The Minnesota Grassland model

Citation:  Haroldson, K. J., R. O. Kimmel, M. R. Riggs, and A. H. Berner.  2006.  Association of ring-necked pheasant, gray partridge, and meadowlark abundance to Conservation Reserve Program grasslands.  Journal of Wildlife Management 70:1276-1284.

This model resulted from research examining the effects of grassland abundance on the abundance of several grassland birds.  Study areas averaged 23.5 square km (9.1 square miles), land cover was estimated by aerial photography and ground surveillance, and species abundance was based on visual counts along spring and summer roadside survey routes averaging 19.3 km (12 miles) through each study area.  Study areas ranged in grassland coverage from 2.2 to 32.6 percent, with CRP being the most common grassland type.  Only the summer (20 Jul – 20 Aug) pheasant abundance model is discussed below because most states that run roadside counts do so during August.  



In the original paper, the authors presented the following response surface to illustrate model predictions.  The dependent variable is the square root of the number of birds seen per 161 km (100 mi.).


If responses by year are averaged over the study period and pheasant abundance is normalized, the following relation is estimated.


If the above estimates are converted to percentage changes in grassland and pheasant abundance, the curve below results.  For example, if percent grassland in a study area increases from 0% to 5%, a population increase of 33% is predicted, if grassland further increases from 5% to 10%, an additional population increase of 28% is predicted, etc.