers of Herring Gulls in landfills were significantly related to human populations served by those landfills (r$\sp2$ = 0.63, P = 0.0003), Great Black-backed Gull numbers were not. Although both species consumed refuse and natural foods, more Herring than Great Black-backed gull stomachs contained refuse $(X\sp2$ = 7.88, P $<$ 0.005 in 1988; $X\sp2$ = 13.93, P $<$ 0.0002 in 1989). Refuse was a minor dietary component for both species' chicks. The passage of large gulls over nest sites had little impact on Monomoy's Common Tern (Sterna hirundo) and Laughing Gull (L. atricilla) colonies. Few fly-overs (49, 0.04%) at altitudes of 20 m or less resulted in mobbings or disturbances. Terns mobbed large gulls more often than did Laughing Gulls (G = 16.61, P $<$ 0.005), but numbers of birds per mobbing did not differ between species (t = 0.95, P $>$ 0.05). Both species habituated to fly-overs and ignored most large gulls. We developed a deterministic simulation model to identify the effects of different types and levels of control on gull population dynamics. Effectiveness of gull control varied with technique, but no method completely eliminated Herring or Great Black-backed gulls. Model construction and interpretation of outputs are described.

This dissertation is the first critical analysis of patterns of species diversity in a deep-sea taxon (the Gastropoda) on a global scale. My analysis is based on 85 epibenthic sled samples collected from soft sediments in ten deep-sea regions of the Norwegian Sea and Atlantic Ocean. Over 21,000 gastropod individuals were sorted to species and their mode of larval development was documented. The basins from which the samples were taken represent a broad range of ecological circumstances that may affect species diversity and taxonomic composition. Bathymetric patterns of species diversity were shown to vary significantly among deep-sea basins. Unusually high or low values of diversity appear to be associated with environmental disturbance or rates of nutrient input. The most important and unexpected finding was that on a global scale there are latitudinal gradients in deep-sea species diversity similar to those in shallow-water and terrestrial environments. Gastropods show a clear and highly significant latitudinal decrease in diversity from the equator to 77$\sp\circ$N in the North Atlantic and a significant decrease from the equator to 37$\sp\circ$S in the South Atlantic. Depressed diversity at higher latitudes in the deep-sea, in part, may be related to seasonal nutrient loading and frequent physical disturbance from bottom currents. At lower latitudes, these same ecological factors may become less intense and variable. A multiple regression analysis showed that the North Atlantic local (sample) diversity is highly correlated with regional (basin) diversity suggesting that evolutionary-historical processes play a role in the development of deep-sea diversity through the mechanism of regional species enrichment. The percentage of species with planktotrophic development in the regional species pool was also correlated with local diversity. For the South Atlantic, the pattern of latitudinal diversity is less clear. On global scales, diversity gradients may be shaped by the interplay between the ecological potential and the history of evolutionary diversification at different latitudes. Underlying evolutionary processes that influence local diversity are difficult to recognize on a local scale, but may be revealed through large-scale patterns of species diversity. (Abstract shortened with permission of author.)

On Georges Bank, skates (F. Rajidae) appear to be replacing depleted populations of economically-important demersal fishes like Atlantic cod (Gadus morhua) and yellowtail flounder (Pleuronectes ferrugineus). Because skates share some food resources with these teleosts, National Marine Fisheries Service biologists are concerned that food resources historically available to demersal fishes are now being consumed by the abundant skates, which could indirectly and negatively impact the growth of demersal species. The main objectives of this study were to estimate the consumptive impact of little skate (Raja erinacea) and winter skate (Raja ocellata) on their invertebrate prey, and to test whether these skates may indirectly impact the growth of yellowtail flounder on Georges Bank. The diet of little skate was comprised mainly of amphipods and decapods on Georges Bank. Both proportions of the two taxa and selection of prey varied depending on season, site, and body length of skates. Seasonal overlap was highest (0.42-0.92) between little skate $<$39 cm and all lengths of yellowtail flounder, indicating little skate could affect yellowtail flounder indirectly through resource competition. The consumptive impact of little skate and winter skate on their benthic prey was determined by estimating the percentage of benthic production consumed by these skate species. Annual consumption ranged from 0.085 kg fish$\sp{-1}$ yr$\sp{-1}$ for little skate 10-19 cm in length to 0.860 kg fish$\sp{-1}$ yr$\sp{-1}$ for skates 50-59 cm, and from 1.186 kg fish$\sp{-1}$ yr$\sp{-1}$ for winter skate 40-49 cm to 5.528 kg fish$\sp{-1}$ yr$\sp{-1}$ for 90-99 cm skates. The percentage of benthic production consumed by little skate and winter skate from 1969 to 1990 ranged from 5% to 15% and 11% to 43%. This indicated only a small to moderate proportion of benthic biomass is consumed by these species. Linear regression and Spearman correlation analyses indicated growth of age 1+ to age 4+ yellowtail flounder was significantly and inversely correlated to its own density, indicating growth of this species is density-dependent. Growth was either not, or positively, correlated to the abundance of skates, groundfishes, and other flounders, suggesting interspecific competition is not an important regulatory mechanism of growth. Therefore, skates appear not to have a measurable impact on the growth dynamics of yellowtail flounder.