Results of single terms regressions showing the relationships of response variables to predictors, subsequently used as covariates in the minimum adequate models (MAM) comparing Czech Republic reserves and military training areas.
收藏资源简介:
AIC-Akaike information criterion; b – the darts indicate directions of the relationships, ↑standing for increasing, ↓ for decreasing, ↑↓ for polynomial trends with darts indicating the fitted functions directions. Upper-case P indicates a situation when polynomial performed better than linear trend, but still did not improve the model. Similarly, Upper-case L indicates a situation when log10-transformed predictor performed better than untransformed predictor, but still did not improve the model.The predictors were: Area, Latitude (Lat), Longitude (Long), Altitude (Alt), Altitude range (ARange) and Phytogeography region (Veg; a ranked variable, from coldest to warmest region of the country).
AIC(赤池信息准则,Akaike information criterion);图中箭头标注了变量间关联的方向:↑代表正向递增,↓代表负向递减,↑↓则代表带有拟合函数方向指示的多项式趋势。大写P表示多项式拟合效果优于线性趋势,但仍未对模型性能产生增益的场景;同理,大写L表示经log₁₀转换的预测变量表现优于未转换的预测变量,但仍未对模型性能产生增益的场景。本次研究涉及的预测变量包括:面积(Area)、纬度(Latitude,缩写Lat)、经度(Longitude,缩写Long)、海拔(Altitude,缩写Alt)、海拔范围(Altitude range,缩写ARange)以及植物地理区(Phytogeography region,缩写Veg,为按该国从最冷到最暖区域排序的等级变量)。




