Premium
Integral occurrence probability: combining cover and relative shoot frequencies based on bounded point‐to‐plant distances
Author(s) -
Van Calster Hans,
Damgaard Christian
Publication year - 2017
Publication title -
journal of vegetation science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.1
H-Index - 115
eISSN - 1654-1103
pISSN - 1100-9233
DOI - 10.1111/jvs.12527
Subject(s) - mathematics , sampling (signal processing) , relative species abundance , measure (data warehouse) , species richness , statistics , ecology , abundance (ecology) , computer science , biology , filter (signal processing) , database , computer vision
Aims To introduce a new distance‐based field method for (herbaceous, terrestrial) plant species that relates cover to relative shoot frequency as a continuous process of occurrence probabilities and to explain how these data can be analysed. Methods We propose to measure shortest distances from a sample of sampling points to the nearest above‐ground part of plant species (up to a maximum search distance). We show how, after appropriate transformation of the point‐to‐plant distances to a 0–1 interval scale, cover as well as relative shoot frequency at any area up to the searched area can be read from the same curve. This leads to the notion of an integral occurrence probability , which we propose as a new species abundance measure. For estimation and regression modelling we make use of the zero‐and‐one inflated beta distribution. We supply all code required for these analyses. Results Simulations of plant distribution patterns showed that the integral occurrence probability is able to differentiate between plant abundance patterns that differed in terms of relative cover, density and type of spatial distribution pattern. It is more sensitive to these differences than either cover or relative shoot frequency alone. The method allows summing occurrence probabilities over species to predict expected species richness as a function of the area searched. Aggregation across species, while accounting for overlap in species spatial distributions, is a simple matter of taking the minimum among the point‐to‐plant distances at each sampling point. The latter was demonstrated with data from a field trial in Nardetea grassland. Conclusions The method could be a viable alternative for currently employed field methods, such as visual cover estimates, point‐intercept sampling and recording the frequency of plant species in equal‐area plots. Applications include, but are not limited to, conservation management monitoring and ground‐truthing of remote sensing data.