The shape of occupancy distributions in plant communities: the importance of artefactual effects

Abstract

Occupancy frequency distributions are commonly used as an approach to describe and analyse interspecific distribution patterns. However, the relative importance of biological versus artefactual mechanisms in shaping occupancy distributions is still largely undetermined. We evaluated the importance of different and interacting artefactual effects on the shape of occupancy distributions in local plant communities. The effects of sampling protocol parameters (i.e. size and number of sample units, sample extent, coverage, and intensity) on the shape of the occupancy distributions were examined separately. We identified the mechanisms that cause the effects by tracking the shifts of individual species between occupancy classes with varying parameters. Furthermore, the impact of different species abundance distributions and increasing levels of intraspecific aggregation on occupancy distributions was investigated by means of artificial assemblages. We show the following results: 1) increases in the number of sample units, sample extent, coverage, and intensity all result in a unimodal occupancy distribution with the mode in the lowest occupancy class; 2) an increase in sample unit size leads to a bimodal distribution; 3) changes that occur in the shape of the occupancy distributions with varying sampling protocol parameters can be explained by the movements of the species between occupancy classes; 4) different species abundance distributions may cause occupancy distributions with a left-hand mode or a bimodal distribution; and 5) the number of species in the highest occupancy class decreases with increasing degree of aggregation. The mode that almost always occurs in the lowest occupancy class is most likely due to the high number of rare species existing in most communities; the mode in the highest class emerges as a pure artefact that occurs when the sample unit size is relatively large compared to the sample extent. Consequently, the exclusion or separation of concurrent artefactual mechanisms is crucial when investigating the biological causes for the shape of occupancy distributions

The ranked species occupancy curves reflect the dominant process of species sorting: Evidence from forest/scrub communities

“…Several previous studies (e.g., Franzén & Eriksson, 2001;Leite & Lopes, 2001;Perelman et al, 2001) also revealed the overwhelming prevalence of satellite over core species, resulting in unimodal distributions at metacommunity level. Moreover, some authors have suggested that the core mode (bimodality) emerges as a consequence of the statistical properties of species occupancy probability distributions (Hui & McGeoch, 2007) or a pure artefact of the sampling protocol (Kammer & Vonlanten, 2009). With an increasing fraction of ruderal species and canopy openness, the slopes of RSOC's head became milder and eventually, in ruderal herbaceous communities, these curves are expected to form a 'shoulder' and approach a sigmoidal shape.…”

“…The estimation of the relative importance of each of these factors is further complicated by their confounding effects as well as by sampling artefacts. Indeed, the form of OFDs has been demonstrated to be scale-dependent (Collins & Glenn, 1997;van Rensburg et al, 2000;Kammer & Vonlanten, 2009). For instance, the inflation of the satellite mode at higher scale levels may be due to the low detection rates of rare species whose frequencies are underestimated (Nee et al, 1991).…”

“…For instance, the inflation of the satellite mode at higher scale levels may be due to the low detection rates of rare species whose frequencies are underestimated (Nee et al, 1991). On the other hand, the core species mode is more likely to occur when the plot size is larger, the sample number and coverage is lower, and the extension of regional area considered is smaller (McGeoch & Gaston, 2002;Kammer & Vonlanten, 2009). Hui and McGeoch (2007) went even further in suggesting that the bimodality in OFDs may simply represent a statistical outcome of the shape of the species occupancy probability distribution.…”

The spatial extent and the dispersal strategy of species shape the occupancy frequency distribution of stream insect assemblages

“…All these factors could modify the occupancy of the species through an increasing proportion of satellite or core species. Obviously, the observed OFD pattern reflects the interplay of these factors and is also influenced by sampling parameters (Heatherly et al., 2007 ; see Table 5 in Kammer & Vonlanthen, 2009 ). It follows that the individual contribution of each factor to OFD pattern is rather complicated to explore (Gafta et al., 2021 ).…”

Limited impacts of extensive human land use on dominance, specialization, and biotic homogenization in boreal plant communities

“…Jenkins [ 10 ] built on the intermediate disturbance hypothesis [ 18 ] and studies of species occupancy [ 19 - 22 ] to develop hypotheses for RSOC shapes, suggesting those shapes should vary along a successional gradient, and as a result, on a spatial gradient in disturbance. He suggested that with high disturbance, recruitment limitation should cause an exponential RSOC where some disturbance-adapted species are prevalent but most others rare.…”