Making ecological indicators management ready: Assessing the specificity, sensitivity, and threshold response of ecological indicators

Caihong Fu (1) , Yi Xu (1) , Alida Bundy (2) , Arnaud Gruss (3) , Marta Coll (4) , Johanna J. Heymans (5, 6) , Elizabeth A. Fulton (7, 8) , Lynne Shannon (9) , Ghassen Halouani (10) , Laure Velez (11, 12) , Ekin Akoglu (13, 14) , Christopher P. Lynam (15) , Yunne-Jai Shin (12)

1 DFO - Fisheries and Oceans Canada
2 IOB - Institut océanographique Bedford - Bedford Oceanographic Institute [Dartmouth, Canada]
3 University of Washington - School of Aquatic and Fishery Sciences
4 ICM - Institute of Marine Sciences / Institut de Ciències del Mar [Barcelona]
5 SAMS - Scottish Association for Marine Science
6 European Marine Board
7 CSIRO Marine and Atmosphere Research [Hobart]
8 Institute for Marine and Antarctic Studies and Centre for Marine Socioecology
9 MA-RE - Marine Research Institute, University of Cape Town, Rondebosch, South Africa
10 UR03AGRO1 - Ecosystèmes et Ressources Aquatiques
11 UM - Université de Montpellier
12 UMR MARBEC - MARine Biodiversity Exploitation and Conservation
13 OGS - Istituto Nazionale di Oceanografia e di Geofisica Sperimentale
14 IMS - Institute of Marine Sciences [Erdemli-Mersin]
15 CEFAS - Centre for Environment, Fisheries and Aquaculture Science [Lowestoft]

Moving toward ecosystem-based fisheries management (EBFM) necessitates a suite of ecological indicators that are responsive to fishing pressure, capable of tracking changes in the state of marine ecosystems, and related to management objectives. In this study, we employed the gradient forest method to assess the performance of 14 key ecological indicators in terms of specificity, sensitivity and the detection of thresholds for EBFM across ten marine ecosystems using four modelling frameworks (Ecopath with Ecosim, OSMOSE, Atlantis, and a multispecies size-spectrum model). Across seven of the ten ecosystems, high specificity to fishing pressure was found for most of the 14 indicators. The indicators biomass to fisheries catch ratio (B/C), mean lifespan and trophic level of fish community were found to have wide utility for evaluating fishing impacts. The biomass indicators, which have been identified as Essential Ocean Variables by the Global Ocean Observing System (GOOS), had lower performance for evaluating fishing impacts, yet they were most sensitive to changes in primary productivity. The indicator B/C was most sensitive to low levels of fishing pressure with a generally consistent threshold response around 0.4*FMSY (fishing mortality rate at maximum sustainable yield) across nine of the ten ecosystems. Over 50% of the 14 indicators had threshold responses at, or below ∼0.6* FMSY for most ecosystems, indicating that these ecosystems would have already crossed a threshold for most indicators when fished at FMSY. This research provides useful insights on the performance of indicators, which contribute to facilitating the worldwide move toward EBFM.

Mots clés: Primary productivity; Environmental drivers; Indictor performance; Random forests; Regime-shifts; Shelf; Gradient forest method; Fishing pressure; Ecological modelling; Marine ecosystem; Framework; Sea; Southern benguela; Fisheries; Fishing impacts

Domaines: Biodiversité et Ecologie; Ecosystèmes

Caihong Fu, Yi Xu, Alida Bundy, Arnaud Gruss, Marta Coll, et al.. Making ecological indicators management ready: Assessing the specificity, sensitivity, and threshold response of ecological indicators. Ecological Indicators, 2019, 105, pp.16--28. ⟨10.1016/j.ecolind.2019.05.055⟩. ⟨hal-02434142⟩