Till startsida
To content Read more about how we use cookies on gu.se

Taxonomic resolution

Reliable taxonomic identifictation of biological data is fundamental for modelling and mapping biodiversity. Therefore, the methods selected for sampling must be suitable for the level of taxonomic resolution that is relevant to the aim of the mapping study.

When samples are collected and brought back to the laboratory, for example using corers or nets to collect invertebrates or fish, limits to taxonomic resolution is set by the taxonomic skills of the available personnel. In situations where data on vegetation or epibenthic fauna are collected visually in the field by diving, snorkelling or with underwater photography (video or still photo), the appropriate taxonomic resolution differs among methods.

Detectability affects data quality
Some species are more conspicous than others; larger species are easier to see and can also hide other, smaller species. Some species are easier to identify whilst others require closer inspection. The method of data collection; remote sampling, scuba diving or video observation contributes to detectability. For example, many species of vegetation and invertebrates in the Baltic Sea are small and live under large algae, and can not be sampled by observations from video records. This necessitates close inspection by divers or for the collection of samples. On the other hand, sampling methods capable of high taxonomic resolution is often much more expensive than simpler photographic methods, therefore the performance of models based on such data is often limited by the sample size.

PREHAB recommendation
The research within PREHAB has shown that broad taxonomic categories (”total vegetation", ”red algae” etc.) can be modelled. Therefore, inability to distiguish among all species is not a fundamental problem for modelling. The problem is rather that uncertain quality in taxonomic identification leads to poor models and maps because they introduce errors associated with false absences and presences, as well as biased estimates of abundance. The important recommendation with regards to sampling method is that the taxonomic resolution suitable for a particular method should be carefully considered and defined a priori. This will ensure an even quality of data! If this is not possible, for example because models are based on available data, modelling methods based on presence only (e.g. Maxent) might be preferred.

© University of Gothenburg, Sweden Box 100, S-405 30 Gothenburg
Phone +46 31-786 0000, About the website

| Map

The University of Gothenburg uses cookies to provide you with the best possible user experience. By continuing on this website, you approve of our use of cookies.  What are cookies?