Identifying the factors that structure host–parasite interactions is fundamental to understand the drivers of species distributions and to predict novel cross-species transmission events. More phylogenetically related host species tend to have more similar parasite associations, but parasite specificity may vary as a function of transmission mode, parasite taxonomy or life history. Accordingly, analyses that attempt to infer host−parasite associations using combined data on different parasite groups may perform quite differently relative to analyses on each parasite subset. In essence, are more data always better when predicting host−parasite associations, or does parasite taxonomic resolution matter? Here, we explore how taxonomic resolution affects predictive models of host−parasite associations using the London Natural History Museum's database of host–helminth interactions. Using boosted regression trees, we demonstrate that taxon-specific models (i.e. of Acanthocephalans, Nematodes and Platyhelminthes) consistently outperform full models in predicting mammal-helminth associations. At finer spatial resolutions, full and taxon-specific model performance does not vary, suggesting tradeoffs between phylogenetic and spatial scales of analysis. Although all models identify similar host and parasite covariates as important to such patterns, our results emphasize the importance of phylogenetic scale in the study of host–parasite interactions and suggest that using taxonomic subsets of data may improve predictions of parasite distributions and cross-species transmission. Predictive models of host–pathogen interactions should thus attempt to encompass the spatial resolution and phylogenetic scale desired for inference and prediction and potentially use model averaging or ensemble models to combine predictions from separately trained models.

To assess the importance of climate variables for the distribution of modern pollen data in China, we present a continental-scale dataset consisting of 1374 samples. Boosted regression trees and constrained ordination techniques are employed to quantify the importance of six climate variables (annual precipitation, PANN; actual/potential evapotranspiration ratio, Alpha; mean annual temperature, TANN; mean temperature of the warmest month, MTWA; mean temperature of the coldest month, MTCO; annual sum of the growing degree days above 5°C, GDD5) for the distribution of individual pollen taxa and modern pollen assemblages. The results show that taxon-specific responses to the climate variables display a wide regional diversity and that the climate variables with low collinearity that best account for the spatial variability of modern pollen assemblages differ regionally. PANN is the most important variable in northwestern and northeastern China and the Tibetan Plateau, while MTWA and MTCO are the dominant variables in east-central and southern China. This suggests that the climate variables that can be optimally reconstructed from fossil pollen data vary in different bioclimatic regions of China. This feature is typical to many continental-scale modern pollen datasets and needs to be considered in pollen-based climate reconstructions.

Prediction of invasive species distributions from survey data is widely recognized as a significant component of forest management and conservation planning. Leucaena leucocephala is the most aggressive invasive shrub and tree in the Hengchun peninsula in southern Taiwan. We analyzed geo-referenced data to identify potential variables of invasion and to predict likelihood of further invasion using boosted regression trees. Our results, which classified 92% of the cells correctly with regard to species presence and absence, indicated probability of invasion is correlated with climatic conditions (temperature and precipitation), landscape features (altitude; slope ratio and aspect; percentages of natural or secondary forest, agriculture land, developed area, and water bodies; and distances to the nearest forest edge and river), and anthropogenic factors (length of forest edge, and distances to the nearest road and agriculture land). The most influential variables are average annual temperature, altitude, precipitation, and slope. Continued range expansion by L. leucocephala is most likely to proceed (1) from the eastern and western portions toward the central portion of Hengchun township and (2) throughout the southern and toward the eastern portions of Manjhou township. Our model should provide useful information to aid forest managers in the development of long term monitoring and control strategies for L. leucocephala, in the early detection and eradication of newly established invasions, and also a framework for the integration and analysis of new presence and absence field data as they become available.