No description available.

A study was undertaken to determine and compare the grazing ecology of goats, red deer and South American camelids when pastured on three contrasting vegetation types typical of hill and upland ecosystems in Scotland: an established sown sward (<I>Lolium perenne </I>dominated); an indigenous grassland (<I>Nardus stricta</I> dominated); and a dwarf-shrub community (<I>Calluna vulgaris</I> dominated). Plots at each site were grazed in sequence with data collected during one spring (May/June) and two summer (August/September) experimental sessions. For each period at each site plant species composition, canopy structure and herbage biomass were characterized. Samples of the diet selected by up to five mature castrated goats, red deer and guanacos fistulated at the oesophagus were used to establish diet composition and <I>in vitro</I> digestibility of the diet. Herbage intake and diet digestibility of an additional five intact animals of each species were determined using n-alkane faecal markers. Total grazing time was estimated using vibracorders and bite rate data collected by observation. The influence of vegetation type and season on the diet composition and ingestive behaviour of each species was evaluated. Between species comparisons on each vegetation type were also made. Differences in selectivity for and against vegetation components indicate that guanacos are principally grazers while goats and red deer are intermediate feeders. On the Lolium sward the goats and red deer selected green leaf of both monocotyledonous and dicotyledonous plants whereas the guanacos selected only graminoids. On the Nardus community the main dietary component for all three animal species was broad-leaved grasses selected from the species-rich intertussock areas. On the dwarf-shrub community there was strong selection for and against graminoid species and <I>Calluna vulgaris </I>respectively, with the dominant species of plant only making a substantial contribution to the red deer diet. Where differences between estimates of diet digestibility were identified the digestibility of material consumed by each species of animal on the sown sward was higher than that from either indigenous community; and the digestibility of material selected from the indigenous grassland was higher than that from the dwarf-shrub community.

A study of forest elephant ecology was initiated in the remote Ndoki Forest of northern Congo. The goal was to identify the ecological determinants of elephant distribution and ranging, and to determine the impact of human activity, at a relatively intact site. Data from a local, intensively surveyed site, and repeated extensive foot surveys over a 253km swathe of the Ndoki Forest, which traversed and northwest-southeast drainage gradient, revealed a spatial and temporal partitioning in the availability of resources important o elephants on several scales. Dicotyledon browse was most abundant in open canopy <i>terra firma</i> forest, light gaps, and swamps, while monocotyledon food was not concentrated in <i>terra firma </i>forest to the southeast, and was super-abundant in localised swamp patches. Mature and old leaf abundance was correlated with rainfall, but new leaves were not. During low rainfall periods, new leaf production was highest in the southeast, becoming widespread as rainfall increased. Forest clearings, clumped in the northwest, contained high mineral abundance in seep-hole water, most concentrated during dry periods. Fruit availability was negligible in swamps, high in closed canopy <i>terra firma </i>forest, and while correlated with rainfall, its temporal and spatial distribution was highly irregular. Drinking water, confined to rivers, was widespread and abundant. Elephants ate leaves, bark, wood, stems, roots, and fruit from over 350 plant species. Leaves dominated food selection, and browsing rates were highest in open canopy forests, particularly swamps. Fruit consumption increased dramatically as its availability increased. Elephants constructed trail systems that allowed efficient exploitation of high payback resources, notably water, minerals, and fruit. Applied research is required to identify potential conservation sites, improve survey methods, quantify the impact of logging on elephant ecology, ranging and demographics, and to understand the role of elephants in ecosystem function, and how it is disrupted by range restriction, population reduction, and logging.

This study presents a description of the life cycle and biology of the macadamia nutborer <I>Ecdytolopha torticornis</I> (Meyrick), and identifies factors which influence the abundance of the nutborer in a macadamia orchard in Atirro Farm, Turrialba, Costa Rica, during 1991-1993. Four larval instars were identified. The life cycle was completed in 36 days, and sex ratio of 1:1 was recorded. Fecundity was found to be 37 eggs per female. Macadamia clones differed in their susceptibility to nutborer attack: clones 344 and 246 were more heavily attacked than clones 508 and 660. The maximum level of nut damage was 37%. Nut damage decreased between years and sites. Four larval parasitoids (<I>Apanteles</I> I, <I>Apanteles</I> II, <I>Pristomerus</I> sp. and <I>Ascogaster</I> sp.) and one egg parasitoid (Trichogrammatidae) were reared. Parasitism presented a delayed-density dependence response to nutborer, abundance, as measured by nut damage. <I>Apanteles</I> I accounted for 40% of parasitoids reared, while <I>Ascogaster </I>accounted for 50%. Predation was found to be negatively affect the abundance of larvae in nuts on the ground, and the fire ant, <I>Solenopsis</I> invicta, was considered to be the most important predator. The nutborer preferred to oviposit upon nuts on branches within the first three meters from the ground, and the number of eggs laid was affected by weed management, more eggs were found when husk was applied as a mulch than in unweeded plots. Nitrogen content decreased with nut maturation. The nitrogen and tannin content of different clones were found to have no effect on the abundance of the nutborer. However, proanthocyanidin content and dry matter may affect larval performance. Evidence is given which supports the hypothesis that <I>E. torticornis</I> is a native species rather than an introduced pest. Four factors affect <I>E. torticornis</I> abundance, the quality of its food source, the quantity of the food source, predation and parasitism.

<I>Nicrophorus vespilloides</I> Herbst (Coleoptera; Silphidae) bury the carcasses of small vertebrates which provide food for their developing larvae. The use of such a valuable, yet rare and ephemeral resource has led to the evolution of a complex and variable social system that ranges from parental care by a single female, through monogamy and biparental care to co-operative breeding. Concentrating on broods raised by either one or two parents, I used laboratory and field populations to study the benefits of parental care in <I>N. vespilloides</I>. A decline in the apparent clutch size of <I>N. vespilloides</I> with successive generations of laboratory culture was correlated with an increase in numbers of the phoretic mite <I>Poecilochirus</I> <I>davydovae</I> Hyatt. Beetles with their mites removed had significantly larger apparent clutch sizes than beetles whose cargoes were left intact. Observations of mite behaviour suggested that the mite eats the eggs of its host. A comparison of the apparent clutch sizes of beetles in the presence or absence of carrion fly eggs implied that <I>P. davydovae</I> is predatory on burying beetle eggs in the field. Adult beetles eclosing from larvae raised on small carcasses (where brood size is regulated by the parents killing a proportion of the first instar larvae) were at least as big as beetles that successfully secured, and raised broods on carcasses placed in the field. There was no difference in the degree of filial cannibalism in broods raised by a single female or a male and female. In the laboratory, broods tended by a single female were vulnerable to being taken over by conspecifics resulting in the death of the larvae and the production of a smaller replacement brood. The additional presence of a caring male completely eliminated this risk.

Raptor predation was studied by direct observation of sparrowhawks, peregrines, and merlins, hunting a known prey population, and subsequent recovery of kills. Raptor predation was shown to be the most significant cause of mortality in most wader species. Kleptoparasitism of raptors carrying prey, by carrion crows, significantly increased the over winter mortality of some waders. Redshank populations were most affected by raptor predation; over 50% of the total population and over 90% of the juvenile population were taken in two winters. No selection for body size in redshank was found, but juveniles were more likely to be killed by raptors. This was a consequence of adult redshank risk-averse foraging, and excluding juveniles from low-risk and low-feeding profitability areas. Juveniles, even though feeding in a relatively profitable area compared to adults, still showed risk-prone foraging within that area. Flocking reduced an individual redshank's probability of being killed by a raptor. Larger flocks were preferentially attacked, but an attack was significantly more likely to succeed on a smaller flock. Within a larger flock a redshank was less at risk through the 'dilution' effect, vigilance effects (which were shown to be a direct consequence of flock size) and probably also the 'confusion' effect. Redshank did not gain any foraging benefits within larger flocks. Reduced individual risk of predation appeared to be the main reason for flocking. Redshank responded differently when attacked in a similar way by the three species of raptor. During an attack the probability of capture depended on the escape response. The response that was most likely to lead to escape from a sparrowhawk was most likely to lead to capture on peregrine attack, showing that raptor discrimination was an important determinant of over-winter survival.

Validated sampling, measuring and ageing techniques were developed to facilitate a study of the population dynamics of larval and young-of-the-year juvenile bass. Drag seines were proven non size-selective when used for sampling juveniles. Repeated measurements of larvae and juveniles, following death and during preservation, indicated that their lengths decreased. Accordingly, conversion factors were calculated to permit live lengths to be determined from measurements of preserved length. Bass eggs were incubated from fertilization at 15 temperatures from 6.8-20.6<SUP>o</SUP>C, and a system to age eggs by reference to developmental stage and temperature was developed. Laboratory reared bass larvae, exposed to temperature and photoperiod regimes similar to those encountered by wild larvae, deposited growth increments on their sagittal otoliths at a rate not significantly different from one per day following mouth opening. In order to permit egg fertilization (birth) dates to be back-calculated from increment counts, relationships between environmental temperature and the time from fertilization to mouth opening were determined experimentally. Larvae, of 5.0-11.0 mm notocord length and 10-45 days old, were captured during ichthyoplankton surveys in the Bristol Channel in May 1989. Their distribution and growth was reported, and hypotheses for their recruitment to estuarine nursery habitats proposed. Back-calculated birth dates of juveniles recruiting to three sites in South Wales during 1988 and 1989 were not significantly different, suggesting they originated from the same larval populations. Growth rates at different sites, and for cohorts at the same site but with early, mid or late season birth dates were rarely significantly different in 1988 or 1989. However, comparison on growth in 1988 and 1989 indicated that it was significantly faster at all sites in 1989. This was attributed to consistently higher water temperatures in 1989. The effects of birth date and growth rate on subsequent overwintering mortality were discussed.

No description available.

The aim of this thesis was to determine how ecology shapes spatial learning and behaviour in different populations of three-spined sticklebacks (<I>Gasterosteus aculeatus).</I> These fish occupy a wide range of marine and freshwater habitats and their behaviour is, therefore, subject to a range of different selective pressures. Under experimental conditions, fish sampled from three river habitats exhibited preferences for developing algorithmic behaviour (i.e. learning a series of turns or movements), while fish from two pond habitats showed a preference for visual landmark information. Furthermore, two river populations were found to be more adept at using direction of flow to orientate than two pond populations. Habitat survey for each of he populations tested were carried out. These surveys were used to assess candidate ecological correlates that may be responsible for the observed population learning and memory differences. In addition, morphological assessment of the different populations revealed a degree of morphological variation. Inferences about feeding ecology and predation pressures in the various habitats were drawn from these results. To determine whether the observed variable spatial ability arose through genetic differences between the populations or was learned anew by each generation, artificial breeding and rearing experiments were carried out. These revealed that the spatial cue preferences and abilities of the three-spined sticklebacks were influenced by an interaction between experience during development and inherited factors. The results of this work have revealed intra-specific variation in both spatial cognition and morphology among natural populations of three-spined stickleback. This variation appeared to be related to the different selective pressures exerted by their respective environments.

I have examined variation in the reproductive behaviour of the Panamanian bishop <i>Brachyrhaphis episcopi</i>, a tropical poeciliid fish. Firstly, I explored male mate choice, which is an understudied but rapidly growing area of research. I found that even in this promiscuous species, males were choosy about the females that they mated with. Males showed mating preferences for unfamiliar females and for larger females, preferences that could increase the number of offspring sired. Male mate choice showed some degree of plasticity both seasonally and on a much shorter time-scale, which may reflect fluctuations in the costs and benefits of being choosy. Secondly, I examined between and within population variation in reproductive behaviour to find out whether ecological factors driving variation in closely-related species could be more generally applied. Field observations revealed that, despite large inter-population variation in predation pressure, there was little inter-population variation in male reproductive activity or behavioural repertoire. Instead, male reproductive behaviour showed significant variation within populations, which may be linked to local operational sex ratio. Reproductive success in this thesis is likely to be linked to male-male competition rather than female choice. Thirdly, I examined variation in individual reproductive investment by males. I looked at both reproductive tactics and other traits associated with mating success, such as testes mass. I found relationships between male size and mating behaviour, but no size-based differences in relative testes investment.