It is an axiom of science that, in order to measure a process, one ends up changing it in some way. This has been apparent to physicists for a long time but it is perhaps less apparent as a consideration in the biological literature. In many cases the degree to which one changes the process that is being measured is insignificant but there are exceptions. This problem becomes particularly difficult in wildlife biology when one is sometimes faced with gaining information from and protecting small populations of endangered animals. The combination of small sample size, and the high value of each individual to the population, mean that any interference could have negative consequences for the population. The resulting data from the study may have statistical characteristics that make it very difficult to use as a basis for conclusive results.

The use of breeding dens and the early kitten development was studied in a free-ranging population of Iberian lynx Lynx pardinus. Radio-tracking of female lynx allowed intensive monitoring of three litters during 1997, and additional data to be obtained from nine other litters in the previous 4 years. Females gave birth a maximum of one litter per year in spring; nine of the 10 births were recorded between the last 2 weeks of March and the first days of April. Kittens were born semi-altricial and reared for nearly 20 days in secluded hollow trunks. Afterwards, they were moved between different bushes, which served as auxiliary dens, where they began to develop walking and senses. Kittens first consumed meat at the age of 4 weeks, although they were not fully weaned until they were at least 10 weeks old. At the age of 2 months, kittens began to leave the dens, accompanying their mothers on outings. It is suggested that the pattern of den use was related to the stage of kitten development, to fulfil the kittens' need for both protection and space as they grew. Data on kitten denning and development should be accounted for in in-situ and ex-situ conservation programmes for this endangered felid.

The taxonomy and/or natural history of 33 species of scincid lizards from Madagascar are reviewed. Most of the material, consisting of 187 specimens now housed in the Museo Regionale di Scienze Naturali, Turin, Italy, has been collected recently, but some dates to the 19th century. Nine new species are described in three genera: Amphiglossus mandady, A. nanus, A. spilostichus, A. stylu, A. tanysoma, Paracontias manify, P. hafa, P. tsararano and Pseudoacontias menamainty, increasing the known skink fauna of Madagascar (62 species) by 14.5%. All the new species come from the closed forests of north-eastern Madagascar. Additionally, Amphiglossus praeornatus is resurrected from the synonymy of A. frontoparietalis, and Androngo alluaudi, A. crenni and A. elongatus are transferred to the genus Amphiglossus. There is also a key to the species of Paracontias and a new interspecifically variable character for Mabuya. Information on morphological variation, basic reproductive biology, habits and habitats is provided, as available, for all species. Perhaps most interestingly, Amphiglossus nanus n. sp. represents the eighth skink lineage in which an oviduct has been lost and Amphiglossus melanopleura decreases the number of longitudinal scale rows at mid-body and the number of subdigital lamellae with altitude.

Didelphids use vertical strata in different ways, suggesting the existence of a wider variety of niches than simply arboreal or terrestrial. This variety could be related to the differential ability to deal with support diameters and inclines, and might be important for the coexistence of species in local assemblages. From data obtained with the spool-and-line technique, the use of the vertical space and above-ground supports is described in four species of an assemblage of didelphid marsupials in south-east Brazil. The stratification in this and other assemblages in the Atlantic forest and in the Amazon is also compared. Animals were trapped in Serra dos Órgãos, state of Rio de Janeiro, and equipped with a spool-and-line device before release. The paths were tracked by measuring variables related to support diameter, incline, distance and height moved above ground. The diameter of supports used was positively related to body size, and the incline to the dominant direction of movement (horizontal vs vertical). The more cursorial species, Metachirus nudicaudatus, only once moved above ground in > 3200 m of paths followed. Didelphis aurita moved mostly on the ground, but was the only species that occasionally reached the canopy. Philander frenata also moved mostly on the ground, occasionally used the understorey, but never reached the canopy. The more arboreal species, Marmosops incanus, moved mostly in the understorey, but never in the canopy. This pattern of stratification is similar to that observed in other sites in the Atlantic forest and in the Amazon.

During terrestrial locomotion, frogs use two distinct gaits: out-of-phase leg movements associated with slow crawling behaviour and in-phase leg movements during fast jumps. In Rana esculenta, crawling occurs during feeding, while jumping is used as an escape strategy. We examined whether a similar velocity-dependent gait shift appears in swimming R. esculenta. Typically, swimming frogs propel themselves by kicking both hind limbs simultaneously. Observations of out-of-phase leg movements in swimming frogs have been reported, but were usually assumed to be associated only with directional changes. We demonstrate that alternating-leg swimming is used quite frequently and that it results in a significantly lower velocity to the one obtained by using in-phase leg movements. This difference is likely to be associated with energetic costs. Mathematical estimates of positive mechanical work required to move the centre of mass revealed that out-of-phase swimming is energetically less expensive than the in-phase gait at a comparable speed, but may not be efficient at high speeds. Possible explanations for this phenomenon are higher inertial energy losses for in-phase swimming, but at high speeds jet propulsion and an interaction effect may gain importance.

Film sequences were used to measure functional hind limb-segment proportions and joint angles at successive stages of walking strides for 22 species of quadrupedal mammals from five orders. Similar data were obtained from three trotting dogs. For each animal, nine Fourier series coefficients describing the movement of each of the hip, hock and ankle joints were obtained. Multivariate statistical analysis of these coefficients revealed that animals with similar limb-segment proportions move their joints in similar ways. However, it should be noted that animals with similar leg-segment proportions also tended to be phylogenetically related. Most of the discriminatory power was between animals of different stances (plantigrade, digitigrade or unguligrade) and between walking and trotting.

We have read with interest the paper by Alibhai, Jewell & Towindo: Effects of immobilization on fertility in female black rhino (Diceros bicornis) (J. Zool. Lond. (2001). 253: 333--345). As conservation professionals who have devoted our careers to serving the needs of endangered species while employing only the highest standards of veterinary care, we feel it necessary to point out that Alibhai, Jewell & Towindo's experimental design, selection of data for analysis, and basic conclusions may be flawed in several critical ways.

Animals enclosed in amber often provide a unique insight into their surface structure. Such fossils of reptiles are rare and usually not extremely ancient, the earliest being no more than 40 million years (my). A recently discovered 120 my lizard from the Lower Cretaceous of Lebanon provides direct evidence that several common external features of autarchoglossan lizards had evolved by this time. Ecomorphology indicates that the lizard concerned had considerable climbing ability on open surfaces and perhaps in vegetation, and probably lived in a mesic forested environment, something supported by associated plant and invertebrate remains.

Amongst the watering pot shells of the Clavagellidae (Anomalodesmata), the least well known is Humphreyia, with a single species H. strangei. Unlike representatives of the other extant clavagellid genera, i.e. Clavagella and Brechites, however, where juveniles are unknown, there is a single juvenile (plus adult specimens) of H. strangei available for study. The Clavagellidae contains the weirdest of all bivalves, encased in a huge (in proportion to the tiny juvenile shell) adventitious tube with an anterior ‘watering pot’. The juvenile H. strangei is dimyarian with a large foot and byssal groove, and a large pedal gape; the entire body is enclosed within a bag-like periostracum-covered mantle cavity. Upon permanent residence, however, the anterior watering pot component of the adventitious tube is cemented to the chosen substratum, the adductor muscles are lost and the foot and pedal gape greatly reduced; the fourth pallial aperture closes and the pedal gape and the tubules of the watering pot are similarly occluded. Once cementation is achieved, further growth anteriorly is impossible and the adventitious tube is secreted posteriorly, to house the long siphons, and subsequently can be added to as growth increments. Anatomically, H. strangei has a typical anomalodesmatan arrangement of mantle cavity and digestive organs so that it is, probably, a suspension feeder. Inside the pericardium, however, is a pair of unique proprioreceptors which probably serve to monitor rectal tonus and thus control defecation, co-ordinated with siphonal retraction and extension. They probably also prevent over-filling of the capacious rectum. It is believed that cementation and adventitious tube formation occur at the time of sexual maturity and this change in lifestyle between juvenile and adult represents a form of metamorphosis – quite distinct from that which occurs in all bivalves between the pediveliger and juvenile stages – and seems to be unique to Humphreyia and probably Brechites. This family of anomalodesmatan bivalves is thus actually stranger than the already aberrant watering pot would suggest.

We appreciate the concerns of Atkinson et al. regarding our paper (Alibhai, Jewell & Towindo, 2001), and will try to give reassurance that these are unfounded.

With regard to the drugs used, Atkinson et al. are incorrect in stating that all the agents used in combination with M99 were alpha-2 adrenoceptor agonists. Azaperone is a butyrophenone neuroleptic sedative (Hall, Clarke & Trim, 2001). Whether ACP or another sedative agent was used in combination with etorphine hydrochloride, the risks we have discussed remain valid since the most potent and potentially dangerous is the opioid etorphine, which is used for wildlife immobilization in a high dose to produce an anaesthetic rather than analgesic effect (Branson, Gross & Booth, 1995). The stresses involved in the capture circumstances are very likely to carry a further risk for the pregnant rhino.

Foraging behaviour of the black-necked stork (BNS) Ephippiorhynchus asiaticus was studied in 1996 and 1997 in Dudwa National Park (DNP), Uttar Pradesh, India. The storks were observed using tactile and visual techniques to catch fish. Of the 929 fish seen caught in 2 years, 894 (96%) were caught by a tactile mode of feeding and the remaining 35 (4%) by a visual mode of feeding. The rate of foraging attempts by BNS fluctuates with that of season in DNP and coincides with prey abundance. Immediately after the monsoon when the water level was high, BNS had to search for prey more often, as the prey became widespread. Whereas in summer when the water level decreased, the concentration of the fish was higher, which helps BNS to catch fish in quick succession. Prey behaviour and the condition of the wetland determined the selection of the tactile foraging technique among the BNS in DNP. The black-necked storks were more successful in the early hours of the day (06:00–10:00) and they were generally more successful or preferred to feed on medium-sized fish (i.e. 5–10 cm) in DNP. Prey profitability was highest for larger size fish and decreased as the prey size decreased. Principal component analysis showed that prey size, handling time of each prey, the month and water depth determined foraging success in 1996 and 1997.

The integrity of genetically determined morphs of the same species in sympatry may be facilitated by sexual incompatibility (and thus reduced gene flow) between morphs. A test for such incompatibility was carried out between fully metamorphosed and paedomorphic forms of the North American mole salamander Ambystoma talpoideum, which differ greatly in life-history characteristics. A few significant differences were found in the content and timing of courtship between homomorphic (within-morph) and heteromorphic (between-morph) encounters during 2-h observation sessions. However, no significant differences were apparent in the incidence of spermatophore deposition (which is an index of the potential for sperm transfer), both as recorded at the end of observation sessions and as determined for pairs left to interact for 16 h overnight. Fertilized and viable eggs were laid by females of both morphs that were exposed to spermatophores from either homomorphic or heteromorphic males, but not to both. We conclude that sexual incompatibility probably plays little or no role in preventing gene flow between sympatric morphs of A. talpoideum. Gametic incompatibility may also be of little or no importance. We suggest that differences in the phenology of breeding are of greater significance in the maintenance of discrete morphs of the mole salamander in sympatry.

Over the past 50 years European populations of the lesser horseshoe bat Rhinolophus hipposideros have severely declined, probably because of the loss of foraging habitat. To date, studies of the foraging behaviour of this species have been limited as its low mass (4–8 g) precluded the use of radio-telemetry because commercially available radio-transmitters exceeded 10% of its body mass. In this study, radio-transmitters weighing < 0.35 g were built. These increased the body mass of the animals from 4.5% to 8.1%, with no demonstrable adverse effect on their flight behaviour. The habitat selection of eight female lesser horseshoe bats was studied in Monmouthshire, U.K. The bats had foraging ranges between 12 and 53 ha (100% kernel). Although one bat foraged 4.2 km from the roost, for 50% of the time tracked bats were recorded within 600 m of the nursery roost. The estimated density within 200 m of the roost was 5.8 foraging bats/ha. This decreased to 0.01 bats/ha at 1200 m. Compositional analysis revealed that this species used woodlands, predominately broadleaf, more than any other habitat. In addition, the bats foraged in areas of high habitat diversity. Conservation management of this species should concentrate on such areas within 2.5 km of the nursery roost.

Population subdivision among mouflon sheep Ovis gmelini ewes and ranging behaviour of rams were investigated in the Caroux-Espinouse massif (southern France). Radio-tracking data from 32 ewes and 41 rams monitored in the course of a 10-year study were used to identify female units and to document the movement patterns of males during the rutting period. Cluster analyses, based on the distance between arithmetic centres and on the degree of overlap of locations, were computed. Ewes were found to be partitioned in three or four main units to which they were faithful. Movement patterns of rams were much more variable: half were faithful to a rutting range, overlapping one or more unit(s), and the others were not. We suggest that a combination of social and spatial attachment could explain this ranging behaviour. Females build up their spatial habits following their mothers and, when adults, they occupy the area where they have been reared. Their spatial attachment is consistent with their social preferences for other females. Males balance between antagonistic spatial attachment and social preferences for other males, and differ in terms of spatial behaviour. The genetic implications are discussed.

Moult and pre-migratory fuelling are energy-demanding processes in birds, usually showing a limited degree of temporal overlap. The barn swallow Hirundo rustica, a strictly aerial insectivorous, long-distance passerine migrant (15–20 g), accumulates conspicuous energy stores (up to 30–40% of lean body mass) before embarking on the autumn migratory flight between western European breeding areas and African sub-Saharan winter quarters. During the pre-migratory fattening period (September–October), both adult and juvenile birds replace body (contour) feathers. The seasonal overlap between moult and energy storing was studied in barn swallows in Italy. Data from 23 000 birds, measured during eight study seasons, were analysed to investigate whether the simultaneous replacement of body feathers and fuelling was associated with reduced energy stores. Among all age and sex-classes, non-moulting individuals showed a relatively greater level of energy stores compared to moulting birds. This was particularly evident during the last phase of the pre-migratory fuelling period, when birds were building up energy stores at maximum rate. The proportion of moulting birds was greater among adults as compared to juveniles, and adult males moulted faster/earlier than females. The frequency of individuals with growing feathers among the innermost primary feathers was greater among birds replacing body feathers. The presence of moulting wing feathers in adult birds only marginally affected energy stores, as compared to the replacement of body feathers. Hence, our data show a negative association between replacement of contour feathers and energy stores in a long-distance migratory bird, suggesting that a physiological trade-off between these two energy-demanding processes may occur, and that a late moult may ultimately have a negative effect on the successful termination of the migratory journey.

The age of crabeater seals Lobodon carcinophagus was first estimated from the teeth by counting layers in the dentine in transverse sections of the canine teeth. Reliable estimation of age from canine teeth in animals older than c. 10 years is not possible because of root closure at a relatively early age. The present investigation compared various methods of estimating age by examination of tooth sections from a large sample of crabeater seals from Marguerite Bay, Antarctic Peninsula. The most reliable estimations were shown to be obtained by counting layers in the cementum of the third postcanine tooth. Ages of 1233 animals (58% female, 42% male) ranged from 0.5 to 39 years. It is suggested that time- and age-specific historical material is potentially relevant for contemporary growth and life-history studies as well as for trophic relations or concentrations of pollutants in animal tissues.

The spatial resources used by a population of lace monitors (Varanus varius) were examined over an active season (September–April) and an inactive season (May–August). In total, 33 monitors were observed, of which 23 monitors were radio-tracked for up to 11 months. Radio tracking provided new information on the spatial ecology of V. varius. During summer (December–February), V. varius moved often, over large (184.5 ha) overlapping home-range areas. In the intermediate seasons of spring (September–November) and autumn (March–May), monitors moved less often and used less than 39% of the summer home-range. Finally, in the cold winter season (June–August) many monitors did not move at all and most used less than 5% of their summer home-ranges. The thermal environment, and reproductive status of V. varius affected its use of space, and the importance of these factors varied seasonally. No spatial segregation was noted between monitors, with home-ranges overlapping on average for eight other radio-tracked monitors. The home-range size of V. varius was accurately predicted using published data on body mass and home-range size for the Varanidae.

Spatial distribution of a deep-sea crinoid Pentametrocrinus tuberculatus was investigated in the Izu-Ogasawara Arc, western Pacific. Photographic and video recordings coupled with direct visual observations were conducted from the manned submersible Shinkai 2000 between the depths of 1340 and 1840 m on the south-western slope of the An'ei Seamount (29° 15'N, 138° 37'E). Close to the seabed, temperature, salinity and dissolved oxygen concentration were in the range of 2.2–3.1 °C, 34.38–34.52 and 1.40–2.15 ml l−1, respectively, and the bottom substratum consisted mainly of flat-surfaced lava rocks covered by thin manganese oxide crusts. Pentametrocrinus tuberculatus individuals were attached to the surface of lava rocks with their five arms extended horizontally and showed no recognizable movement. Incidence of arm loss (which may indicate mechanical damage or predation pressure) was very low. The highest density of P. tuberculatus occurred in an area termed ‘terrace’ where the slope was reduced above a relatively steep hillside. Analysis of nearest neighbour distances showed that P. tuberculatus individuals generally demonstrate regular spacing with either no or partial overlap of their ambits (= circular area defined by the extended arms), due to a combination of aggregative tendency (to form patches) and behavioural spacing among individuals within a patch. The observed pattern of distribution was clearly different from the pattern expected from random occurrence of individuals in space. Thus, on spatial scales of 101 to 103 m, P. tuberculatus showed a contagious or aggregated distribution, while on smaller scales and in areas where crinoids were concentrated they showed a combination of patch formation and regular distribution of individuals.

The aim of this study was to investigate age-related growth in the Sumatran elephant Elephas maximus sumatranus and to use the derived relationship to determine the age structure of the wild elephant population in Way Kambas National Park (WKNP), Sumatra. Shoulder height, forefoot circumference and diameter of dung bolus were found to be related to age of captive Sumatran elephants using the Von Bertalanffy growth function. All length measurements were highly correlated with age in the Sumatran elephant and provide growth models for determining the age structure of wild populations. Female captive elephants reached their growth plateau earlier than male elephants who continued growing throughout the ages observed. There was no clear evidence of a secondary growth spurt in male elephants. The growth model relating dung diameter to age was used to predict the age structure of the wild elephant population in WKNP from dung measured along random line transects. The wild elephant population in WKNP is young and dominated by sub-adults (between 5 and 15 years of age). There are marked differences between the age structure of the population as revealed in the current survey and that reported from previous studies, suggesting that changes have occurred within the population in the intervening period. The use of dung diameter to predict age offers a robust field technique for use in situations where direct observations are limited, and the use of other age estimation methods is impractical. It is easily coupled with dung counts for estimating the size, age structure and biomass of elephant populations, and has considerable potential for investigating the effects of poaching on age structure and identifying where priority action should be directed in human–elephant conflict situations.

Antechinus agilis mate within a 2–3 week, highly synchronized period each year, with copulation consisting of short bouts of thrusting interrupted by longer bouts of rest during which time the male remains mounted. In the laboratory, mating can last as long as 8–12 h without any break in intromission, with captive paired animals generally copulating once per day (Woolley, 1966a,b). The mating programme used in this study examines the effects on copulatory behaviour of changing: (1) timing of access relative to ovulation; (2) order of mating; (3) the delay between the first and second males' mating access. The total mount time was divided into quarters and the changes in behavioural patterns assessed by examining changes in the frequency of five key activities; thrusting, pelvic side-to-side movements, walking, female resistance, and dismounts. No significant differences were observed in the time from initial pairing to first mount regardless of mating order or time of mating, nor was the total time mounted significantly affected by mating order. Significant reductions in the total time mounted were evident, however, for those males mating closer to the time of ovulation. The behavioural strategies associated with copulation in A. agilis significantly enhance arguments for equality between males and females in determining overall mating strategies.