Structuring the unstructured: estimating species-specific absence from multi-species presence data to inform pseudo-absence selection in species distribution models

Abstract

1. Species distribution models (SDMs) are an increasingly popular tool in ecology which, together with a vast wealth of data from citizen science projects, have the potential to dramatically improve our understanding of species behaviour for applications such as conservation and wildlife management. However, many of the best performing models require information regarding survey effort, specifically absence, which is typically lacking in opportunistic datasets. To facilitate the use of such models, pseudo-absences from locations without recorded presence must be assumed. Several studies have suggested that survey effort, and hence likely absence, could be estimated from presence-only data by considering records across "target groups" of species defined according to taxonomy.2. We performed a probabilistic analysis, computing the conditional probability of recording a species given a particular set of species are also recorded, to test the validity of defining target groups by taxonomic order and to explore other potential groupings. Based on this quantification of recording associations we outline a new method to inform pseudo-absence selection comparing predictive performance, measured the area under curve (AUC) statistic, against the standard method of selection across a series of SDMs.3. Our findings show some support for target grouping classification based on taxonomy but indicate that an alternative classification using survey method may be more appropriate for informing effort and consequently absence. Across 49 terrestrial mammal species, pseudo-absence selection using our proposed method outperformed that of the standard method showing an improvement in the predictive performance of presenceabsence models for 17 out of 22 with sufficient data to elicit a significant difference. Based on our method we also observed a substantial improvement in the performance of presence-absence models compared to that of presence-only models (MaxEnt) with a higher AUC for all 22 species showing a significant difference between approaches.4. We conclude that our method produces sensible robust pseudo-absences which either compliment patterns in known presences or, where conflicts occur, are explainable in terms of ecological variables potentially improving our understanding of species behaviour. Furthermore, we suggest that presence-absence models using these pseudo-absences provide a viable alternative to MaxEnt when modelling using presence-only data.

Modeling current and potential distributions of mammal species using presence‐only data: A case study on British deer

“…Here, we considered records of other deer species together with common mammals easily identifiable by a similar method of visual observation alone (both direct and indirect, e.g., including evidence of species presence such as burrows, mounds and scat). Specifically, we considered records of fox ( Vulpes vulpes ), gray squirrel ( Sciurus carolinensis ), rabbit ( Oryctolagus cuniculus ), hare ( Lepus europaeus ), mole ( Talpa europaea ), rat ( Rattus norvegicus ), and cat ( Felis catus ); see Croft and Smith () for details. Using these records, we computed binomial datasets for each of the deer species describing the number of successes as individual visits (unique 1 km 2 BNG cell and date) where the target species was reported and the number of trials as visits where any of the species considered, including the target species, was reported.…”

Dealing with sampling bias and inferring absence data to improve distribution models of a widely distributed vulnerable marsupial

“…This is not surprising, because biases in presence‐only data are known to be exceptionally difficult to deal with (Støa et al., 2018; Yackulic et al., 2013). Approaches for inferring absences have been used elsewhere, often with good results (Bradter et al., 2018; Croft & Smith, 2019; Huang & Frimpong, 2015).…”

Enhancing Diversity Knowledge through Marine Citizen Science and Social Platforms: The Case of Hermodice carunculata (Annelida, Polychaeta)

“…In particular, it was noticed that people are more inclined to supply presence than absence data, leading to biased data [100] (i.e., not homogeneous data distribution in the studied area). Some authors suggest overcoming this problem by supplying species lists [101]; however, this solution is in conflict with the strategy based on a fast method. In this study, the issue was fixed by vetting the web ecological knowledge.…”