Ecological Effects Of Human Disturbances On Moose And Roe Deer And Their Adaptation Mechanisms At Multiple Spatial Scales

Ecological effects of anthropogenic disturbances on moose (Alces alces cameloides) and roe deer (Capreolus pygargus bedfordi) at multiple spatial scales were studied by locating 60 2-km transects, surveying 613 10-m×10-m plots and 5265 2-m×2-m subplots, following 105 snow tracks in total length of 53 000-m, collecting 184 fecal samples and assaying 260 snow-urine samples, in the northwestern Lesser Kingan Mountains by using both ecological techniques (such as field survey, modem landscape ecological techniques, mathematic ecological modeling, population ecology techniques, behavioral ecology techniques etc.) and nutritional ecological techniques (such as high performance liquid chromatography and biochemical analysis) from 2005 to 2007. Six conclusions were drawn as follows:1. In our survey, moose and roe deer exhibited differential selection or avoidance behavior for all kinds of resources and human disturbances. Moose did not appear to avoid the occurrences of roe deer. The spatial distribution of moose revealed no avoidance of roe deer, but differential selection of resources and different responses to human disturbance. Models indicated that the inclusion of human disturbance as a factor in the modelling significantly diminished the assessment of habitat quality, and resulted in an assessment showing an increase in the degree of fragmentation of high quality and good quality habitat patches. Such disturbance was extremely significant for moose, as shown in a comparison of the quality layers of GIS maps with and without the inclusion of human disturbance factors. Most habitat assessed as high quality without the inclusion of any human disturbance factor was assessed as poor quality if human disturbance factors were included in the model; only a small patch of high quality remained. This suggested that human disturbance is a key factor affecting distribution patterns of moose and roe deer.2. Moose and roe deer exhibited differential preferences for plant species in the whole study area. Moose primarily preferred willows, while roe deer preferred poplars. They exhibited different characteristics of foraging behavior in zones with differing magnitudes of human disturbances. Moose and roe deer appeared to have different habitat requirements for foraging activity at microhabitat, patch and landscape scales, but they all were correlated with food abundances at microhabitat scale. Human disturbance influenced foraging habitat selection of moose and roe deer at the three scales as above. Individual calculations of GIS map layers of patches and landscapes indicated that both the whole area and mean patch area of high and good forage habitats were smaller than that of poor and low ones, suggesting a serious habitat loss and fragmentation. In addition, the loss of moose foraging habitat was more serious than that of roe deer as revealed by combined calculations.3. For the three different scales, microhabitat, patch and landscape, the moose exhibited different habitat requirements for bed-site selection, but this selection was negatively influenced by proximity to roads, positively influenced by the presence of settlements for forestry workers and researchers, andnegatively influenced by the harvesting interval of the forest patch. It also responded (avoiding or selecting) at similar distances for the three different habitat scales when they were confronted with the same type of human disturbance features. It did not appear to avoid settlement Significantly, in contrast, it showed a slight preference. The avoidance distance for roads occurred at about 3.5km. Habitat preference decreased with the extension of forest harvest intervals. Habitat requirement of roe deer was a little different from that of moose. Bed-site selection of this species was influenced by the proximity of settlements at all three scales, but was only influeced by forest harvest interval at patch scale. The response distance to a settlement was about 6 km at all three scales. In contrast to the moose, habitat preference of roe deer increased with the extension of forest harvest intervals.4. Fractal analysis of habitat utilization characteristics indicated that fractal value of moose movement path changed with the scales and revealed an apparent abrupt disjoint point, while the fractal value of roe deer showed straight line and self-similarity. These results suggested that moose demonstrated a preference for open habitat and similar fecal pellet excreting behavior at all three fractal scales, but only correlated with foraging in the proximity of food plants and use of mixed coniferous and deciduous forest at 6-15m, and correlated with foraging in the proximity of food plants and avoiding coniferous forest at 6-10m. Roe deer habitat utilization was mainly correlated with avoidance of denser arboreal forest, strategies increasing forage quantity and adopting suitable bedding behavior.5. Snow depth, the proportion of annual willow shoots and forest harvest intervals were significant variables to differentiate the microhabitat selection of both moose and roe deer. Forest harvest intervals contributed to their microhabitat separation the most, followed by snow depth and the proportion of annual willow shoots. Consequently, both deer species exhibited differences in adaptibility to harvested patches at different successional stages, tolerance to snow depth and preferences in food plant species selection. Hence, suitable forest harvest activities may promote the separation of their microhabitats, benefiting their coexistence.6. No dead or moribund individuals (UN:C≥20.0mg:mg) were found in these surveys, but a considerable proportion of moose were observed suffering from serious nutritional restriction (UN:C≥3.5mg:mg) in Jan. of 2006 and Jan. 2007 years in Zhanhe region, as well as roe deer in Mar. 2006 in the same region. In contrast, the nutritional condition of roe deer in Shengshan region was better than those in Zhanhe region from Jan. to Feb. in 2006 and 2007. Food competition between moose and roe deer might be a main factor impacting nutritional trends of moose population. At the time that harsh habitat occurred, moose exhibited endogenous protein catabolism. They reduced food overlap with roe deer by regulating foraging frequency for poplar and birch etc. and obtained energy and crude protein by foraging more quantity of faber fir, which has higher energy and protein contents; Snow depth and food abundance might be the main factors impacting the nutritional trends of roe deer population. Roe deer also exhibited endogenous protein catabolism, obtained energy and crude protein met their energy and crude protein requirements by foraging a higher proportion of poplar in their diets (poplar has higher crude protein levels) and a higher proportion of those few grasses which yield more digestible energy, while they were experiencing harsh habitat. Our comprehensive analyses of the relationship between nutritional trend and the temporal and spatial changes of food abundances, snow depths and diets suggested that the two species adopt different series of adaptable combinations of leg morphology, foraging behavior and nutritional metabolism etc. while they were experiencing harsh habitat conditions. In addition, trends of death rate changes of moose and roe deer populations might be indirectly reflected by refining the distribution of UN:C of individuals, which could allow us to analyze its correlation with habitat factors. Therefore, snow-urine analysis might be a valuable tool in practical management of deer populations under conditions of increasing habitat loss.