In 2003, the date for publication of the annual report for 2002 was postponed and it is therefore printed in this issue of the Journal. 2002 was, however, another very successful year during which we received 281 submissions and published 167 papers in volumes 256–258. The Journal continues to receive high-quality, original research papers and reviews from a wide range of taxa and a variety of disciplines. Our aim to attract a truly international set of contributors has been met once again, with papers from 41 countries being submitted during 2002.

A 40-species assemblage of jumping spiders inhabiting a small (0.6 ha) tropical suburban study site in Zimbabwe was studied over 4 years. It was hypothesized that marked habitat separation is one of the critical factors permitting the close coexistence of so many species from the same family (Salticidae). The 25 most common salticids (i.e. species with 10 or more records) were found to occupy six primary habitat types: tree trunks (with three salticid species), tree leaves (two species), shrubs (five species), walls (three species), low herbaceous plants and grasses (three species) and ground and litter (six species). The remaining three species, termed habitat generalists, frequented three or more of these habitats. Finer-scale microhabitat separation was also recorded in many species. The remaining 15 less common species (i.e. <10 records) could not be assigned to particular habitats with certainty. Twenty-one (84%) of the 25 common salticids showed a significant association with a single primary habitat type. Three-quarters of species were present for 3 or 4 years of this 4-year study. It was concluded that the narrow spatial niches occupied by most species, combined with the numerous microhabitats within the site, were conducive to the highly diverse salticid assemblage. Attention is drawn to the problems of long-term sampling of such a small area without undue disturbance to the fauna. The impacts of long-term weather cycles on habitats, ephemeral niches, age-dependent use of habitats, inter- and intraspecific associations of individuals within habitats, and the relationship between suburban and wild habitats are discussed. Suburban environments, by providing a wide range of microhabitats, can support a high diversity of salticid species and thus contribute to their conservation.

Frequency-modulated sweep calls of Otomops martiensseni were recorded from individuals as they emerged from nine different building roosts near Durban, South Africa. Multiple analyses of call features, including duration (ms), lowest frequency (kHz), highest frequency (kHz) and frequency with most energy (kHz), indicated significant inter-individual variation. Discriminant function analysis of call features correctly classified the calls of individuals from four roosts > 70% on 19 of 28 times. Although other species of molossids (Chaerephon pumilus, Tadarida aegyptiaca, and one unidentified species) produced social calls as well as frequency-modulated sweep calls, O. martiensseni produced just the latter vocalizations and they were longer and lower in frequency than those of the sympatric molossids. Other species of molossids, but not O. martiensseni, produced feeding buzzes as they attacked flying insects. The frequency-modulated sweep calls of O. martiensseni seem to serve a communication function, but they may not be used in echolocation unlike similar calls by other molossids. Individually distinct communication signals (frequency-modulated sweep calls) enhance communication in a species that lives in year-round social groups (one adult male, females and dependent young).

Nearly all animals show altered musculo-skeletal phenotypes when subjected to captive conditions. Whether such changes affect biomechanical performance is for the most part unknown. In American alligators Alligator mississippiensis such modifications include shortened jaws, more robust body form, and broadened heads. Bite-force performance was assessed for a variety of sizes of wild-captured alligator specimens and the results correlated with morphological indices. Bite forces ranged from 217 to 13 172 N, with the latter being the highest value ever measured for a living animal. These data were statistically compared with those for long-term captive specimens using ANCOVA. Bite-force performance showed similar patterns of increase between captive and wild-reared animals, and bite forces with respect to snout–vent length and body mass were statistically indistinguishable. Nevertheless, with respect to head size, captive alligators were found to bite more forcefully than their wild counterparts. These findings illustrate the importance of considering biomechanical performance differences between wild and captive individuals if meaningful ecological ties are to be made. Furthermore, before concluding that wild-reared or captive animals show similar or different biomechanical performances, it is important to understand that standardization to different morphological parameters can reveal conflicting results. Consideration as to which measures are the most germane to the question at hand is essential.

Earlier authors demonstrated that wetland species of the genus Brachinus (Coleoptera: Carabidae) are pupal ectoparasitoids of water beetles of the families Dytiscidae, Gyrinidae and Hydrophilidae. Here we demonstrate that the dryland species, Brachinus explodens and B. crepitans, develop on pupae of ground beetles of the genus Amara (Carabidae). As Brachinus and Amara species often occur together at high densities in crops, the rearing of both Brachinus species was followed in the laboratory. Females laid eggs in the soil and the first instar larva searched for a host. After finding an Amara pupa, a Brachinus larva bit it and then fed on haemolymph exuding from the wound. The second instar larva continued feeding in this way. The third instar larva attached the Amara pupa to its dorsum and ate it completely within c. 2 days. Feeding was followed by a resting phase and moulting to pupa and adult. The total post-embryonic development was completed within 20 days in B. explodens and 24 days in B. crepitans at 24.7 °C.

Earwigs (Insecta, Dermaptera) show astonishing diversity in penis morphology. In several families, males have a single penis (termed virga), whereas males of other families possess two functional virgae. Taxonomists have assumed that the two-virgae state is ancestral; however, ecological reasons why the ancestor acquired two virgae have not been explored. This study investigated in detail male and female genital structures, mating behaviour and insemination processes in the earwig Diplatys flavicollis (Diplatyidae). Diplatyidae are considered to be the most primitive family of earwigs. SEM and light-microscopy revealed that males of this species have two gonopores on each of two virgae, similar to those reported in other diplatyids (i.e. two double-barrelled penises), while females have four to six independent sperm-storage organs (spermathecae). Rapid fixation of mating pairs and insemination success of males from which one virga had been removed clearly revealed that only one virga was used for mating and was usually sufficient for inseminating multiple spermathecae. This finding rejects the one-to-one correspondence between male gonopores and female spermathecae. Based on allometric analysis of spermathecal variation, the possible significance of multiple spermathecae in relation to sperm-storage strategies of females is discussed. Compared to studies of male genital morphology, few studies have described spermathecal morphology. Based on compiled data of spermathecal and virgal morphology among earwigs, a parallel evolutionary trend between spermathecae and virgae from complicated (multiple) to simple (single) ones is suggested, and further investigation of the phylogeny, female genital morphology and insemination processes among earwigs is encouraged.

Partial (DNA) sequences were examined for one nuclear (28S rRNA gene) and one mitochondrial (16S rRNA) locus for nine species of pomatiopsid snail (Gastropoda: Rissooidea: Pomatiopsidae) from south-east Asia and south-west China. Fresh field samples were collected for the following taxa: Delavaya dianchiensis (Triculinae: Triculini) from Dianchi lake, Yunnan Province, China; Neotricula aperta (Triculinae: Pachydrobiini) from north-east Thailand; Neotricula burchi from northern Thailand; Oncomelania hupensis robertsoni (Pomatiopsinae: Pomatiopsini) from south-west China; Tricula bambooensis (Triculinae: Triculini) from Dianchi lake; Tricula bollingi from northern Thailand; Tricula hortensis from Sichuan Province, China; Tricula ludongbini from Dianchi lake; and Tricula xiaolongmenensis from Dianchi lake. This work represents the first published DNA-sequence data for the Dianchi lake taxa and the first 28S sequence data for all nine taxa. All of these taxa with the exception of N. burchi and the Dianchi taxa transmit Schistosoma in nature; N. aperta and O. h. robertsoni transmit Schistosoma to humans. The data were used to estimate phylogenies using the maximum likelihood method, a Bayesian method, and the maximum parsimony method. The paper aims to examine the relationships between N.aperta, N. burchi and T. bollingi and the relationship between the Chinese taxa and the south-east Asian taxa; these relationships being important in evaluating certain historical biogeographical hypotheses. Good congruence was found between the phylogenies estimated by the three methods for both the 16S and 28S loci. However, poor congruence was found between the phylogenies based on the 16S and 28S data when the maximum likelihood and Bayesian methods were used. The lack of congruence is explained as a consequence of a rapid, recent, endemic radiation of the Yunnanese Tricula spp.; this hypothesis is consistent with current historical biogeographical models for the Triculinae. The paper puts forward dispersal hypotheses, based on palaeogeographical changes, as possible explanations for the phylogenetic relationships estimated here and for the current biogeography of these taxa.

Tibetan argali Ovis ammon hodgsoni and blue sheep Pseudois nayaur have almost completely overlapping distributions encompassing most of the Tibetan plateau and its margins. Such a sympatric distribution of related species with similar ecological requirements implies that there is some degree of resource partitioning. This may be accomplished on the basis of habitat and/or diet separation. This study evaluated such ecological separation on the basis of physical habitat partitioning by these two sympatric ungulates in Hemis High Altitude National Park, Ladakh, India, in an area where the argali established a small new population in 1978. Such separation was tested for on the basis of expected difference between the species in their proximity to cliffs, associated with species-specific anti-predator behaviour. Tibetan argali selected habitats away from cliffs while blue sheep selected habitats close to cliffs. Blue sheep also selected steep slopes whereas argali selected gentle slopes. The two species did not differ in their use of habitats in terms of elevation. They did, however, differ in their use of plant communities; blue sheep selected sub-shrub and grass-dominated communities whilst argali selected forb-dominated communities. We suggest that the two species coexist in this site as a result of the differential use of habitat associated with their species-specific anti-predator strategies.

American mink Mustela vison are an invasive species in many parts of the world (e.g. Europe, Iceland and South America) and in many of these areas they are a threat to the native wildlife. It is therefore important to establish a reliable and efficient method to estimate their relative abundance in order to manage their populations. In this study, the use of surveys of signs (scats and footprints) as a way to estimate the relative abundance of American mink was evaluated. Populations of mink in three areas of England were studied to evaluate the impact of three factors (surveyor, marking places and month) on the probability of deposition and detection of signs. Among these factors only month had a significant effect. In the second part of the study, sign surveys were compared with live trapping and it was found that the proportion of sections with mink signs was only loosely correlated to the abundance of mink as estimated from live trapping. In conclusion, this study suggests that, provided that surveyors are trained, that there is a sufficient number of marking places in each section, and that the surveys are carried out in comparable seasons, sign surveys are an appropriate way to monitor the relative abundance of mink. Sign surveys are better than trapping when it is necessary to monitor mink populations over large areas because they are less time consuming, require less manpower and are cheaper.

The claw geometry of birds can be used to predict their mode of life. Previous studies, however, have not considered how bird size might affect these predictions. Thus, in the present study, the geometric scaling relationships of bird claws are examined for a variety of extant birds with different modes of life: predatory, climbing, perching or ground-dwelling. Measurements of hind-limb claw radius (i.e. claw size) and claw angle (i.e. claw ‘hookedness’) of the third digit claw were made on 120 species of bird ranging from 0.0057 kg to 44 kg in body mass. Claw radius was found to be proportional to (body mass)0.34 across all species. Claw angle was found to increase with body mass for predatory and climbing birds (i.e. bigger birds have relatively more hooked claws), and decrease with body mass for ground-dwelling birds (i.e. bigger birds have relatively less hooked or flatter claws). No significant relationship was found between claw angle and body mass for perching birds. Mode of life could not be predicted with any certainty using measurements of either claw radius or claw angle, suggesting difficulty in assigning fossil species such as Archaeopteryx to a specific locomotor category. As claw design should enable the claw to withstand the forces placed upon it, further work is needed to establish the stresses experienced by the claws of different types of bird.

In ungulates, a set of deciduous teeth is replaced by a set of permanent teeth during their first 2 years of life. Little is known about factors causing variation in timing of tooth eruption, but overall size, condition or mineral deficiencies may play a role. The pattern of incisor (I) and canine (C) eruption was investigated in 2241 yearlings from five populations of red deer Cervus elaphus in Norway during 1964–2001. Before the sampling period (10 September–15 November), I1 had erupted in most yearlings (97%), while eruption of I3 and C had barely started (erupted in 7% and 5% of individuals, respectively). Exactly 50% of the yearlings had erupted I2. Jawbone length was shown to be a main determining factor for eruption of I2. However, for a given jawbone length, there was large variation among populations in the proportion of yearlings with erupted I2 (from <20% to >75%). Females got their permanent I2 earlier than males. Tooth eruption was found to be later with increasing local deer density, even after correcting for body weight. There was a residual effect of body weight (a measure of condition) after body size (measured as jawbone length) was controlled for in one of five populations, indicating that overall condition may influence timing of tooth eruption in this area. There was no evidence that calcium was limiting for eruption, as no correlation was found between timing of eruption and limestone in bedrock. Since there is large variation in timing of tooth eruption among red deer populations even after controlling for differences in phenotype and environmental factors, genotypic variation may also play an important role for tooth eruption patterns.

To determine the effects of their movement patterns, seven Cape clawless otters Aonyx capensis were caught in two rivers and radio-tracked between 1993 and 1995. Total range length varied from 4.9 to 54.1 km and core length from 0.2 to 9.8 km. Total area of water used varied between 4.9 and 1062.5 ha, and core areas from 1.1 to 138.9 ha. As predicted using the resource dispersion hypothesis, total home-range length was correlated with mean reed bed (high food density patch) nearest neighbour distance. The pattern of home-range use by females was suggestive of territoriality. Male Cape clawless otters had overlapping home ranges, both with other males and with females.

Four species of canids occur in the Cerrado of central Brazil. Three of them, the maned wolf Chrysocyon brachyurus, the crab-eating fox Dusicyon thous and the hoary fox Dusicyon vetulus, were studied in Emas National Park between 1996 and 1999 to investigate niche separation. The diet of the three species was studied to understand niche breadth and degree of overlap. Habitat and activity patterns were used as second and third ecological parameters to define niche dimensions, and were estimated using camera-trap data. The maned wolf is the largest species, weighing c. 21 kg, and is about three times larger than the crab-eating fox and six times larger than the hoary fox. The major ecological differences between the three species were found in their food niche and habitat use, where crab-eating fox presented higher differences from the hoary fox (Pianka's index of niche overlap (O) = 0.405). Despite differences in niche breadth, habitat use between the hoary fox and the maned wolf were more similar, explaining their larger overlap, in comparison with habitat use by the crab-eating fox. Activity patterns among the species showed less divergence. The three species presented two activity peaks, one in the dusk–night period and another in the morning period. These data permit a better understanding of the ecological separation of the three Cerrado canids that enables their coexistence.