Padrões distintos de congruência climática em duas espécies invasoras de prosopis em zonas semi-áridas da América do Sul

Based on climate data and occurrence records, ecological niche models (ENM) are an important opportunity to identify areas at risk or vulnerable to biological invasion. These models are based on the assumption that there is a match between the climatic characteristic of native and invaded regions predicting the potential distribution of exotic species. Using new methods to measure niche overlap, we chose two exotic species fairly common in semi-arid regions of South America, Prosopis juliflora (Sw.) D.C. and Prosopis pallida (H. ; B. ex. Willd) HBK, to test the climate matching hypothesis. Our results indicate that both species occur with little niche overlap in the native region while the inverse pattern is observed in the invaded region on South America, where both species occur with high climatic overlap. Maybe some non-climate factor act limiting the spread of P. pallida on the native range. We believe that a founder effect can explain these similarities between species niche in the invaded region once the seeds planted in Brazil came from a small region on the Native range (Piura in Peru), where both species occur sympatric. Our hypothesis of a founder effect may be evident when we look at the differences between the predictions of the models built in the native and invaded ranges. Furthermore, our results indicate that P. juliflora shows high levels of climate matching between native and invaded ranges. However, conclusions about climate matching of P. pallida should be taken with caution. Our models based on climatic variables provide multiple locations suitable for occurrence of both species in regions where they still don t have occurrence records, including places of high interest for conservation.

Análise da variabilidade climática sobre a produtividade da soja no município de Bom Jesus (PI)

The climate is still main responsible for the variations soybean productivity (Glycine max (L.) Merrill), exerting a limiting action on these agricultural systems. The bomjesuense cerrado, this culture has proved, over the years, an increase of cultivated areas, however, productivity does not keep the same pace, going through periods of oscillations. Thus, although the crop is added to high technology, culture has great vulnerability to climatic adversities. Thus, the present study aims to analyze possible trends in meteorological variables, which can influence the soybean yield in Bom Jesus. For this purpose, different datasets were used, as follows: i) two periods of daily data (1984-2014 and 1974-2014), both obtained from the National Meteorological Institute (INMET); ii) climate normals from 1961-1990 as defined by INMET; iii) local agricultural production data of soybean-year (1997/1998 to 2012/2013) obtained from the Municipal Agricultural Production (PAM) dataset, which is management by Brazilian Institute of Geography and Statistics (IBGE). The analysis procedures included calculations of climate normals for 1984 to 2014 period and some statistical applications, as follows: i) the Wilcoxon test, used to evaluate differences between climate normals (1961 to 1990 and 1984 to 2014); ii) the Mann-Kendall nonparametric test, in order to analyze the linear trend of agrometeorological variables (rainfall, maximum temperature, minimum temperature and diurnal range of temperature; iii) cluster analysis by Ward method and the Spearman correlation test (rs) to identify the relationship between agrometeorological variable and soybean annual productivity. We adopted a statistical significance level of 5%. The results indicate changes in seasonality of the 1984-2014 climatology with respect to past climatology for all variables analyzed, except for insolation and precipitation. However, the monthly analysis of precipitation indicate negative trend during October and positive trend in December, causing a delay in start of rainy season. If this trend is persistent this result must be considered in futures definitions of the soybean crop sowing date over the region studied. With Mann-Kendall test was possible to identify positive trends with statistical significance in maximum temperature for all month forming part of soybean cycle (from November to April), which in turn tends to cause adverse effects on crop physiology, and consequently impacts on the final yield. Was identified a significant positive correlation between soybean yield and precipitation observed in March, thus precipitation deficit in this month is harmful to the soybean crop development. No statistically significant correlation was identified among maximum temperature, minimum temperature, and DTR with annual soybean productivity due these range of meteorological variables are not limiting factors in the final soybean yield in Bom Jesus (PI). It is expected that this study will contribute to propose planning strategies considering the role of climate variability on soybean crop final yield.