Impacts of environmental pollution on mangrove phenology: Combining remotely sensed data and generalized additive models.

Mangrove ecosystems worldwide have been affected by anthropogenic activities that modify natural conditions and supply trace elements that affect mangrove health and development. In order to gain a better understanding of these ecosystems, and assess the influence of physicochemical (granulometry, pH, salinity and ORP) and geochemical variables (concentrations of V, Cr, Co, Ni, Cu, Zn, Pb, Rb, Sr and Zr) on mangrove phenology, we combined field and satellite derived remotely sensed data. Phenology metrics in combination with Generalized Additive Models showed that start of the season was strongly influenced by Pb and Cu pollution as well as salinity and pH, with a large percentage of deviance explained (92.10%) by the model. Start of season exhibited non-linear delays as a response to pollution. Other phenology parameters such as the length of season, timing of the peak of season, and growth peak also indicated responses to both trace elements and physicochemical and geochemical variables, with percentages of deviance explained by the models ranging between 33.90% and 97.70%. While the peak of season showed delays as a response to increased pH and decreased salinity, growth peak exhibited a non-linear decrease as a response to increased Sr concentrations. These results suggest that trace element pollution is likely to lead to altered phenological patterns in mangroves.

Country-scale mapping of ecosystem services provided by semi-natural grasslands.

Semi-natural grasslands harbour high biodiversity and play a key role in the supply of ecosystem services (ES). However, abandonment, changes in traditional management practices and agricultural intensification constitute a major threat to these grasslands worldwide and these practices have led to declines in species diversity. In this paper the multi-functionality of semi natural-grasslands is assessed from the ES perspective, within a range of common semi-natural grassland types throughout Estonia. The analysis follows a stepwise approach based on the ES cascade model. Firstly, analyses of the relationships between plant species distribution patterns and environmental factors are described. Secondly, the effect of grassland abandonment on plant species diversity, as well as on the presence of rare and protected plant species is tested. In order to overcome the lack of data on ES at the national scale, plant species diversity and soil organic carbon are tested as surrogate indicators for five ESS: pollination, herbs for traditional medicinal use, nutrient cycling, nutrient retention and biomass production. In the final step, the spatial distribution of ES is assessed, based on an ES hotspots map obtained by detecting areas where high levels of plant species diversity and soil organic carbon overlap. The results show that the majority of ES hotspots are present in wooded meadows and pastures. However, there is an important threat to these hotspots because 45% are not eligible for agri-environmental support.