We investigated how landscape structure links to water quality in agricultural watersheds draining into Lake Erie and if such structure may serve as an effective water quality indicator. Specifically, we used correlation and regression analyses to assess if water quality correlates with metrics of riparian forests landscape structure and with proportions of land cover types in non-riparian areas. We also quantified the relative contribution of these spatial parameters to water-quality benefits within selected watersheds. We used historical data to compute yields for suspended sediments (SS), total phosphorus (TP), nitrate+nitrite (NO23), and total dissolved solids (TDS, estimated from conductivity) and computed landscape metrics in GIS using land cover and other spatial data. In the riparian corridor, there were no significant (p≤0.05) correlations between water quality parameter yields and landscape metrics computed for forests cover or combined natural land cover types. Within non-riparian portions of watersheds, %row crops was positively correlated with SS (p=0.01) and TP (p=0.05) and showed weak but not significant correlations with NO23 (p=0.07). Conversely, SS was negatively correlated with cover of forests, wetlands, woody vegetation, and all natural vegetation combined. Except for wetlands cover, this was also the case for TP, while all NO23 correlations were non-significant. TDS showed weak negative correlations with the amount of pasture/hay and wetlands and positive correlations with developed cover. According to multiple regression results, proportions of land cover in non-riparian areas and other landscape factors within entire watersheds were generally better indicators of water quality, while the %row crops and %pasture/hay were the only riparian metrics that contributed to explaining water quality variability. These results suggest that the reduction of pollutants in streams and rivers depends primarily on improving land management practices in non-riparian portions of watersheds. On the other hand, results also suggest that an increase in vegetation cover (natural or planted herbaceous buffer strips) in the riparian areas may contribute to further mitigate pollutant loading into streams. Although not significantly correlated, pollutant yields generally showed a negative trend with the amount of natural vegetation cover types in the riparian corridor. Furthermore, model results indicated that pasture/hay cover in riparian areas reduces NO23 and TDS loading.

In spite of its size and biological significance, we know little about the ecology of the Pantanal, a 140,000 km2 floodplain in west-central Brazil. Increasing human pressures make this lack of understanding particularly critical. Using transects and 1 m2 circular plots, we documented floristic composition and interacting-environmental conditions associated with littoral herbaceous vegetation along inundation gradients at two ecologically-distinct sites in the Pantanal. We recorded water depth and percent cover for each species in Baía Piuval, a bay in the Bento Gomes River (Mato Grosso), and in a bay in the Acurizal Reserve (Mato Grosso do Sul). Baía Piuval and Acurizal plots contained a total of 22 and 18 macrophyte species, respectively. At both sites Eichhornia azurea and Salvinia auriculata occurred most frequently as dominant or co-dominant species. Chi2 analysis, used to quantify zonations along depth gradients, generated four different groups of species ( p < 0.05) for Baía Piuval. For Acurizal, two significantly different groups ( p < 0.05) occurred with an intermediate assemblage of species that could be assigned to either group. Canonical correspondence analysis, used to analyze species distributions, showed a pattern consistent with the Chi2 results for Baía Piuval but not for Acurizal. Higher species richness and diversity occurred where dry season and low water levels coincided and richness was generally highest in proximal plots where water depths were lowest. Our results are consistent with the few other plant ecological studies reported for the Pantanal. This study can be considered additive to needed baseline data on biota and ecology of this region of South America.

Recommended Citation Gottgens, Johan F.; Perry, James E.; and al, et, "The Paraguay-Paraná Hidrovía: Protecting the Pantanal with Lessons from the Past: Large-scale channelization of the northern Paraguay-Paraná seems to be on hold, but an ongoing multitude of smaller-scale activities may turn the Pantanal into the next example of the “tyranny of small decisions”" (2001). VIMS Articles. 1475. https://scholarworks.wm.edu/vimsarticles/1475