The role of vegetation structure in nest site selection by clay-colored sparrows (Spizella pallida)

Background/Question/Methods

Over the last 50 years, many North American grassland bird populations have declined significantly.  Essential to restoration is the discovery of factors that influence habitat selection.  This study investigates nest selection criteria for the clay-colored sparrow (Spizella pallida) in a 9ha field site consisting of 3 different restoration types: warm season grasses, cool season grasses and mixed forb plantings (>40 species of forbs) located in western Wisconsin.  These plots were laid out side-by-side, allowing birds to distribute themselves among the field types.  The objectives of this study were to compare the use of three habitat types for nesting by S. pallida and investigate the influence of vegetative characteristics on nest site selection.  During 2003 and 2005 we conducted bi-weekly nest searches by walking through each field until completely transversed, with nests located by observation or through intensive searches in areas where birds were flushed.  Nests were monitored until fledged or predated, after which we obtained vegetation measures (foliage height and cover, floristic diversity and lateral cover) at the nesting locations.  Using 5 apriori models we evaluated vegetation surrounding nesting locations of S. pallida compared to those of other species using adjusted delta AIC values and Akaike weights (w).

Results/Conclusions

In total, 44 S. pallida nests were located.  S. pallida exhibited a preference for mixed forb plots (84% of nest locations).  Nests were located between 18.3 (+10.7) and 23.0 (+10.6) cm above the ground in 2003 and 2005 respectively (df=30, p=0.18).  Nests occurred with microhabitat characteristics of 80.0 (+ 25.2) cm mean foliage height and average diversity (Shannon-Weiner Index) of 1.29 (+ 0.37).  Foliage cover within the first meter averaged 19.84 (+11.52) and 0.67 (+1.68) in the second meter.  Nest site selection results, comparing S. pallida with nest sites of other species, were best explained by the structural model (Adjusted AIC of 59.92, w=0.68, pseudo R²=0.36).  Nest site selection appears to be influenced not only by floristic diversity, but by structural predictors like foliage height and cover.  Foliage height and cover in the second meter had a positive impact on nest location (β=0.06, df=9, p=0.04; β=1.43, df=9, p=0.01), as did floristic diversity, although non-significant (β=2.33, df=9, p=0.07).  Areas with high lateral cover (greater than 1 meter above the nest) had a strong negative impact on nesting location (β=-23.23, df=9, p=0.01).  These results highlight the importance of considering structural diversity when making decisions regarding the composition and structure of reconstructed grasslands.