Publication Date:
2014-06-25
Description:
Marine benthic ecosystems are difficult to monitor and assess, which is in contrast to modern ecosystem-based management requiring detailed information at all important ecological and anthropogenic impact levels. Ecosystem management needs to ensure a sustainable exploitation of marine resources as well as the protection of sensitive habitats, taking account of potential multiple-use conflicts and impacts over large spatial scales. The urgent need for large-scale spatial data on benthic species and communities resulted in an increasing application of distribution modelling (DM). The use of DM techniques enables to employ full spatial coverage data of environmental variables to predict benthic spatial distribution patterns. Especially, statistical DMs have opened new possibilities for ecosystem management applications, since they are straightforward and the outputs are easy to interpret and communicate. Mechanistic modelling techniques, targeting the fundamental niche of species, and Bayesian belief networks are the most promising to further improve DM performance in the marine realm. There are many actual and potential management applications ofDMsin the marine benthic environment, these are (i) earlywarning systems for species invasion and pest control, (ii) to assess distribution probabilities of species to be protected, (iii) uses in monitoring design and spatial management frameworks (e.g. MPA designations), and (iv) establishing long-term ecosystem management measures (accounting for future climate-driven changes in the ecosystem). It is important to acknowledge also the limitations associated with DM applications in a marine management context as well as considering new areas for futureDMdevelopments. The knowledge of explanatory variables, for example, setting the basis for DM, will continue to be further developed: this includes both the abiotic (natural and anthropogenic) and the more pressing biotic (e.g. species interactions) aspects of the ecosystem.
Repository Name:
EPIC Alfred Wegener Institut
Type:
Article
,
isiRev
Permalink
