the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 25 Jun 2026
MacroTraits: a global trait data and information system for marine benthic ecology
Abstract. For a long time, biological trait data have been a bottleneck in biodiversity research. Constraints of data unavailability and challenging analytical implementation are still obstacles to the investigation of functional biodiversity patterns. This is especially true in marine zoobenthic ecology where the use of biological traits became common much later than in terrestrial and freshwater ecology. Additionally, most of trait-based marine studies have dominantly been conducted in European waters while large gaps remain in other areas of the world. Therefore, this paper offers a framework to fill this gap by providing the most comprehensive zoobenthic trait data compilation at the global scale. Based on more than 8000 references, 1893 species of the marine macrozoobenthos are documented for life history, dwelling mode, ecosystem function, habitat and biogeography through 41 traits. Next to this compilation, the paper brings clarifications on research directions by means of these data within the dominant paradigms of modern ecology. In particular, the dichotomous expressiveness that opposes response to effect traits (i.e., fitness components versus ecosystem function) is emphasised. The data base is accessible through an R package in the repository https://doi.org/10.5281/zenodo.20555888 and that facilitates data treatment such as trait selection, cross table construction and label handling.
- Preprint
(2156 KB) - Metadata XML
-
Supplement
(1331 KB) - BibTeX
- EndNote
Status: open (until 10 Aug 2026)
-
RC1: 'Comment on essd-2026-444', Anonymous Referee #1, 14 Jul 2026
reply
-
AC1: 'Reply on RC1', Olivier Beauchard, 14 Jul 2026
reply
Dear reviewer,
Many thanks for your time and feedback. Some of the comments certainly deserve some discussion useful to a wide audience.
(1) Manuscript length. Indeed, the manuscript reaches a certain length due to the detailed terminology. Since this is mainly a data paper, a simple summary table of traits and modalities as usually encountered would have been too much superficial. There are often several ways to structure a trait by considering different sets of modalities (e.g., when comparing papers). In each description that we provided, we wanted to show why the trait was built in the way it was, and not another one in order to comply either with biological reality or practical considerations. Also, as there is still no reference terminology, we strove to reference our documentation as best as possible based on the rich old literature of natural history. This leaves a traceability when justifying our developments (as it should be), and, possibly, inspiration for other purposes while keeping alive a valuable literature of several decades.
(2) Bremner’s works. There are at least two highly cited publications of Bremner and collaborators in the beginning of the 2000s (Bremner et al., 2003, 2006). They were not cited due to either lack of explicit reference to life-history strategy (as evolutionary convergences of adaptations) or effect trait concept as paradigmatically different. That does not detract in any way from the quality of those works; many other papers, with similar objectives, could have also been cited. There are now many reviews providing discussion and classification of all those works. However, we acknowledge that citing Bemner et al. (2003) better clarifies the historical context of the use of traits in benthic community ecology. Line 96, we propose:
“Multivariate exploration of life-history strategy as a result of environmental forces has been mostly investigated in freshwater and terrestrial ecology. In marine benthic ecology, since Bremner et al. (2003), only a handful of marine benthic studies specifically addressed the evolutionary concepts of selection and life strategy (Sutton et al., 2021; Beauchard et al., 2022; Gusha and McQuaid, 2025; Mendes et al., 2025; Bergagna et al., 2026).”
Bremner, J., Rogers, S. I., and Frid, C. L. J.: Assessing functional diversity in marine benthic ecosystems: a comparison of approaches, Mar. Ecol. Prog. Ser., 254, 11–25, https://doi.org/10.3354/meps254011, 2003.
Bremner, J., Rogers, S. I., and Frid, C. L. J.: Matching biological traits to environmental conditions in marine benthic ecosystems, Journal of Marine Systems, 60, 302–316, https://doi.org/10.1016/j.jmarsys.2006.02.004, 2006.
As raised and developed in discussion and conclusion, marine benthic community ecology, unfairly neglected research field, has missed an internationally coordinated agenda.
(3) Methodological aspects. Again, this is more a data paper than a review on analytical methods. However, we consider that it is difficult to talk about data as particular as fuzzy data without referring to methods. Separately, we tried to place as much and as relevant as possible analytical aspects in the supplement where the data can be handled and processed through typical multivariate ordinations.
Historically, a key event was the publication of Chevenet et al. (1994) in a special issue of Freshwater Biology as mentioned in Appendix. The authors proposed a data coding method, and accordingly, Fuzzy Correspondence Analysis (FCA) as a specific modification of Multiple Correspondence Analysis (MCA; Tenenhaus and Young, 1985) for multivariate purposes. From there, if the objective remains limited to a simple examination of trait covariances along the different axes (after a careful check of the eigenvalue diagram), there is no particular trait selection required. A priori, by assuming that the set of traits is homogeneously representative of fitness or ecosystem function, species axis scores can be clustered to derive a typology of functional groups; this is typically called “unconstrained analysis”. Our detailed terminology was also intended in this respect, accompanied with an assessment of affinity for response or effect. In the case of environment-trait correlative analysis, “constrained analyses” take place. Nowadays, after a long series of papers, the RLQ/Fourth-corner combination represents the most robust and unbiased way to proceed (Dolédec et al., 1996; Legendre et al., 1997; Dray and Legendre, 2008; ter Braak et al., 2012; Dray et al., 2014; Peres-Neto et al., 2027); additional papers can be found regarding the univariate context, especially niche modelling. Note that a derived version, the Double constrained correspondence analysis, can compete with RLQ (dc-CA; ter Braak et al., 2018). The origin of RLQ can be found in Co-Inertia Analysis (2 matrices; Dolédec and Chessel, 1994), later extended to 3 matrices (Dolédec et al., 1996). Dray et al. (2003) provided a review on the concept of co-inertia; compared to RDA, co-inertia can tolerate more explanatory variables, even collinear while RDA remains unstable in such a case. However, when correlating environmental variables with a set of traits, both R and Q ordinations should exhibit similar gradients in order to ensure a significant pattern. A given set of traits exhibit a certain multidimensionality, and adding or removing a trait, depending of its correlations with the other traits, can substantially alter the multidimensionality of table Q. Therefore, a selection procedure should take place regarding table Q as well as table R. To our knowledge, there is still no such procedure.
In summary. Technically, under constrained analysis through co-inertia (RLQ), there is no limit to the numbers of environmental or trait variables. In practice, and theoretically, selection (e.g., environmental filtering) should limit specific sets of correlated R and Q variables. We think that analyses should be processed with careful ordinations and different combinations of R and Q variables should be separately tested before considering results as definitive.
In our North Sea case study (Supplement), we do not use all the traits. Also, we simplified some of them for which the initial format lead to non-significant outcomes (while showing the usefulness of our R functions). So, we propose to add some text explaining the subtleties inherent to the procedure. Page 25, after the first paragraph of section S5:
“In a more general analytical context, this case study illustrates the complexity of RLQ applications (Dray et al., 2014). In its traditional use, RLQ relates environmental variables (table R) with biological traits (table Q) through species distributions (table L). Importantly, a significant R-Q pattern implies that R variables form gradients that match species distributions, themselves corresponding to trait modality distributions along a matching gradient (Dolédec et al., 1996). In this application example, not all combinations of R and Q variables may lead to a significant RLQ pattern. Therefore, the user is encouraged to run the procedure with different combinations in order to apprehend the need for deep data exploration before considering definitive RLQ outcomes.”
Chevenet, F., Dolédec, S., and Chessel D.: A fuzzy coding approach for the analysis of long-term ecological data, Freshw. Biol., 31, 295–309, https://doi.org/10.1111/j.1365-2427.1994.tb01742.x, 1994.
Dolédec, S., and Chessel, D.: Co-inertia analysis: an alternative method for studying species-environment relationships, Freshw. Biol., 31, 277–294, https://doi.org/10.1111/j.1365-2427.1994.tb01741.x, 1994.
Dolédec, S., Chessel, D., Ter Braak, C. J. F., and Champely, S.: Matching species traits to environmental variables: a new three-table ordination method, Environ. Ecol. Stat., 3, 143–166, https://doi.org/10.1007/BF02427859, 1996.
Dray, S., Choler ,P., Dolédec, S., Peres-Neto, P. R., Thuiller, W., Pavoine, S., and ter Braak, C. J. F.: Combining the fourth-corner and the RLQ methods for assessing trait responses to environmental variation, Ecology, 95, 14–21, https://doi.org/10.1890/13-0196.1, 2014.
Dray, S., and Legendre, P.: Testing the species traits-environment relationships: the fourth-corner problem revisited, Ecology, 89, 3400–3412, https://doi.org/10.1890/08-0349.1, 2008.
Legendre, P., Galzin, R., and Harmelin-Vivien, M.: Relating behaviour to habitat: solutions to the fourth-corner problem, Ecology, 78, 547–562, https://doi.org/10.1890/0012-9658(1997)078[0547:RBTHST]2.0.CO;2, 1997.
Peres-Neto, P. R., Dray, S., ter Braak, C. J. F.: Linking trait variation to the environment: critical issues with community-weighted mean correlation resolved by the fourth-corner approach, Ecography, 40, 806–816, https://doi.org/10.1111/ecog.02302, 2017.
Tenenhaus, M., and Young, F. W.: An analysis and synthesis of multiple correspondence analysis, optimal scaling, dual scaling, homogeneity analysis and other methods for quantifying categorical multivariate data, Psychometrika, 50, 91–119, https://doi.org/10.1007/BF02294151, 1985.
ter Braak, C. J. F., Cormont, A., and Dray, S.: Improved testing of species traits–environment relationships in the fourth-corner problem, Ecology, 93, 1525–1526, https://doi.org/10.1890/12-0126.1, 2012.
ter Braak, C. J. F., Šmilauer, P., and Dray, S.: Algorithms and biplots for double constrained correspondence analysis, Environ. Ecol. Stat., 25, 171–197, https://doi.org/10.1007/s10651-017-0395-x, 2018.
(4). Biogeography. “There are also challenges around biogeographical extent - modalities developed in temperate areas may not be useful in the topics for example.” What do you mean exactly?
Citation: https://doi.org/10.5194/essd-2026-444-AC1
-
AC1: 'Reply on RC1', Olivier Beauchard, 14 Jul 2026
reply
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 50 | 20 | 4 | 74 | 15 | 2 | 2 |
- HTML: 50
- PDF: 20
- XML: 4
- Total: 74
- Supplement: 15
- BibTeX: 2
- EndNote: 2
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
A macrobenthic traits database is long over due, so credit to the authors for delivering on this.
The manuscript is very comprehensive in terms of both philosophical considerations and analyses of the data in the data base. This could be reduced. I am surprised that Bremner's work from the 2000s, the first application of the traits and fuzzy coding approach developed by Chevenet et al is not cited at all. The authors also do not provide a discussion on the need to select traits to avoid issues weighting out comes given that some traits will be biologically linked. There are also challenges around biogeographical extent - modalities developed in temperate areas may not be useful in the topics for example. Again a cautionary note/discussion would be useful.