DOES VEGETATION AFFECT THE HARVEST OF MUSKRATS ON THE UPPER
MISSISSIPPI RIVER NATIONAL WILDIFE AND FISH REFUGE?  

Laurie B. Wlosinski 1, and Joseph H. Wlosinski 2.   1 Upper
Mississippi River National Wildlife and Fish Refuge, 51 East
Fourth Street, Room 101, Winona, MN 55987; 2 U.S. Geological
Survey, Upper Midwest Environmental Sciences Center, 2630 Fanta
Reed Road, La Crosse, WI 54603.  

We used regression analysis to compare the number of muskrats
(Ondatra zibethicus) harvested on the Upper Mississippi River
National Wildlife and Fish Refuge (Refuge) with acres of
vegetation at various levels of resolution.  The Refuge
encompasses most of the wetland areas from Pools 4 through 14. 
Muskrat harvest data were available on a per pool basis for the
1990 and 1992 through 1997 trapping seasons.  Vegetation acreages
on the Refuge were obtained from the 1989 Geographic Information
System land cover/land use database developed by the Long Term
Resource Monitoring Program.  That database was developed from
1:15,000 aerial color infrared photographs and extensive ground-
truthing.  The first level of resolution for vegetation was 142
species or specie groups as classified, although many of these
groups were absent from some or all pools.  The next level of
resolution contained all classes of either Scirpus, Sagittaria,
or Typha, species commonly associated with muskrats.  The 142
species groups were then combined into classes: submergents,
submergent/rooted floating aquatics, submergent/rooted floating
emergents, rooted floating aquatics, rooted floating
aquatics/emergents, emergents, emergents/grasses forbs, and
grasses/forbs.  Various combinations of these groups were also
investigated.  Regression analyses were performed for muskrat
harvest using both the 1990 season or the average harvest for all
seven seasons against each of these groupings.  For the eight
vegetation classes we also performed a stepwise multiple
regression using acres of vegetation and the average size of a
stand as independent variables.  Significance for all tests was
set at the à = 0.05 level.    In general, regressions using the
seven-year average muskrat harvest produced a stronger
relationship than the 1990 harvest data.  The best relationships
at the species or specie groups level were for Sagittaria and
Scirpus.  Both relationships were significant and had R2 values
of 0.67 and 0.89, respectively.  The relationships for most of
the other species or specie groups were not significant.  Any
grouping containing Typha alone was not significant, but all
groupings with Sagittaria and/or Scirpus were.  The R2 values for
the latter groupings ranged between 0.76 and 0.83. 
Submergent/rooted floating aquatics, rooted floating
aquatics/emergents, and emergents showed significant
relationships and R2 values between 0.71 and 0.83.  We found that
the majority of the predictive capability from the stepwise
multiple regression models was due to the number of acres of
vegetation.  Average stand size was not a good predictor in our
models.  

Keywords:  muskrat, Ondatra zibethicus, Upper Mississippi River,
Upper Mississippi River National Wildlife and Fish Refuge,
vegetation