Constructed wetlands host higher levels of avifauna species richness than their conventional counterparts (Figure 11). At sampling efforts under 60 minutes, differences are visible but minimal. However, for samples with greater than 60 minutes of effort (n=204), we can be 95% confident that constructed wetlands hosted at least 3.1 more species than the conventional ponds during an average migration month, or 21% more (Table 4). This was determined by a Welch two-sample t-test (t=5.67, df =150). It should be noted that in general, both pond type groups showed high variability, suggesting that additional classification based on vegetation or other characteristics may be necessary.

This analysis could be refined in the future to specifically determine if these additional species are species of conservation interest or concern, or are from specific taxonomic families. Given that the effect was highest at highest levels of effort suggests that the species that make up this difference are harder to detect or are more rare/less abundant, but further analysis of species composition would show if this is the case. Understanding which features these species need and find attractive in an urban setting could inform design specifications for future constructions.

Inclusion of Islands and Peninsulas in Stormponds Increases Avian Species Richness

Ponds with islands and peninsulas (Figure 12) host higher avian species richness during migration months as compared to featureless ponds (Figure 13). This difference is apparent at all levels of effort. Since effort seems to not be the main factor, pooling the samples for all effort strata shows that we can be 95% confident that ponds with features host 4.3 more species, or 33% more than featureless ponds as determined by a Welch two-sample t-test (t=9.111, df =189). 

Given this effect is similar at all levels of effort, my interpretation is that, unlike in the above result, these additional species are typically present and are common and detectable, which is why they are also found with minimal effort. This could indicate that the features provide necessary habitat for specific species that, while not rare, require these features to be present. Additional shoreline that is protected from disturbance from humans, wild and domestic canids, and other predators may be a factor. Future research on the effect of these features should include an analysis of species composition changes and targeted searches to understand which species are exclusively using the island or peninsulas, if any.

Ponds with Higher Perimeter/Area Ratios Have Less Species Richness

Ponds with higher Perimeter/Area Ratios (PAR) (Figure 14) hosted less species than those with lower (Figure 15). In general, the difference is neglible in the lower search effort class and increases with additional effort. However, high search effort comparisons are less reliable because the High PAR class had very few checklists over 90 minutes (n=11).  Comparing the differences for effort greater than 30 minutes, but less than 90 showed that we can be 95% confident that High PAR ponds have 13% less species richness as measured by a Welch two-sample t-test (t=5.381, df =174).