A silty matrix still adheres to the external left half of the specimen.For logistic reasons it has not yet been possible to clean, reconstruct and fully study the cranium, so here we offer only a preliminary assessment. For similar reasons the pelvic fragments and incisors are not described here.A remarkable feature of UA 31 is its great anteroposterior length (204 mm) in relation to its maximum cranial breadth (130 mm). The greatest cranial breadth is located at the level of the supramastoid crests and with the length gives a cranial index of 63.7. Although the skull has not yet been restored and cleaned, tentative estimates suggest an endocranial capacity of 750-800 cm 3 .The neurocranium is low compared to its maximum length, but high relative to its transverse diameter. The upper-middle facial skeleton is slightly concave and marked prognathism is evident in the subnasal region. The face is narrow and short. The palate is shallow, wide and short, with its length reaching less than one-third of the total cranial length as seen in basal view. The very heavily built, forwardly projecting supraorbital torus is arched and thickened medially and over the middle of each orbit (17 mm). The minimum frontal breadth is 84.0 mm. Substantial elevation of the forehead and absence of sagittal keeling are evident in frontal view.The frontal squame is well curved to bregma and the vault outline becomes gradually flatter on the parietals. The temporal lines originate in the form of a raised crest at the posterolateral margin of the supraorbital torus, but disappear early on the parietals. A slightly thickened angular torus is present. There is no nuchal torus, but a modest external protuberance is detectable where opisthocranion coincides with inion.The specimen also shows a modest protrusion of the mastoidsupramastoid-auriculare complex lateral to the temporal squame. The mastoid process is short and broad. In coronal section, the parietal walls converge slightly inferiorly (Fig. 3); there is a high positioning of the maximum biparietal breadth (125.0 mm); the parietal thickness progressively decreases up to the midline (from 7.4 to 6.0 mm); the thickness at bregma is 6.3 mm.The cranium from Buia shows the following combination of features, which are characteristic of the African specimens referred to Homo erectus and Homo ergaster 11,12 : long ovoid braincase with low endocranial capacity; greatest cranial breadth across the supramastoid crests; massive supraorbital torus; opisthocranion coincident with inion. One trait typical of H. sapiens is the high position of the greatest biparietal breadth with parietal walls converging slightly inferiorly. This mosaic of primitive and progressive features increases the known morphological variation reported for earlymiddle Pleistocene Homo crania and urges caution in the specific assessment of UA 31. Until we have conducted a full comparative study, we prefer to leave open its specific allocation.Given its chronological position, in the middle of the time interval 1.4-0.6 Myr, from w...

Comparatively little is known about reproductive behaviour in wild sharks as it has proved extremely di¤cult to study, especially in large pelagic sharks. Here we describe annual courtship-like behaviour in the second-largest ¢sh species, the basking shark (Cetorhinus maximus), from 25 separate episodes observed and tracked during a ¢ve-year study period (1995^1999) o¡ south-west England. Social behaviours observed between paired, or three or four, sharks were consistent with courtship behaviours seen in other shark species, namely nose-to-tail following, close following, close £ank approach, parallel and echelon swimming. Mature individuals between 5 and 8 m total body length (L T ) exhibited these behaviours whereas smaller sharks (3^4 m L T ) did not. Lead individuals were identi¢ed as female on a number of occasions and interactions were prolonged; the longest continuous observation of socializing was 1.8 h, although intermittent track data indicates bouts may last for up to 5^6 h. Locations of courtship-like behaviour events were not distributed randomly and were signi¢cantly associated with thermal fronts. Our results indicate that putative courtship behaviour occurs between May and July along oceanographic fronts, probably as a consequence of individuals aggregating to forage in rich prey patches before initiating courtship. Thus, locating the richest prey patches along fronts may be important for basking sharks to ¢nd mates as well as food in the pelagic ecosystem. As courtship-like behaviours occur annually o¡ south-west England we speculate that this region may represent an annual breeding area for this protected species, but mating itself probably takes place at depth as it was not seen at the surface.

Movements and distribution of cod (Gadus morhua) in the southern North Sea and English Channel: results from conventional and electronic tagging experimentsThe sub-structure of Atlantic cod (Gadus morhua) stocks in the North Sea has important consequences for fisheries management as the Common Fisheries Policy moves towards a more regional approach. We investigated the movements, distribution and behaviour of cod in the southern North Sea (ICES IVc) and English Channel (ICES VIId) by re-analysing historic data from conventional tagging experiments, and by conducting new experiments with electronic tags. Cod tagged and released in IVc showed a northwards shift in distribution during the feeding season consistent with a homing migration away from spawning grounds along the coasts of the UK and the Netherlands. In contrast, cod tagged and released in VIId did not exhibit a consistent pattern of seasonal movement. Many cod released in VIId were subsequently recaptured close to their release position, although some moved out of the Channel and into the southern North Sea. Overlap between the recapture areas of cod released in the different management areas was no more than 25% in either the spawning or feeding season. Behavioural data from electronic tags suggest that cod in IVc make use of tidal streams to migrate northwards and eastwards in spring, whereas selective tidal stream transport was rarely exhibited by cod tagged and released in VIId. Overall, the evidence suggests that there are behavioural differences between cod in IVc and VIId that limit the mixing of cod from these two areas during the feeding and spawning seasons. sub-groups were found in four areas: the northern North Sea, the Moray Firth, Flamborough Head and the Southern Bight, broadly corresponding to the ICES stock assessment and management areas IVa, IVb and IVc/VIId. These areas are large, however, and evidence is increasing that the movement of cod is much more restricted than previously

The effects of changes in shading (through riparian canopy removal and re-growth) on juvenile salmon, Salmo salar L., trout, Salmo trutta L., and grayling, Thymallus thymallus (L.) populations, and macroinvertebrate biomass and species composition in a chalk stream in southern England were examined. Low levels of in-stream weed growth, because of shading by closed tree canopy, diminished macroinvertebrate production and diversity. 0+ salmon and trout had lower densities under closed canopy, relative to adjacent open sites with substantial weed cover, where fish were also found to be larger. Canopy removal positively affected the growth of aquatic macrophytes and the availability of potential prey for juvenile salmonids. The findings have implications for the management of chalk streams, in particular, that riparian tree canopy should be managed to prevent complete closure, and excessive cutting of weed should be avoided where salmon production is below sustainable levels.K E Y W O R D S : canopy, chalk stream, grayling, macroinvertebrate, salmon, trout.

The post spawning behaviour of sea trout Salmo trutta was studied over a 2 year period in the river and estuary of the River Fowey, south-west England. Forty-five sea trout kelts were trapped immediately after spawning in December and intraperitoneally tagged with miniature acoustic transmitters. The subsequent emigration into coastal waters was monitored using acoustic receivers deployed throughout the river catchment. The levels of gill Na þ K þ ATPase activity in sea trout kelts sampled at the same time as the tagged fish were within the range of 2Á5 to 4Á5 mmol Pi per mg protein per h indicating that the post-spawning fish were not physiologically adapted to salt water. The tagged kelts were resident in fresh water between 4 and 70 days before entering the estuary. Sixty two per cent of the tagged kelts subsequently migrated successfully into coastal waters, with a higher success rate for male fish (75%) than females (58%). There was a significant size related difference in the run-timing of the kelts with the larger fish moving more quickly into coastal waters after spawning than smaller fish. Seaward migration within fresh water was predominantly nocturnal and generally occurred in conjunction with increasing river discharge and rising water temperature. Migration through the estuary continued to be predominantly nocturnal and occurred during an ebbing tide. Residency within the estuary varied amongst individuals although it was invariably short, with most fish moving out into coastal waters within one to two tidal cycles. Five tagged kelts returned from the coastal zone and re-entered fresh water during April and June. Marine residence time varied between 89 and 145 days (mean 118 days) and the minimum estimated marine survival was c. 18%. One of these sea trout was subsequently recaptured after successfully spawning in the vicinity where it had been previously tagged demonstrating a degree of spawning site fidelity. # 2005 Crown copyright