Several species of vespertilionid bat gather at underground sites in temperate zones during late summer and early autumn for an activity called swarming. Up to many hundreds of bats may arrive and depart during a night. Bat activity was recorded automatically with a frequency division logging system at a swarming site in south-east England for 415 nights between 1997 and 2001. In all 5 years, activity was highest between the beginning of August and the end of October (the swarming period), with a peak in September. Activity varied markedly from night to night and was affected by rainfall (which significantly suppressed swarming activity), and residual maximum ambient temperature (with which activity was positively correlated). Moon phase had no detectable influence on swarming activity. Activity was low in the first few hours after sunset of each night during the swarming period indicating that few bats emerged and that there was low daytime occupancy of the site. Activity increased to a peak between 6 and 7 h post-sunset consistent with a large number of bats arriving after the first evening foraging bout. Activity then decreased gradually up to dawn as these bats departed again. Logged activity was positively correlated with the number of bats caught by harp-trap, confirming that loggers are a reliable alternative to catching when monitoring a swarming site.

Odonates (dragonflies) are well known for the ability of the males to displace sperm stored in the female's sperm-storage organs during copulation. By this means, copulating males are able to increase their fertilization success. This ability has been used as an example to illustrate a conflict of interests between the sexes in which males have evolved sperm-displacement mechanisms whilst females have presumably evolved means to avoid sperm displacement. The present review has four aims: (1) to describe the copulatory mechanisms used during sperm displacement; (2) to analyse the causes of sperm usage patterns; (3) to discuss this information using current hypotheses on conflict between the sexes; and (4) to illuminate topics for further research. Four copulatory mechanisms are described: sperm removal (physical withdrawal of stored sperm), sperm repositioning (‘pushing’ of rival sperm to sites where its use will be least likely), female sensory stimulation to induce sperm ejection, and sperm flushing (displacement of sperm using the copulating male's sperm). Sperm-precedence studies in Odonata are scarce and their values vary considerably between species. In those species in which sperm displacement is incomplete, the last copulating male obtains a high but variable short-term fertilization success which decreases with time. Some male and female factors affecting sperm precedence patterns are mentioned: (1) male variation in genital morphology; (2) duration of copulation influenced by the male (the longer the copulation, the more stored sperm displaced); (3) adaptations of the sperm-storage organs that allow the female to manipulate the sperm she has received (i.e. avoiding sperm displacement, re-distributing sperm masses, favouring sperm located in certain sites and ejecting sperm after copulation). We suggest that male and female odonates have co-evolved at the level of genital function with the control of stored sperm as the focus of the conflict. The benefits for males in this co-evolution lie in maximizing their fertilization success. However, it is not clear what females obtain from storing sperm and making it unreachable during sperm displacement. Two hypothetical benefits that females may obtain for which some evidence has been gathered are genetic diversity and viability genes. It is finally suggested that odonates can become excellent subjects of study for testing current ideas related to sexual conflict and speciation processes through sexual selection.

The pattern of correlations between geographic–environmental variables and geometric shape descriptors for the skull and mandible of Thrichomys apereoides was studied by recently developed combinations of geometry and multivariate statistical techniques. Environmental variables were obtained for each locality: altitude, mean temperature, rainfall, human population density, and vegetation type. The three views of the skull (dorsal, lateral and ventral) and the mandible were significantly associated with a latitudinal environmental gradient along the diagonal of open areas (caatinga and cerrado) throughout the range sampled. The populations from xeric environments presented relatively larger coronoid processes, larger jugals and wider snouts (related to the activity of jaw closing muscles). A correlation of patterns of shape variation from data sets of different skull views showed that lateral skull shape is the most informative view. Further studies are necessary to separate completely the contribution of genetic and environmental components to skull and mandible shape divergence among populations of T. apereoides.

With five previously described native species, all of them endemic, and at least four introduced species, the murid rodent fauna of Seram is more diverse than that catalogued from any other Moluccan island. Here, the rodent fauna of Seram is briefly reviewed, and a newly recorded rodent from the island is described as Melomys rufescens paveli, a new subspecies of a species otherwise known from New Guinea and the Bismarck Archipelago. Other non-volant mammals from Seram and the nearby islands of Ambon and Buru are briefly discussed.

For snakes, prey size and sex are two possible determinants of head and body size. In garter snakes Thamnophis sirtalis, females are generally longer and have greater body weights than males, and also have larger relative head sizes, which may facilitate foraging success. The selective pressures that account for sexual size dimorphism in garter snakes have not been unequivocally demonstrated. In this study, the body (length and mass) and head (head length, head width, jaw length, inter-ocular distance) sizes of garter snakes inhabiting two nearby but ecologically dissimilar sites with different types of available prey were compared. Overall, the adult female snakes were larger and had greater relative head sizes than males. Males from the two sites did not differ significantly in body or head sizes. However, the mean body length and mass of females from a site where vertebrates are included in the diet were greater than that of females feeding almost exclusively on earthworms. There were also significant site differences in all four head measurements in females, although the direction of the difference varied by site. Diet-induced morphological plasticity is well documented and was evident in this study, although the relative roles of genotype, ontogeny, and competing selective forces in the expression of such plasticity can only be ascertained through future experimental studies.

A new secondary sex character, here named the ‘tibial gland’, found on the forelegs of the males of a new spittlebug species, Augustohahnia dilatitibia sp. nov. (Hemiptera: Cercopidae) from New Guinea, is reported for the first time. The males of Augustohahnia dilatitibia sp. nov. are very distinct and unusual in having remarkably swollen fore tibiae with flat, bare and smooth prolateral surfaces. Scanning electron microscope observations revealed that the prolateral surface of the swollen fore tibiae is covered with numerous, very fine pores, with a density of 0.0068–0.0076 pores/μm2. The distance between the pores varies between 4.04 and 10.77μm. Each pore unit is composed of a deep cavity in which one central sponge-like, grooved disc is set. The opening of the cavity is c. 5.20–5.78 μm in diameter. The central, sponge-like disc in the cavity has a sponge-like apex with 55–65 longitudinal grooves and measures 4.34–4.55 μm in diameter. There are also one to three fine, nearly ellipsoidal pits (c. 0.35–0.48 μm wide and 0.62–0.85 μm long) on the marginal cuticle rim surrounding the central sponge-like disc. These pores are presumed to be male pheromone-producing glands. If so, this represents the first report of a possible male pheromone structure in cercopid spittlebugs. Two new species of the spittlebug genus Augustohahnia Schmidt (Hemiptera: Cercopoidea), A. dilatitibia sp. nov. and A. sanguinifrons sp. nov., both from New Guinea, are described and illustrated. A key to the four known species in Augustohahnia is provided.

Hyphydrus ovatus beetles inhabit lakes, ponds and slow running streams. In central Europe the larvae develop at water temperatures of >10°C from May to July. All three instars live on or close to plants. The first instar prefers the upper areas of plants. Its prey are insect larvae and small stages of planktonic crustaceans. The second instar chooses deeper areas and catches larger prey, whereas the third instar prefers the root zone and feeds mainly on tubificids and chironomids that are caught in the detritus of the bottom layer. The elongated nasale of Hyphydrus ovatus is equipped with different sensilla. Despite the rather different shape of their outer parts, their ultrastructure indicates that they all have a mechanoreceptive function. Chaetiform sensilla are especially numerous and are especially long near the tip. Plate-like mechanosensilla are positioned in a half circle at the ventral edge of the rostral tip. The assumed function of these sensilla is the location of prey. As in other Hydroporinae, the axis connecting the dorsal and ventral points of articulation of the mandibles is sloped laterally. Thereby, the closing mandibles move upwards and press the prey against the underside of the nasale, which – like the upper jaw in vertebrates – serves as the counterpart of the mandibles. On its ventral side, the nasale is equipped with various cuticular teeth that serve to hold the prey when it is pressed against them by the mandibles.

The first population recapture study of indigenous coypus Myocastor coypus is reported in protected areas of Argentine pampas within a region where the species is hunted for food and fur. Quarterly live trapping at two sites captured a high proportion of all coypus present (96% of juveniles; 85% of adults, independently of season, site and sex). Survival (from mortality or emigration) was constant across sites and seasons, and higher for juvenile males than females (88% and 71%, respectively) and lower for adult males than females (45% and 81%). Most turn-over in numbers was replaced by births, and immigrating males tended to be the largest individuals present. Sex ratios were male-biased amongst juveniles and female-biased amongst adults. Taken together, these results are consistent with local resource competition in habitat rendered low in quality by crowding inside the locally protected areas. The indigenous coypu population matured more slowly, and to a smaller maximum body size, than coypus studied in introduced populations in North America and particularly in northern Europe. Cold European winters may have contributed to the evolution of precocious sexual maturity and larger body sizes in European compared with indigenous populations.

The burrowing habits of thalassinidean shrimps presuppose their high susceptibility to the deleterious effects of gill fouling by particulate debris and epibionts. The structures and strategies used to prevent gill fouling were examined in Nihonotrypaea japonica and Upogebia major by morphological inference and aquarium observations. The gill complement of both species consists of 10 arthrobranchs, with flattened filaments of the phylloid trichobranch type. Nihonotrypaea japonica and U. major lack passive gill-cleaning setal systems in the branchial chamber, but were found to actively groom gills by inserting and brushing the gill filaments with a pad of dense setae on the distal segments of their fifth pereiopods (P5 brush). Grooming setae of the P5 brush in N. japonica are of serrate and plumose types, whereas they are serrate and pappose in U. major. Serrate setae with distally curved spiniform denticules were morphologically similar in both species. Phylogenetic mapping of the gill-cleaning mechanisms suggests that gill-cleaning characters are homoplasic and systematically conservative among the families of the Thalassinidea.

Movement pattern of the Iriomote cat Prionailurus bengalensis iriomotensis was studied on Iriomote Island, Japan by radio-tracking. The influence of sex and reproductive status of cats on their daily movement distance (DMD), straight-line distance (SLD), daily movement range (DMR) and other relative indices of movements was estimated. DMD was longest in males (3.2 km) and shortest in females without kittens (1.9 km). Also, males moved faster than either females with or without kittens (0.34, 0.25 and 0.25 km/h, respectively). During the mating season, males moved faster and their DMRs were larger than during the rest of the year. The SLD/DMD ratio in males was similarly high in both seasons (0.29 and 0.33), which indicates that their daily routes were expanded over large areas year-round, possibly to maintain a continuous control over their home range. In contrast, SLD/DMD varied seasonally in breeding females. It was low (0.09) during the lactating period, when females returned repeatedly to the den with kittens and it became high (0.32) out of the breeding season. Breeding females moved equally long DMDs in both seasons, whereas non-breeding ones moved shorter distances during the breeding period. These differences may be related to seasonal changes of food abundance. Cats moved intensively, as indicated by long (812–1139 m) routes per km2 of their home range, and occasionally visited distant parts of their ranges in a short time (4–5 days). On the other hand, they showed a relatively slow increase of the patrolled area, from 10% to 36% of their home ranges, in the first and fifth day, respectively. Movement pattern of Iriomote cats showed no clear relation to active defence of home ranges.

Eggs of Pelodiscus sinensis were incubated under one fluctuating and four constant temperatures, and hatchlings from different incubation temperatures were maintained under identical conditions to assess the effects of incubation temperature on sexual phenotype and hatchling growth. The incubation length decreased as temperature increased, but it did not differ between sexes within each temperature treatment. Hatching success was higher at intermediate temperatures (28 °C, 30 °C and the fluctuating temperature regime) than at low (24 °C) and high (34 °C) temperatures. The sex ratio of hatchlings did not differ from equality within each temperature treatment. Thus, our data support previous work that P. sinensis does not have temperature-dependent sex determination, and add evidence for the prediction that turtles within the Trionychidae have genotypic sex determination exclusively. Incubation temperature affected hatchling mass, with hatchlings from intermediate incubation temperatures being heavier than those from low (24 °C) and high (34 °C) incubation temperatures. Hatching size was not a predictor of post-hatching growth. Incubation temperature affected hatchling growth, with hatchlings from 24 °C overall growing faster than did hatchlings from higher incubation temperatures. The influence of incubation temperature on hatchling growth was well buffered within the range of constant temperatures from 28 °C to 34 °C. Fluctuating incubation temperatures augmented male growth but reduced female growth, as female embryos were more vulnerable to extremely high temperatures.

Jump performance was studied in seven flea species (Xenopsylla conformis mycerini, Xenopsylla ramesis, Xenopsylla dipodilli, Parapulex chephrenis, Synosternus cleopatrae pyramidis, Nosopsyllus iranus theodori and Stenoponia tripectinata medialis). Jump length was compared among individuals, between sexes and among species, and an examination made of whether: (1) there is a consistent difference in jump performance between male and female fleas; (2) sexual dimorphism in body size affects jump performance; (3) there is a size-specific difference in jump performance among species; (4) there are morphological correlates of jumping ability among individuals, between sexes and among species. Males were smaller than females in all species and size dimorphism did not increase with increasing body size. In all species, except S. t. medialis, both absolute and size-specific jumping abilities of the males were less than those of the females. Males and females of S. t. medialis demonstrated similar jumping ability despite their body size difference. The best jumpers were S. c. pyramidis, N. i. theodori and S. t. medialis, whereas X. dipodilli as well as female X. c. mycerini and X. ramesis demonstrated the poorest ability to jump. Female:male performance ratio differed among species, being the highest in X. c. mycerini and X. ramesis, and the lowest in S. t. medialis. The proportion of the overall dimorphism in jumping ability that was owing to size differences between males and females was low, ranging from 0.10 to 0.34. No correlation between jumping ability and pleural height or length of hind coxa, femur and tibia was found in any species, except in X. dipodilli. Interspecific comparisons using size-corrected and phylogeny-controlled values did not show any relationship between jumping ability and any morphological trait. Fleas occupying habitats with a soft sandy substrate were better jumpers than those living in habitats with a hard stony substrate.

To reach adult habitats, juveniles of three species of Hawaiian gobies (fishes under 3 cm long) climb waterfalls up to 350 m high, over 10 000 times their body length. The demands of moving through such an extreme environment could constrain the range of viable locomotor mechanisms that these fishes use. Previous qualitative observations indicated that Lentipes concolor and Awaous guamensis use ‘powerbursts’ of axial undulation to climb, whereas Sicyopterus stimpsoni ‘inches up’ vertical surfaces by alternately attaching oral and pelvic suckers to the substrate. To compare these propulsive mechanisms and their physiological requirements, high-speed video footage of climbing by juveniles from these three species on an artificial waterfall were collected, and climbing kinematics and performance for the two climbing styles were quantified. Bouts of powerburst climbing by L. concolor and A. guamensis typically begin in or near direct water flow and are initiated by a single, rapid adduction of the pectoral fins. Powerburst climbing bouts are rapid (12.4±1.0 body lengths (BL) s−1), but short in duration (0.07±0.02 s) with few continuous locomotor cycles (3.8±1.3 cycles bout−1). Powerburst climbers use high amplitude undulations along the entire body, but minimum resultant velocities of these undulations are high (>6 BL s−1). This suggests that these species may hybridize terrestrial propulsive mechanisms with aquatic mechanisms. In contrast, climbing by inching in S. stimpsoni involves little axial undulation or fin movement. Sicyopterus stimpsoni typically exit the water outside of direct flow and seem to use terrestrial propulsive mechanisms. As the oral disc attaches to the substrate, it expands to almost twice its resting area, after which the posterior body is pulled upwards; once the pelvic disc attaches, the oral disc releases and the anterior body advances. Climbing bouts include several continuous cycles of disc attachment (11.0±1.4 cycles bout−1) and last several seconds at velocities of 0.21±0.01 BL s−1. Before climbing waterfalls during migration to adult habitats, S. stimpsoni undergo a non-feeding metamorphosis that leads to the development of the mouth as a secondary locomotor organ. The unusual behaviour and ontogenetic strategy of S. stimpsoni seem to be evolutionary novelties, rather than ancestral retentions, suggesting that the evolution of these features may have been closely correlated. The substantial performance and kinematic distinctions between powerburst and inching climbing indicate that considerable locomotor diversity can evolve even in the context of extreme environmental demands.

Three locomotor modes were examined (sprinting, jumping and climbing) in eight species of skinks: seven Niveoscincus spp. and Pseudemoia entrecasteauxii. These species formed four distinct ecological types: ground-dwelling, arboreal, heath/rock dwelling, and saxicolous. Significant behavioural preferences for particular escape modes in the field were found, which reflected the performance capabilities at an animal in the laboratory. This study used both non-phylogenetic and evolutionary based analyses to demonstrate that species occupying different microhabitats and using different escape tactics exhibit corresponding differences in performance abilities. Four Niveoscincus species are specialized in performance abilities and behavioural responses (N. greeni, N. ocellatus, N. pretiosus, N. orocryptus) by excelling in some performance abilities but having behavioural restrictions at attemping other locomotor modes. Only N. microlepidotus and N. orocryptus used escape tactics opportunistically; these species possess a suite of behavioural responses that may reflect the wide range of microhabitats they occupy. Ground-dwelling species N. metallicus, N. coventryi and Pseudemoia entrecasteauxii performed poorly at all performance abilities and seemed to have biomechanical limitations, rather than behavioural restrictions, on locomotor mode. Thus, when making interspecific comparisons, the behaviour of an animal needs to be considered before appropriate performance measures are selected. Phylogenetic analyses demonstrate that escape tactics and performance abilities have co-evolved in Niveoscincus, with an evolutionary trend towards behavioural and locomotor specialization.

Patterns of growth of caudal vertebrae in the chameleon Chamaeleo dilepis were determined using principal component analysis, and compared to growth of the entire tail relative to snout–vent length. Despite significant positive allometry of the whole tail, growth rates of vertebrae differed along the length of the tail. Specifically, there was a proximal region that grew positively allometrically, and an extensive distal portion that grew with negative allometry. Intervening, was a short transitional region of approximate isometry. Positive allometry of the entire tail resulted from the extensive proximal region that grew in this manner. Although the region of positive allometry extended further caudad than the m. caudofemoralis longus, m. retractor penis magnus, and m. ischiocaudalis, its extent correlated more closely with the presence of neural spines (which are used as a proxy for the extent of the m. transversospinalis) and with tail coiling in this species. The positively allometric region housed the non-segmental musculature of the tail and did not bend, and the negatively allometric region identified the portion of the tail that was prehensile.

Ecological densities of large herbivores were estimated using the line transect method in the tropical moist forests of Bhadra Tiger Reserve, southern India, during November 2000. The species of interest were chital Axis axis, sambar Cervus unicolor, muntjac Muntiacus muntjak, gaur Bos gaurus and Hanuman langur Presbytes entellus. Six permanent transects, ranging from 2.6 to 3.6 km (totalling 18.2 km) were each walked 26 times. A total of 473 km were walked during the surveys. Numbers of detections were generally low (51, 25, 68, 17 and 302 for chital, sambar, muntjac, gaur and langur, respectively). Mean estimated densities of different species were [${\smash{\raise6pt\hbox{\gary}}{\hspace*{-6pt}{D}}}$ (${\pm}\,$SE)]: 4.51 (${\pm}\,$1.05) chital km−2, 0.89 (${\pm}\,$0.23) sambar km−2, 3.64 (${\pm}\,$0.63) muntjac km−2, 1.48 (${\pm}\,$0.63) gaur km−2 and 22.62 (${\pm}\,$2.64) langur km−2. Possible shortcomings of our estimates are considered and suggestions made for improving future surveys. A comparison with densities estimated for these species in other well-protected parks showed that the chital, gaur and sambar densities in Bhadra were extremely low. The main causal factors for these low densities seem to be poaching and livestock grazing.

Many areas have experienced disproportionate increases in the number of conflicts between large carnivores and humans, and this is especially true in western North America where urban sprawl has encroached into regions that have historically contained large carnivores. Yet, globally there is a paucity of studies of temporal changes in behavioural and ecological parameters of carnivores associated with human-induced perturbations at the same location. We capitalized on the extent to which human population growth and its coincident food stores offer a quasi-experimental setting to test hypotheses about the impact of novel food resources. Using black bears Ursus americanus and garbage, measures of behaviour and ecology were contrasted between individuals living in urban–wildland interface (‘experimental’) and in wildland (‘control’) settings at the interface of the Sierra Nevada Range and the Great Basin Desert in the western United States. A temporal dimension was included by comparing our data to those from the same population lacking areas of human encroachment 10–15 years earlier. Specifically, an examination was made of the impacts of garbage on bear time budgets, patterns of activity, and den chronology. Individuals at urban interface areas relative to wildland conspecifics were: (1) active for significantly fewer h per day (8.5 vs 13.3 h; P<0.01); (2) shifted their activities to nocturnal periods (P<0.001); (3) entered dens significantly later and remained in them for significantly fewer days (P<0.05). Our results are contrasted with selected carnivores from sites where attendant changes in behaviour and ecology have accompanied landscape changes associated with human activity. Our findings suggest alterations in carnivore ecology may be rapid and occur within shorter periods than had been previously assumed.

The relationships between haematozoan parasites, cell-mediated immune response, territory features and colour morph were investigated in tawny owls Strix aluco, in the Park of Monza in Milan, northern Italy. Rufous and grey birds were found to differ significantly in their blood parasite loads, particularly during the breeding season, when rufous birds hosted more parasites than grey birds. The increase in parasitaemia of rufous owl was not, however, owing to the breeding status of sampled individuals. Although body condition was similar between the two colour morphs, immune response was higher in grey than in rufous owls for the same level of parasites. Moreover, parasites seemed to influence breeding of rufous birds negatively, because only individuals with no or few parasites reproduced. Both these results suggested a higher susceptibility to parasites of rufous owls. Nevertheless, a habitat effect existed because parasite loads increased significantly with woodland extent and tree-density within owl territories, and rufous birds defended more wooded territories than grey ones. Thus, the differential parasitaemia between colour morphs in tawny owls may be the result of both a differential exposure to flying vectors depending on habitat selection and a differential colour-based susceptibility of individuals.

Eight aardvarks Orycteropus afer were fitted with radio-transmitters at Tussen die Riviere Nature Reserve between June 1996 and December 1997. Two became habituated and could be approached to a distance of 1 m. Activity patterns, home ranges and burrow use were investigated using telemetry and direct observations. In winter, aardvarks emerged above ground earlier, returned to burrows earlier and were active for shorter periods than in summer. They were nocturnal except occasionally in winter when some individuals foraged above ground before sunset. Home ranges varied between 133 ha and 302 ha and range overlap occurred within and between both sexes. Densities were c. eight animals per 1000 ha. Aardvarks used steep slopes (>15°) less than would have been expected according to availability. Average lengths of consecutive burrow use for three aardvarks were 4.9, 6.9 and 8.6 days, with some burrows being abandoned after only 1 day and one burrow being used for 38 days.

Procellariform seabirds provide a good model for studies of the causes and consequences of variability in avian egg size, because females can only adjust reproductive investment by breeding intermittently, or by altering the size of their single egg. Maternal characteristics such as age, breeding experience and body size, as well as environmental variability, can influence egg size, but the effect of these factors has rarely been assessed simultaneously in the same study. Previous studies in Scotland have shown that the egg size of northern fulmars Fulmarus glacialis increased in relation to breeding experience. At this colony the influence of breeding experience, body size and inter-annual variability upon egg size was tested simultaneously. Data collected over seven breeding seasons between 1975 and 2002 showed that egg size varied significantly both between years and in relation to the length of the breeding experience of females, but that female body size explained most variation in egg size. Inter-annual variability in egg size was not related to the winter North Atlantic Oscillation, which had recently been shown to influence other measures of reproductive success at this colony. Larger eggs also seem to be more likely to produce successful fledglings. These findings are discussed in relation to the relative contribution of egg quality and parental quality on increased reproductive success.