Because lactation has high energetic costs, females should vary their foraging behaviour according to reproductive status. In ungulates, however, some studies found no differences in feeding behaviour between non-reproductive (yeld) and lactating females. Despite the importance of rumination in determining digestive efficiency, no study has attempted to identify tactics involving this parameter in free-ranging ungulates. Whether or not females varied their ruminating behaviour as a function of the presence/absence of offspring was tested by observing marked bighorn ewes Ovis canadensis of known reproductive status, age, and body weight. Lactating ewes ruminated 1.21 times faster than yeld ewes and showed less inter-individual variability in rumination speed, suggesting an energetic constraint. After considering the potential physiological advantages of this behaviour, I suggest that differences in ruminating parameters may allow the synchronization of activities in groups made up of individuals with different energy requirements. Lactating females may increase rumination effort in response to increased energetic demands and risk of predation.

Socioecology has traditionally assumed that differences in social form between species derive from selected responses to particular evolutionary environments, while differences within a species reflect plastic responses to contemporary differences in ecology. The issue of behavioural or social plasticity is particularly salient for primatologists as it ties into conceptualizations of the evolution of cognitive capacity. In this regard, savanna baboons Papio hamadryas have provided a focus for research by virtue of their ecological plasticity and a problem for researchers by virtue of marked social differences in one subspecies – hamadryas baboons P. h. hamadryas – that cannot be explained solely by reference to current ecology. It is argued that the problem extends beyond hamadryas and that at least some differences between other populations also reflect selected responses to restricted, local evolutionary conditions. Of particular importance to the understanding of contemporary differences in baboon sociality is the extent to which evolutionary environments have constrained group size and perturbed adult sex ratios and, thereby, structured responses to the ever-present threat of infanticide. The view that there is much to be gained from a phylogenetic approach to population differences in behaviour is endorsed.

The Editors are pleased to report another successful year for the Journal of Zoology in 2003. During the year we received 257 submissions and published 140 papers in volumes 259–260; and our rejection rate for the period was 45%. Our aim to attract a truly international set of contributors has been met once again, with papers from 41 countries being submitted during 2003.

A new genus and species of gecko from the Lower Eocene of north-western Russia is described from a superbly preserved specimen in Baltic amber. It is the oldest gekkonid lizard to be represented by more than fragmentary skeletal remains. The digits of the specimen are mostly intact and reveal a unique combination of characters not seen in any living form. Expanded sub-digital scansors on the toes, however, are essentially similar to those of modern climbing geckos and verify the existence of a complex adhesive system 20–30 million years earlier than supported by previously discovered fossil geckos.

Population density is likely to determine the form of competition in which males are engaged for access to females. At low density, scramble competition should be of paramount importance because of the low probability of encounter between males and females. Consequently, sexual selection should favour characters that allow rapid detection of females. Conversely, at high population density, sexual selection should favour attributes that raise the fighting ability of males because of the more frequent contacts between males. These general predictions were tested in this study for the precopulatory mate-guarding isopod, Asellus aquaticus. In this species, male-biased sexual dimorphisms are reported for body size and antennae length and are, respectively, ascribed to contest and scramble competition over females. Therefore, the relative strength of sexual selection on male body size and antennae length at two different densities was experimentally assessed. Multivariate analyses indicated that density affected morphological correlates of the mating success of males, with body size being the main determinant of pairing success in males at a high density, whereas only antennae size significantly affected access to females at a low density. The antennae length of males was manipulated to examine how antennae length affects the ability of males to detect females in three experimental conditions varying in the probability of random contacts with receptive females. An advantage of having long antennae was only observed when there was an intermediate difficulty in finding females. Our results are discussed in relation to the influence of density on selective regimes in A. aquaticus.

Thirty-five museum specimens collected in 1926 and 1947–48 vouch for the distinctiveness of an undescribed large mammal, a form of lechwe antelope. Their preservation has allowed comparative analyses of morphological characters to reveal this new species, the Upemba lechwe Kobus anselli sp. nov. It is most similar to the black lechwe K. smithemani and quite distinct from all other known taxa, including topotypical red lechwes K. leche. Prevailing threats and conservation concerns underwrite its formal description after neglect by science for decades. This lechwe is restricted to the Upemba wetlands, Kamalondo depression, south-east Congo basin (Katanga Province, Democratic Republic of Congo), and has declined greatly since the 1970s. Commercial poaching through the 1980s reduced c. 20000 individuals to under 1000 estimated today. Recognition of K. anselli as a distinct evolutionary entity was previously ignored, because the entrenched taxonomy assumed it was just another red lechwe. Its speciation seems to be recent, probably Pleistocene. Lechwes evolved in an archipelago of wetlands formed through vicariance of a more extensive drainage system, the Palaeo-Chambeshi, which extended from Katanga and north-east Zambia across the south-central African plateau, into the Kalahari basin. The Palaeo-Chambeshi was a major endoreic tributary of Palaeo-Lake Makgadikgadi. Its fragmentation isolated lechwes in the Kamalondo depression from wetlands in the Upper Zambezi–Okavango, Upper Kafue, and the Chambeshi and Luapula drainages. The belated discovery of K. anselli emphasizes the region's conservation significance. Because of its high species richness and high endemism, Katanga is a biodiversity hotspot within the encompassing Katanga–Chambeshi region, also covering eastern Angola and much of Zambia. A secure future for the critically endangered Upemba lechwe hinges on reducing adverse human impact and maintaining the integrity of its wetland habitat. Support for protected areas is critical. Maintenance of ecological processes, focused on aquatic systems, is especially important to conserve biodiversity. Recognition of the complex evolutionary history of the region (since the Late Neogene) underpins the scientific foundation for all conservation plans and activities. It prescribes why a regional conservation strategy should encompass the landscape mosaic, structured across neighbouring drainage systems (Lufira, Upper Lualaba and Chambeshi–Luapula rivers). A trans-frontier conservation area will consolidate the protected areas and land use systems of the region in an ecological context.

Some anurans have a peculiar casqued head with the skin co-ossified with the underlying bones. This type of skull usually is associated with phragmosis, a protective behaviour in which the animal enters a hole and closes it with the head. Although co-ossification of the head in lissamphibians frequently has been associated with water economy, recent studies of Corythomantis greeningi, a casque-headed tree frog from semi-arid areas in north-eastern Brazil, suggest that cranial co-ossification contributes little to conservation of water in the frog. Instead, during phragmotic behaviour, the co-ossified head protects the animal against predators and indirectly enhances water balance. Thus, the primary role of co-ossification is defence, a hypothesis that is the focus of this study, which describes the morphology of the head of C. greeningi with an emphasis on the co-ossification and the venom glands. We report on behavioural features and on the toxicity of the cutaneous secretion produced by the abundant venom glands that are associated with large spicules on the skull.

Mongolian gazelles Procapra gutturosa were observed, hand-captured and radio-monitored in the eastern steppe of Mongolia during 1998–2003 to understand better their reproduction and survival. During early June, 92% (range = 87–96%) of adult (≥2 years old) females observed (mean n = 735/year) were pregnant, and nearly all gave birth during a 10-day interval (24 June–3 July) each year. Mean mass of neonate (1–2 days old) calves (mean n = 47/year) was greater for males than for females, and masses were directly correlated with mean monthly temperatures the previous winter (December–April). Survival during the first 10 days of life for 111 radio-marked neonate calves monitored during 2000–03 (n = 9–55/year) was 0.83 and did not seem to vary with birth weight. Also, deaths owing to hypothermia, abandonment, or unknown causes (interval cause-specific mortality rate = 0.16) outnumbered deaths owing to predation (0.02) during this interval. Survival rates were similar during the rest of the year (0.86 for 355 days) when most mortalities were owing to predation (interval mortality rate = 0.12 vs 0.01). Annual survival of calves was 0.71 (95% CI = 0.61–0.82). Gazelle births are highly synchronous, probably to take most advantage of the short summer growing season, and perhaps to avoid deleterious spring weather and to minimize predation. High fecundity and relatively high calf survival, especially during the first weeks of life, support the notion that gazelle populations can recover fairly quickly from demographic catastrophes.

Direct observation of Eurasian otters was used to demonstrate the characteristics and interannual use of small patches of habitats by females and litters in stretches of two rivers in north-east Spain. During censuses and other vigils, 215 otter sightings (involving 265 otters) were obtained, 14–18% of which were observations on females and litters. Thirty-five additional sightings were obtained in stretches used by litters of large cubs (47 individuals). Six small cub (2- to 6-month-old) and 15 large cub (>7-month-old) different family units (litters) were detected in both rivers as a whole. Average sizes of litters of small and large cubs were significantly similar, being respectively 1.17 (SD=0.41, range 1–2, n=6) and 1.27 (SD=0.46, range 1–2, n=15), so mortality of cubs was low at the period before dispersal. Litters were detected in significantly different patches from other otters, and litters of small and large cubs were also found in significantly different river sections. Litters of large cubs used small home ranges and core areas, and selected sections at reservoir tails, reservoirs, wider stretches, rough waters and refuges. Small cubs were found in narrow river stretches with calm waters. Selected patches had sheltered zones with abundant fish. This reduced use of space may result in the low rate of mortality.

The small mammal fauna of a forest area in southern Nigeria is studied in order to understand the general patterns regulating the community composition. A set of different methods of trapping was used in order to evaluate the composition of the species at the level of both the ground-dwelling and the arboreal guilds. For the Insectivora, six shrew species were trapped, but only two (Crocidura nigeriae and C. poensis) appeared abundant and widespread in the various habitat types. No arboreal species were found. For the rodents, 17 species were captured, which were arranged by a Principal Component Analysis into four groups: (1) Rattus rattus and Anomalurus derbianus; (2) Xerus erythropus and Grammomys rutilans; (3) Graphiurus cf. murinus and Hylomyscus stella; (4) the remaining, more generalist species. For the ground-dwelling rodents, 14 species were found, which were arranged in three groups: (1) Malacomys edwardsi and Grammomys sp.; (2) Mylomys dybowskii and Rattus norvegicus; (3) all the other species, with stronger similarities between Praomys tullbergi, Mus musculoides, Aethomys sp. (unknown), and Lemniscomys striatus. Although an approximately equal number of species constituted the arboreal and ground-dwelling rodent fauna, the number of trapped animals was much higher on the ground than at the level of the arboreal guild (ratio 8.7: 1; total n=2322). The general implications of the patterns observed are discussed.

Patterns of formation and sequence of ossification of the entire skeleton are described for the snapping turtle Chelydra serpentina, based on new and previously published data. Differences in the sequences of ossification events are described and demonstrate considerable intraspecific variation in these developmental data. The chondrocranium of a late developmental-stage specimen is described and illustrated as baseline data for comparative studies. Patterns of formation and chondrification of forelimb and hind limb elements are generally consistent with those of previous studies; however, conspicuous differences in the patterns of ossification are observed among metapodial and phalangeal elements. In the cranium, patterns of ossification of dermal elements are generally more variable than are those of endochondral elements of the braincase, and as in previous studies, endochondral elements ossify after dermal elements. Documented intraspecific variation in sequence and timing of formation, chondrification, and ossification of skeletal elements should serve as a caveat for those conducting studies of phylogeny, heterochrony, and evolution with these data. Causes of natural and induced variation are discussed briefly.

Ecomorphs of Anolis lizards evolved convergently via niche partitioning in separate radiations. Each ecomorph has a suite of traits including body size, morphology, microhabitat, anti-predatory behaviour, and foraging, but foraging movements of ecomorphs have not been characterized quantitatively. Data are presented on aspects of foraging for seven species representing four ecomorphs in the cristatellus series from Puerto Rico. Using these and published data for Jamaican and Hispaniolan anoles, an examination is made of predictions based on definitions of ecomorphs by Williams (1972, 1983). Grass-bush and trunk-crown anoles exhibited higher movement rates than trunk-ground anoles. Data for other radiations agree and support a prediction of relatively high movement rates for twig and trunk ecomorphs. All the species studied were ambush foragers and all used the head-down survey posture, indicating phylogenetic stability of these traits. Arboreal species adopted survey posture in about 75% of observations, and grass-bush species used it when on elevated perches. Rates of foraging movements increased as perch height decreased in grass-bush ecomorphs owing to visual obstruction by vegetation limiting the scanning area close to the ground. A similar relationship occurred unexpectedly in a trunk-crown II species, A. stratulus, whereas movement increased marginally with perch height in a trunk-crown I species, A. evermanni. Perch height and movement rates were unrelated in three trunk-ground species. Movements per minute and per cent time moving were highly correlated among species owing to the brevity of individual movements. Data from other Greater Antillean radiations are needed to separate the effects of ecology and phylogeny on anole foraging.

Data from a long-term study (1993–2003) using artificial nest-boxes, were analysed to examine competition for nesting between blue tit Parus caeruleus and common dormouse Muscardinus avellanarius in Sicily. Occupation rates and the reproductive biology of the blue tit in sample woodlots outside the distribution area of the common dormouse were used as a control in sample areas where the two species were syntopic. A selection test showed that the two species, when living in syntopy, actively chose the small nest-boxes, thus overlapping in the use of the same spatial resource. The experimental exclusion of the common dormouse from nest-boxes caused an increase of blue tit occupation rate. Once the effect of nest-box density was removed, complementary density changes between the two species clearly appeared. The population cycle of both species in the areas of syntopy showed a swinging pattern, i.e. in areas or during years of common dormouse high abundance few blue tits bred and vice versa. The abundance of the blue tit was affected not only by common dormouse abundance, but also by some habitat and climate variables. Blue tits had low breeding success and a high percentage of nest failure (38% in syntopy vs 3% in control areas), owing to nest destruction and predation by common dormouse, as well as probably also by other rodents (23%) and Martes martes (15%). Another experiment, using stuffed models, showed that the blue tit recognized the common dormouse as an enemy, and behaved as if confronted by a competitor or predator species to protect their nests.

Most regions of the world support moth species that use caves as diurnal roosts and/or as overwintering and mating sites. Although subterranean roosts are often also occupied by insectivorous bats, nothing is known about moth–bat interaction at these sites. The frequency of cohabitation and the dynamics of moth–bat interaction at roosts were examined during surveys at 30 known bat roosts in eastern Australia; 15 disused mines and 15 natural caves. Moths and bats cohabited in 20 roosts (67%); however, only two species of eared moths, Speiredonia spectans and S. mutabilis (Noctuidae), were observed. These large (wingspan up to 75 mm) species show a classic escape response to ultrasound. Moths used roosts occupied by bats with both frequency modulated (FM) (four species) and constant frequency (CF) (two species) echolocation calls. Signal frequencies of these bats had ranges of 23–25 and 44–71 kHz. Moths were absent from roosts occupied by high-frequency (>140 kHz) CF bats. Predator–prey interaction was assessed by studying a sub-sample of six roosts occupied by S. spectans and the horseshoe bat Rhinolophus megaphyllus. Speiredonia spectans and R. megaphyllus cohabited roosts throughout the year. Bats captured and consumed moths in these roosts. The numbers of moth wings on the floor of roosts were counted to assess rates of predation. Bats consumed 209 moths at the six roosts over a 12-month period, representing 19% of the moth population at the start of the study. The study establishes that moth–bat cohabitation of roosts is relatively common in eastern Australia and that bats do capture moths in these sites. Bat predation seems to be a significant source of mortality of moths that occupy subterranean roosts.

The honey badger, or ratel, Mellivora capensis has not been well studied and data on life-history variables have previously been incomplete and contradictory. Here we present data on life-history variables that were collected from visual observations as part of the first intensive study of free-living honey badgers (25 radio-marked individuals) and we make comparisons with other similarly sized mustelids (5–25 kg). The honey badger is a non-social species and only the mother rears the young. The data do not support delayed implantation as gestation was a maximum of 50–70 days with the unusually small litter size of one cub. The long time to independence of both male and female cubs (12–16 months) resulted in birth intervals longer than 12 months with no distinct breeding season. Overall the honey badger appears to have a more K-selected life history compared to other species within the Mustelidae.

A new species of Congosorex is described. The new species occurs in Tanzania and is larger in external characteristics than the other two known species in the genus, with the longer tail being especially distinctive. The skull of the new species is wider, on average, than other Congosorex, but intermediate in greatest length. Previously undocumented characteristics that differentiate Congosorex from Myosorex are described. The presence of Congosorex in the Eastern Arc Mountains supports the theory that these mountains are refuges of ancient lineages of organisms with ties to the Guineo-Congolian region of Africa and that there was once a continuous swath of forest connecting these two areas.

Spacing patterns of individuals within an island population of the Central American spiny rat Proechimys semispinosus were studied by radio-tracking eight adults for 2 months during the rainy season and eight adults for 1 month during the dry season in central Panama. All trees ≥10 cm diameter at breast height on the island were mapped, tagged, identified to species, and inventoried monthly to determine fruit production. Diurnal telemetry was conducted to locate daytime resting burrows. Burrow habitat was quantified by measuring nine variables at each burrow site. Individuals had larger home ranges and greater home-range overlap during the rainy season, and males had marginally larger home ranges than females. No evidence was found of attraction, avoidance or exclusive use of space by adult spiny rats. Thirty-six located burrows had nine instances of co-occupancy. Burrows were located close to water and had little herbaceous cover. The presence of broadly overlapping home ranges, larger home ranges of males vs females, greater numbers of burrows of females in home ranges of males vs those of females, and larger home ranges of sexually active males suggest that P. semispinosus had a promiscuous mating system within this population.

The overlap of large carnivores, livestock and people can engender conflicts that often threaten the future viability of carnivore populations in the pastoral systems of Africa. A playback survey of lions Panthera leo, spotted hyenas Crocuta crocuta and black-backed jackals Canis mesomelas and a transect count of wild herbivores was conducted in the Maasai Mara National Reserve and adjoining pastoral ranches to assess the effect of pastoralism and protection on the density and distribution of carnivores in June 2003. Reliability of the prey counts depended on an assumption of similar observability between the protected and pastoral areas, which we assessed using distance sampling in November 2003 and computer simulations and determined to be similar. Estimates of wild prey biomass density was 2.6 times higher in the ranches (14212 kg/km2) than in the reserve (5472 kg/km2) during this wet season count. Apparent hyena density estimates were 1.3 times higher in the ranches (0.561 hyenas/km2) than in the reserve (0.404 hyenas/km2), in correspondence with the regional pattern of prey density. This distribution of hyenas is biased towards the reserve, if it is dependent on prey density. Estimates of apparent jackal density were similar in both areas whereas lion density was anomalously 8.0 times lower in the ranches (0.046 lions/km2) than in the reserve (0.369 lions/km2). Lion and hyena densities and prey biomass did not differ between June 1991 (5172.273 kg/km2) and June 2003 (5472 kg/km2) in the reserve, but jackal density increased in the same period. Lions never responded to playbacks in the ranches, so the potential shift in lion behavioural response for different land use zones is another potential explanation for the patterns found here. We think a real shift in lion populations is a better explanation than a behavioural change in relation to playbacks based on additional data from independent systematic and intensive censuses and playback surveys conducted in the ranches. Lion populations in the pastoral ranches seemed headed for extinction, probably owing to conflicts with pastoralism, necessitating urgent conservation interventions that integrate pastoral economic welfare with large carnivore conservation goals to foster long-term viability of lion populations in the pastoral systems.

The species–area relationship was used to test the effects of habitat reduction on the proportion of tetrapod species that would be considered threatened with extinction, assuming the current estimate of 19% of remaining forest coverage in Rio de Janeiro State, Brazil. Here, species–area estimates from deforestation were compared with a Red List produced by experts. Analyses were performed with: (1) the total pool of species that occur in Rio de Janeiro State (159 amphibians, 115 reptiles, 635 birds and 157 mammals); (2) the pool of species endemic to the Atlantic forest (109 amphibians, 46 reptiles, 149 birds and 35 mammals). The sensitivity of the analyses to variations in the Z parameter of the species–area relationship was evaluated. In general, the species–area overestimated the total number of species threatened with extinction for all tetrapods, while the estimate of endemic mammals predicted to be extinct was close to the number of endemic species listed as threatened with extinction. Furthermore, changes in the Z-values did not affect those results. The species–area relationship accurately predicted the number of endemic birds assumed to be extinct since it used values of Z smaller than 0.25. The analyses with endemic species of reptiles and amphibians of Atlantic forest overestimated the number of species predicted to be extinct for both groups. In conclusion, at the considered scale, mammals and birds endemic to the Atlantic forest seem to be more affected by deforestation than reptiles and amphibians.

Radio-tracking locations of 25 individuals (13 females; 12 males) and visual observations of nine habituated individuals were used to investigate the spatial organization and movement patterns of the honey badger Mellivora capensis in the southern Kalahari. The home ranges of adult male honey badgers (541±93 km2) were significantly larger than the home ranges of adult females (126±13 km2). Female home-range size was five times larger than predicted from body mass. The extensive home ranges of females were likely to be a function of relatively low prey availability in the semi-arid Kalahari and the long period of cub dependence (12–16 months). While mean home-range overlap in females was moderate (13%) and home-range centres were regularly spaced, females did not appear to actively defend a territory and no direct interactions between females were observed. Scent marking appears to mediate spatio-temporal separation and females show a loosely territorial spacing pattern. In contrast, males did not support the typical mustelid pattern of intra-sexual territoriality but instead encompassed the overlapping home ranges of up to 13 females. Males and females differed significantly in their rate of travel (3.8±0.3 km/h vs 2.7±0.2 km/h), straight line (6.2±0.5 km vs 2.4±0.2 km) and actual distance (13.8±0.9 km vs 7.7±0.7 km) moved during an active period but do not differ in the percentage of their home-range area traversed in a single day (3%). Young males tended to have smaller home ranges (151±45 km2) than adult males and showed a spacing pattern more similar to adult females than adult males. In common with other solitary mustelids, the spatial organization suggests a polygynous or promiscuous mating system.