Development and evaluation of pattern recognition habitat models for the ruffed grouse, gray squirrel, and fox squirrel in Missouri
Abstract
Pattern recognition (PATREC) models are single-species models that use habitat characteristics to arrive at an estimation of population abundance. Currently, personnel of the Mark Twain National Forest in Missouri are using this type of model as a basis for making management decisions on 13 management emphasis species. Before the outputs of these models can be used with confidence, however, relational functions that support the models must be explored and quantified by research. The primary purpose of my study was to develop and evaluate PATREC models for the ruffed grouse (Bonasa umbellus), gray squirrel (Sciurus carolinensis), and fox squire! (Sciurus niger) in Missouri. I measured vegetative structure and population abundance of ruffed grouse and both squirrel species on 13 and 14 study areas, respectively, to identify habitat features associated with high and low densities of each species~ Six habitat parameters were determined to be related to ruffed grouse density. Areas with high spring densities(> 13.0/405 ha) had significantly (P < 0.01) greater amounts of disturbed forest habitats and lower amounts of upland sawtimber than areas with low densities. structural characteristics influencing the quality of disturbed cover included basal area, woody stem density, and canopy closure. Woody stem density was also an important habitat condition of the upland sawtimber cover type. The PATREC model developed with these 6 parameters correctly classified 100% of the study areas into either high o~ low density classes. Based upon model outputs, areas with low densities were deficient in either the percent occurrence or structural characteristics of the disturbed cover type indicating that management for ruffed grouse in Missouri should be directed toward providing more of this habitat type. Population estimates of squirrels concurred with 2 years of poor acorn mast production. During and immediately following poor mast years, it appears that both fox squirrels and gray squirrels use areas where timber is interspersed with crops and old fields. These areas may provide a source of supplemental food. In contrast, both species were negatively correlated with the occurrence of the hardwood sawtimber cover type. However, the horizontal density of vegetation from o.o to 1.5 min height and overstory canopy closure of shagbark hickory (Carya ovata) within the hardwood sawtimber cover type exhibited positive relationships with the abundance of both species. During periods of poor mast, both species also appear to be positively influenced by the combined effects of increased dominance of hickory and soft mast producing tree and shrub species. Management programs that facilitate the maintenance of openings and crop fields, an abundance of ground cover, and a variety of dominant and subdominant trees should not only improve habitats for year-round use by squirrels but also during periods of low mast availability.
Degree
M.S.
Rights
OpenAccess.
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 License.