Strategies to optimize monitoring schemes of recreational waters from Salta, Argentina: a multivariate approach.

Several recreational surface waters in Salta, Argentina, were selected to assess their quality. Seventy percent of the measurements exceeded at least one of the limits established by international legislation becoming unsuitable for their use. To interpret results of complex data, multivariate techniques were applied. Arenales River, due to the variability observed in the data, was divided in two: upstream and downstream representing low and high pollution sites, respectively, and cluster analysis supported that differentiation. Arenales River downstream and Campo Alegre Reservoir were the most different environments, and Vaqueros and La Caldera rivers were the most similar. Canonical correlation analysis allowed exploration of correlations between physicochemical and microbiological variables except in both parts of Arenales River, and principal component analysis allowed finding relationships among the nine measured variables in all aquatic environments. Variable's loadings showed that Arenales River downstream was impacted by industrial and domestic activities, Arenales River upstream was affected by agricultural activities, Campo Alegre Reservoir was disturbed by anthropogenic and ecological effects, and La Caldera and Vaqueros rivers were influenced by recreational activities. Discriminant analysis allowed identification of subgroup of variables responsible for seasonal and spatial variations. Enterococcus, dissolved oxygen, conductivity, E. coli, pH, and fecal coliforms are sufficient to spatially describe the quality of the aquatic environments. Regarding seasonal variations, dissolved oxygen, conductivity, fecal coliforms, and pH can be used to describe water quality during dry season, while dissolved oxygen, conductivity, total coliforms, E. coli, and Enterococcus during wet season. Thus, the use of multivariate techniques allowed optimizing monitoring tasks and minimizing costs involved.

Range-wide genetic structure and diversity of the endemic tree line species Polylepis australis (Rosaceae) in Argentina.

PREMISE OF THE STUDY: Knowledge on the range-wide distribution of genetic structure and diversity is required to facilitate the understanding of historical tree migration and for predicting responses to current climate change. With respect to post-glacial migration patterns known from the northern hemisphere, we tested the prediction that the southernmost populations of a subtropical tree line species have lower within-population genetic diversity and higher genetic differentiation than the central and northernmost populations. METHODS: We used AFLP to assess the genetic structure of 18 populations of the wind-pollinated Polylepis australis (Rosaceae) sampled over its entire distributional range in three Argentinean high mountain regions. Genetic diversity was calculated as a percentage of polymorphic bands (P) and Nei's expected heterozygosity (He); genetic differentiation was assessed using AMOVA, Φ(ST)--statistics, and Bayesian cluster analysis. KEY RESULTS: Contrary to our expectations, the northernmost Polylepis australis stands had lower within-population genetic diversity and higher genetic differentiation than the central and southernmost stands. Populations grouped into two major clusters, the first containing the southern populations and four central populations and the second containing the northern and one central population. CONCLUSIONS: Patterns of Polylepis australis genetic structure and diversity differ from historical migration scenarios observed for the northern hemisphere. The decline in genetic diversity toward the north may point to an equatorward migration following past climatic changes. Populations within the south and central part appear to be connected by effective long-distance pollination while gene flow in the northern part is probably hampered by geographic isolation.