Green vegetation and fecal protein relationships in two southeastern Montana mule deer populations

A study was conducted on 2 areas in southeastern Montana in 1987-1988 to determine environmental factors important to survival of mule deer in that region, and the times of year when individual factors become critical.Weights of forbs and grasses clipped from 4 sites reached a peak in May or June, then declined throughout summer and fall.In 1987 weights of grass apparently began to increase again in mid-summer after declining from the first peak.Amounts of vegetation clipped were smaller in 1988 than in 1987, probably due to decreased rainfall in 1988.Percent nitrogen values for deer pellets collected from the 2 areas ranged from 1.31 on 11/25/87 to 3.78 on 5/21/88.Fecal N varied with grams of water in vegetation clipped.Fecal N was only slightly higher in summer 1987 than in summer 1988.Fecal N from the 2 study areas was similar except during 17 October-7 December 1987, when fecal N was significantly higher (p < 0.02) on the Ashland area than on the Boxelder area.Radio telemetry and observations of unmarked deer provided data on movements and use of cover types.Fourteen cover types were identified on the Ashland study area, 11 on the Boxelder area.Riparian areas were used heavily in summer but very little in winter.On the Ashland area the skunkbush type was important in summer and the alfalfa type was important in October.On the Boxelder area deer were more likely to be found near the creeks in summer and fall than in winter, and on cold winter days than on warm winter days.Deer on the Boxelder area used the same general areas yearlong.Some deer on the Ashland area migrated between distinct summer and winter ranges.Movement of many deer on the Boxelder area to the largest creek in late summer or early fall suggested that succulent forage was scarce at that time on other parts of the area.Deer appeared to use smaller areas in summer than in winter.Fawn:100 doe ratios showed a generally increasing" trend from July 1987 to February 1988 and again from June 1988 to September 1988.Reliability of these ratios is doubtful due to small sample sizes.Dates and causes of death of marked deer were recorded when they were known.A mule deer population model developed in New Mexico was tested for its usefulness in southeastern Montana.The model accurately predicted a peak in deer numbers in fall 1984, with a subsequent decline.Predicted birth rates appeared to be too low.A major problem with the model was inaccurate prediction of overwinter mortality.A population model for use in southeastern Montana should include some measure of winter severity.vii