Local distribution patterns of fleshy‐fruited woody plants – testing the orchard hypothesis

Abstract

Plant distribution patterns are influenced by many different factors. We examined mechanisms behind local distribution patterns of boreo‐nemoral fleshy‐fruited woody plants with seed dispersal mainly mediated by birds. It has been suggested that guilds of these plants develop ‘orchards', i.e. locally aggregated occurrences composed of several species. We analysed spatially explicit occurrence data of different life stages of a local guild of fleshy‐fruited woody plants in south‐eastern Sweden, and conducted a seedling recruitment experiment for a subset of ten species. Spatial point pattern analyses showed that the guild of fleshy‐fruited species was aggregated at small (< 10 m) spatial scales. Saplings were more common under canopies of heterospecific reproductive individuals than expected by chance. These results show that the local guild of fleshy‐fruited species is distributed as orchards, i.e. clusters consisting of individuals of different species and life stages. We found no evidence of negative distance dependence between saplings and reproductive conspecific individuals. Results from the recruitment experiment suggest that recruitment is seed limited and generally low among the studied species. At the site‐scale (circular areas with 50 m radius), there was no difference in seedling recruitment between sites with and without reproductive conspecific individuals for most species included in the recruitment experiment. This further suggests that the aggregated patterns found are not simply a result of spatial concordance in suitable habitats across life stages. Instead, we suggest that the sheer number of seeds from species in the guild deposited under the crowns of fruit bearing individuals is the main mechanism behind the build‐up of orchards. Although further studies are needed to fully disentangle the processes underlying the observed patterns of local diversity, we argue that describing patterns and contrasting them to the predictions of ecologically relevant hypotheses is a useful first step.