Elsevier

Ecological Indicators

Volume 115, August 2020, 106393
Ecological Indicators

Ability of spatial indicators to detect geographic changes (shift, shrink and split) across biomass levels and sample sizes

https://doi.org/10.1016/j.ecolind.2020.106393Get rights and content

Highlights

  • The ability of 12 spatial indicators was tested to correctly identify four spatial changes: shifts, shrinks and split.

  • Indicators were tested using demersal fish species of the Bay of Biscay and Celtic Sea.

  • Inertia and isotropy detected all spatial changes in the data.

  • Center of gravity identified the shift, several indicators the shrink and none the split.

  • A minimum of 40 samples was required to detect the spatial changes.

Abstract

Spatial indicators are widely used to monitor species and are essential to management and conservation. In the present study, we tested the ability of 11 spatial indicators to quantify changes in species’ geographic patterns: (1) spatial displacement of a patch of biomass (‘shift’), (2) a spatial decrease in a patch, accompanied either by a loss of biomass (‘shrink0’) or (3) a relocation of the same biomass (‘shrink1’), and (4) splitting of a patch into smaller patches (‘split’). The geographic changes were simulated by manipulating the spatial distributions of the demersal species (observed during bottom trawl surveys). Hence, the spatial distributions of the latter being used as input data on which the manipulations were done. Additionally, other aspects of the indicators affecting the responses to the geographic changes were also tested, (1) homogeneous increase in biomass throughout the patch and (2) different sample sizes.

The center of gravity (defined by latitude and longitude) was the only indicator that accurately detected the ‘shift’ in biomass. The index of aggregation identified a decrease in the area and biomass of the main biomass patch (‘shrink0’), while the Gini index, equality area and spreading area were accurately identified a decrease in the area of the main biomass patch when total biomass did not decreased (‘Shrink1′). Inertia and isotropy responded to all geographic changes, except for those in biomass or distribution area. None of the indicators successfully identified ‘split’ process. Likewise, one of the indicators were sensitive to a homogeneous increase in biomass or the type of spatial distribution. Overall, all indicators behaved similarly well when sample sizes exceeded 40 stations randomly located in the area. The framework developed provides an accessible and simple approach that can be used to evaluate the ability of spatial indicators to identify geographic processes using empirical data and can be extended to other indicators or geographic processes. We discuss perspectives of the development of spatial indicators especially within the application of EU’s Marine Strategy Framework Directive.

Introduction

All life in Earth is both product and contributor to its place in space and time (David Attenborough, launch of ‘Our Planet’, 2019). Spatial indicators have been developed to represent and summarize species spatial patterns and their dynamics. They are often used in management (e.g. assess the state of species and ecosystems) (Bock et al., 2005, Greenstreet et al., 2012, Modica et al., 2016, Piet and Jennings, 2005, Rochet and Trenkel, 2009), and in ecology (e.g. to understand a species’ relationship with its environment, in face of habitat and climate change (Persohn et al., 2009, Yalcin and Leroux, 2017). Thus, the ability of indicators to identify an underlying geographic process accurately is crucial for their appropriate use in practical situations.

Selecting indicators from the large list of those available is not straightforward and usually only a few indicators can be used. Previous studies have attempted to identify a small set of indicators that identified most of the spatial patterns observed and have better statistical properties (e.g. robust to outliers and changes in the distribution, regardless of abundance, Bock et al., 2005). Further, most indicators’ results are often highly statistically correlated with each other, and thus may be redundant. For example, Rufino et al. (2018) suggested grouping indicators into three categories that, reflect the main ecological patterns of species spatial distribution: occupancy, aggregation and quantity. Doing so would reduce the number of indicators to only three, each representing one category.

Another important aspect of indicators that can influence their selection is their ability to identify spatial or geographic change. To our knowledge, one aspect of sampling design that has been poorly addressed when using empirical data is the number of samples required to identify a change in species distribution. However, Rindorf et al. (2012) analyzed the properties of several indicators analytically and by simulating abundance-occupancy relationships, in response to changes in species distribution and sample size.

The aim of the current study was to determine the ability of several spatial indicators to identify changes in geographic patterns of demersal species, when the species main biomass patch moves (shift), when a larger patch splits into smaller ones (split) or when the area of highest biomass decreases (shrink, with a decrease in or relocation of biomass). The indicators were also assessed at higher levels of biomass (two and five times as high) and multiple sample sizes (20–160 stations). Conclusions are then drawn about management applications, especially in the European Union’s Marine Strategy Framework Directive (2008/56/EC) (MSFD), which requires that species and ecosystems be monitored using indicators that are operational and have clearly defined targets.

Section snippets

Data used

The data analyzed came from a bottom trawl fishery survey (EVHOE)(Evaluation Halieutique de l’Ouest Européen, EVHOE cruise, RV Thalassa, IFREMER, Leaute and Pawlowski, 2015) that was performed in Autumn 2015 in the Bay of Biscay and the Celtic Seas. The survey covered a bathymetric range 20 up to 700 m deep and consisted of 148 randomly stratified sampling stations (Figures:). The distribution of the biomass of 29 demersal species (Supplementary material 1) was interpolated onto a grid with

Results

The latitude of the center of gravity accurately identified the southward shift of the species biomass, although its relative change was smaller than that of the corresponding number of grid rows shifted (i.e. a 20% shift in latitude, for a 30% shift in rows, i.e. 7 out of 23 in the target area). No change was detected for the shrink or split processes (Fig. 1; Table 1). The longitude of the center of gravity did not change for the shift and split processes, but did changed slightly (by 0.02)

Discussion

A good suitable indicator should be calculated by a simple direct equation and clearly interpret the underlying process (Baddeley et al., 2015). Although previous studies recommend including multiple indicators (Petitgas and Poulard, 2009, Woillez et al., 2007b), monitoring programs, such as the MSFD often require parsimonious and non-redundant indicators. It is thus necessary to select few indicators, if possible, using objective criteria. We tested the ability of several indicators to detect

CRediT authorship contribution statement

Marta M. Rufino: Conceptualization, Formal analysis, Investigation, Methodology, Writing - original draft, Writing - review & editing. Nicolas Bez: Methodology, Validation, Writing - review & editing. Anik Brind’Amour: Funding acquisition, Writing - review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The authors acknowledge all participants in EVHOE surveys, as well as the developers of the indicators. Marta M. Rufino is funded by a research contract awarded by IFREMER within the MSFD framework, and also by the Loire-Brittany and Adour Garonne French Water Agencies. The authors would also like to acknowledge the contributions of two anonymous referees.

References (29)

  • Leaute Jean-Pierre, Pawlowski Lionel, S.M., 2015. EVHOE 2015 cruise, RV Thalassa....
  • C. Persohn et al.

    Habitat preferences of selected demersal fish species in the Bay of Biscay and Celtic Sea, North-East Atlantic

    Fish. Oceanogr.

    (2009)
  • P. Petitgas

    Biomass-dependent dynamics of fish spatial distributions characterized by geostatistical aggregation curves

    ICES J. Mar. Sci.

    (1998)
  • P. Petitgas

    Sole egg distribution in space and time characterised by a geostatistical model and its estimation variance

    ICES J. Mar. Sci.

    (1997)
  • Cited by (1)

    1

    Presente address: MARE ULisboa - Marine and Environmental Sciences Centre, FCUL - Faculdade de Ciências da Universidade de Lisboa, Campo Grande, 1749-016 Lisboa, Portugal.

    View full text