The population genetics of species occupying formerly glaciated regions are not only
impacted by glacial retreat but also agricultural land use that is typical of such regions. Areas
which have experienced glaciation often display a lowered amount of genetic variability and
minimal population structure, and these effects become more predominant with increasing
distance from a potential refugial population. Meanwhile, agricultural land use over the recent
past has also been demonstrated to disrupt population structure distribution through disturbance
regimes. The purpose of this study was to assess potential post-glacial and agricultural effects on
populations of creek chub (Semotilus atromaculatus) in two agricultural watersheds that differ in
the glacial history. The Saint Joseph River (SJR) watershed, Indiana and Michigan, USA was
entirely glaciated during the last glacial maxima, while the Little Miami River (LMR) watershed
in Ohio, USA, is situated on the boundary of the glacier. The degree of agricultural land use also
varies between and within the two watersheds. Using eight microsatellite loci, 312 individuals
were genotyped from 13 sites in SJR and 2,318 individuals from 29 sites in LMR. Measures of
genetic differentiation showed that there was strong differentiation between watersheds.
Analyses within watersheds recovered additional but weaker differentiation that was mostly
associated with the geography of sub-watersheds and isolation by distance. Proximity to the
glacial boundary appeared to play a minimal role in genetic differentiation and genetic variation.
Differentiation among localities was not directly associated with the glacial boundary within
LMR, and localities in this watershed had lower allelic richness and heterozygosity than those in
the fully glaciated SJR. After accounting for the positive correlation of stream distance in LMR
using partial Mantel test, both glacial history and agricultural land use were positively correlated
with genetic differentiation. However, these predictor variables were also strongly correlated
with one another which prevented disentangling the two potential effects. Within SJR, no
10
relationship of genetic differentiation with agricultural land use was recovered. My study shows
that there is not a simple relationship between glacial history, contemporary land use, and genetic
differentiation in creek chub. Rather, it appears that the patterns of genetic variation observed
may be more closely linked to the dispersal behavior of creek chub within and among
watersheds, and the history of effective population size within watersheds.