Between September 1983 and June 1987, 175 caribou (Rangifer tarandus caribou) from the George River caribou herd were captured and radio-collared. By October 1987, 42 of those had died: 15 were apparently killed by predators, 8 were legally shot, 4 appeared to have died proximally from malnutrition, 1 died from an accident, and cause of death could not be determined for 14 others. Seasonal and annual survival rates were estimated most precisely for adult (≥ 24 months) females, but a consistent trend towards lower survival at the end of the study period was evident in all age and sex groups. Decreasing summer survival was responsible for the lowering of annual survival rates, as winter survival was relatively constant over the study period. Using our calculated age-specific survival rates, and estimates of age-specific fecundity, the survival–fecundity rate of increase (rs) for the female segment of the herd was 0.12 for 1983–1984 and −0.01 for 1986–1987. Decreasing summer survival thus appears to have been of greatest importance in reducing rs to zero and halting the growth of the herd. Density-dependent mechanisms acting on the George River caribou herd appear different from those described for caribou herds regulated by winter forage.

The Rivière George caribou herd (RGCH) was numerically low during the middle part of the century but apparently erupted in the sixties and the seventies. Puberty was early and pregnancy rate was high among RGCH females from 1973 until the mid-eighties when fecundity decreased significantly, in particular among sub-adults; productivity remained low in 1992. Autumn calf:female ratios reflected this trend in pregnancy, exceeding 50 calves: 100 cows between 1973 and 1983, but dropping thereafter progressively to a low of 24 in 1992. In 1993, this ratio rebounded back to 42. Annual adult survival rate of radio-collared females was high (0.95) at the beginning of the monitoring in 1984, but exhibited a variable but declining trend until 1992. Simulations were conducted to estimate the demographic trend of the RGCH between 1984 and 1992, using annual survival rates of radio-collared animals and annual autumn calffemale ratios to estimate calf production. Age structure played a minor role in estimating the finite rate of increase (Lamda). According to the simulations, the RGCH increased in size until 1987, and showed a slight decrease thereafter. The herd should have decreased by 12-15% between 1988 and 1993, according to the simulations. Productivity first caused a decline in Lamda, but in recent years decreased survival contributed slightly more than productivity to the reduction in Lamda. Estimation of the herd size by means of aerial censuses in 1976, 1984, 1988 and 1993 suggested a similar pattern in demographic trend, differences being statistically meaningless. We speculated on the future of the RGCH, that could have erupted after many decades of unfavourable weather. The herd will exhibit a rapid descent to low numbers if wolves show a numerical response to current caribou abundance, or if lichen availability on the winter range decreases due to competition with the adjacent and increasing Rivière aux Feuilles herd; otherwise it will exhibit dampened oscillations, whose amplitude will depend on the time-lag of vegetation reaction to decrease grazing pressure on the summer range

We determined demographic attributes of an arctic hare (Lepus arcticus) population in the Southern Long Range Mountains of Newfoundland. Twenty-four adult hares were livetrapped, radio collared, and monitored from July 1981 through July 1984. From livetrapping and an aerial census we estimated mean density at about 1 adult/km2. A single litter was born annually and natality averaged 3.0 young/female. Mean dates of conception and parturition were 19 April and 8 June, respectively. Annual survival rate of radio-collared adults was 0.78; calculated 1st-year survival rate of juveniles was 0.15. Natural mortality was mainly from red foxes (Vulpes vulpes). One-year-old individuals comprised 18% of 68 adult hares shot in May. Sex ratios were biased towards males in both trapped and shot samples. Summer home range size of three adult females (52–69 ha) averaged half that of three adult males (116–155 ha). Movements increased in March and April with onset of breeding activity. It is unlikely that winter food shortage affected survival, as hares collected in May had both kidney fat and relatively high levels of bone marrow fat. Annual differences in these condition indices were, however, reflected in several reproductive parameters. We hypothesize that the single litter and small litter size which combine to give Newfoundland arctic hares the lowest reproductive potential of any known hare population are a consequence of (i) the late spring that delays onset of breeding until April, and (ii) the shorter day length in April and hence lower gonadotrophin levels at this southernmost limit of distribution. We believe that predation on juveniles is the paramount limitation on growth of this arctic hare population, and primarily responsible for the stabilization at low densities.

We analyzed scats (n = 679) and stomach contents (n = 25) collected from 1980-2003 to assess the relative frequencies of food types utilized by Newfoundland Marten (Martes americana atrata) during summer and winter. Meadow Voles (Microtus pennsylvanicus) were the most prevalent food item occurring in 80% and 47.5% of samples from summer and winter, respectively. Apart from Snowshoe Hares (Lepus americana), which occurred in 28% of winter samples, all other food types occurred in <16% of samples during each season. Diet breadth was widest during winter and may be related to a lower availability of Meadow Voles during this time of year.

We evaluated multiscale habitat selection by threatened Newfoundland (Canada) martens (Martes americana atrata) across landscapes composed of a range of habitat types to test the hypothesis that martens preferred mature (61-80-yr-old) and overmature (.80 yr) conifer stands. At both the landscape and stand scales, adult ( L 1 yr) resident martens (n 5 58, 92 yr-specific home ranges) selected for, or used in proportion to availability, a broad range of habitat types, including recent cuts M 5 yr old, regenerating forest ,6.5 m, precommercially thinned stands, and mature and overmature forest. Marten home ranges were not dominated by mature and overmature forest; median availability of mature and overmature forest within individual home ranges was only 30% (range 5 11-76%). Age distributions were not different among martens with high, intermediate, and low quantity of mature and overmature forest in their home range; our data do not indicate that martens inhabiting ranges with little mature and overmature forest were compromising fitness. Habitat selection by martens in Newfoundland was more generalized than has traditionally been inferred, and we suggest that inherent landscape fragmentation, in combination with absence of many predators and competitors (i.e., ecological release), caused the Newfoundland marten to evolve to use a more generalized habitat niche than many mainland populations of American martens. We recommend that landscapes suitable for marten not exceed .29% younger aged forest. Maintaining resident martens in landscapes where forest harvesting is occurring requires prescriptions that recognize the highly fragmented nature of the natural landscape, the prevalence of mature and overmature forest, younger forests, avoided landcover types, and human access.

Adaptive management of social-ecological systems requires integration and collaboration among scientists, policy makers, practitioners, and stakeholders across multiple disciplines and organizations. Challenges associated with such integration have been attributed to gaps between how human systems are organized and how ecosystems function. To address this gap, we explore the application of information ecology as a theoretical basis for integrating human systems and natural systems. First, we provide an overview of information ecology with reference to its relationship with information theory and how we define "information." Principles governing whole-part relationships, i.e., holons and holarchies, are then used to develop a general information flow model for evolutionary, complex adaptive systems. This general model is then applied to examine a number of issues related to science-policy integration and in the development of a reference framework for practical application in adaptive management. A number of additional considerations for practical use of the framework are also discussed.

The Humber River Basin (HBR) responds to critical issues facing decision-makers concerning climate change and the sustainability of the HRB and its environments. The basin and its component watersheds represent an excellent platform from which to develop and integrate science and policy. The research presented herein is considered an important prerequisite to developing a more scientifically sound and ecosystem-based strategy for management of the basin ecosystem(s). Scientists and policy-makers in Newfoundland and Labrador, Canada recognise that our most vital natural resources are directly or indirectly linked to water in the form of marine, fresh and estuarine systems.

We report on the first capture of the Southern Red-backed Vole (Clethrionomys gapperi), the eleventh non-native terrestrial mammal established on the island of Newfoundland over the last 150 years. Red-backed Voles may have been accidentally introduced by unknown sources in pulpwood imports or may have been deliberately introduced in an attempt to augment the depauperate small mammal fauna as a vigilante recovery effort for the endangered Newfoundland Marten (Martes americana atrata). We anticipate significant utilization of the Red-backed Vole as prey by both Newfoundland Marten and Red Fox (Vulpes vulpes) with associated demographic responses within and between these species. Red-backed Voles will likely change habitat utilization patterns for the endemic subspecies of Meadow Vole, Microtus pennsylvanicus terraenovae.