Hello Nathalie, I've never done it, but I know that network analysis is used a lot for plant-animal interaction, and by assessing simple indicators of the network (conectance, modulation, etc.) you can infer about ecosystem health, i.e. how do those indicators change under disturbance and the ecological consequences of those network changes. I hope this was useful!

Vero

I have no idea what "ecosystem health" is. Exactly how would one measure it?

Sure : this is the central question ! In fact, we are faced for marine systems in Europe with the necessity to define what we consider as a Good Environmental Status of marine food webs under the implementation of the Marine Systems Framework Directive. What I wonder is wether we could use Ecological or Environ Network Analysis to define symplified proxies that could be monitored in the scope of this implementation.

The central issue of these questions is to define something that is both ecological meaningful and for which you are likely to have enough data to analyse. In practice, I find these nearly always comes down to monitoring species, or a set of species. For fish, you may also have some additional information on their diet.

I've no personal experience of Ecological or Environmental Network Analysis, but it seems that this approach relies on an "equilibrium" view of the system. In my opinion, ecosystem health can be defined by its resilience and to assess it you must know the dynamics of the system, the boundary conditions for stable states, and the role of adaptive cycles in its functioning. This is better achieved by a simple model than by a simple proxy, I guess...

Either for network or resilience analysis you will need a lot of information of the system. Network analysis are very complicated and highly dependent upon the sampling design and method. As Stuart said, I also believe that to make it simple it always comes down to species...although with diet data you may have an idea of the interactions held in the system. Functional diversity approach are also useful to asses ecosystem health (meaning maybe a comparison with a reference state??), where you analyze more than the species level only, and may get an idea of the processes of interest. Even more, it can help you assess the resilience of a state.

First of all, I have to say that I cannot give my expertise as I am a beginner on research (just starting my PhD). But as this issue is partly related with my study topic I’ll try to contribute to your question!

Your question mixes two aspects: a tool, and then an ecological concept.

I am not very familiarized with the statistic tool of Network Environ Analysis. But thanks to your question I’ve been reading about it and looks pretty interesting. This tool analyzes how transfers of any given reference unit (money, matter, energy) take place between different levels/elements of a system. In a stable system, we will have to consider a closed cycle of transfers, or otherwise, a compensation of fluxes (external inputs that compensate imperfect performance of the cycle).

On the other hand, a "healthy ecosystem" is conceptually, one able to supply ecosystem functions responsible for the maintenance of the own ecosystem properties (feedback mechanisms), and from the anthropogenic point of view, the ecosystem services that we require (food resources, buffering properties etc.).

Combining the tool and the concept, we should fix then what the "maximal capacity of the ecosystem is" in terms of biomass, energy, productivity... understand it's natural regime (as a stable succession of states in time), and understand the features responsible for it's resilience and resistance, to guaranty the stability of the ecosystem and the properties that define it.

A frequent proxy for ecosystem’s health is biodiversity assessment at different functional/trophic levels. The more species supply a same function, or constitute a trophic level, the more we think it is stable (also depends on the evenness of the function display by the implicated species). Furthermore, the more diverse the traits of species participating on a similar function are, the more the ecosystem is likely to be resistant and resilient to external perturbations.

As full biodiversity assessment is difficult or impossible for the most of the ecosystems, the concept of "surrogate taxa" emerges as groups of beings (a sensible taxon, or functional group) that are likely to represent accurately/ summarize several ecosystem properties. Also functional diversity (traits of species rather than species) has been revealed as a trustworthy studyline to understand resilience of ecosystems.

In order to build the matrices required for the Environ Network Analysis, the surrogates must be identified and characterized/quantified in relation to the transfer units (biomass / productivity / energy). As an example if surrogates are trophic levels, we can consider the biomass of each, and study how biomass transfers across levels takes place: 1ary consumers -- predators -- super predators -- decomposers. If we consider functional groups, we will identify the species at each functional group, and estimate/quantify their participation on the unit (again biomass/ nitrogen/ resource consumed etc), then evaluate how different groups perform on functions.

The interpretation of ecosystem health can be related to the evenness of the different elements integrated at each level of transfers, and how system fluxes vary when we replace one or more elements randomly and then specifically at different levels (simulating specific pertrubations affecting particular ecosystem settings) .

Hope I wasn't too long, and clear.

Some references:

- Balvanera, P. et al., 2006. Quantifying the evidence for biodiversity effects on ecosystem functioning and services. Ecology letters, 9(10), pp.1146-56.

- Cadotte, M.W., Carscadden, K. & Mirotchnick, N., 2011. Beyond species: functional diversity and the maintenance of ecological processes and services. Journal of Applied Ecology, 48(5), pp.1079-1087.

- Fath, B. D., & Patten, B. C. (1999). Review of the Foundations of Network Environ Analysis. Ecosystems, 2(2), 167–179.

- Grant, F. et al., 2008. Ecosystem Services and Drivers of Biodiversity Change. Report of the RUBICODE e-conference, (April).

- Sundstrom, S.M., Allen, C.R. & Barichievy, C., 2012. Species, Functional Groups, and Thresholds in Ecological Resilience. Conservation Biology, 26(2), pp.305-314

Bonjour Nathalie,

You might also get something useful from looking at the work published by Jeff Dambacher and colleagues while he was based at Oregon State - this will get you started with that literature:

http://www.ent.orst.edu/loop/pubs.aspx

I think Dr Dambacher is in Australia now (with CSIRO) working on marine ecosystems.

Good luck

Thank you very much !

Bonsoire Nathalie!

The only researcher I know of that has anything like a monitoring program of ENA health indicators is Joanna Patricio. See her book: "Ecological Indicators for Coastal and Estuarine Environmental Assessment - A User Guide". See also . Stuart, for some leads on the meaning of "ecosystem health", see ,, and . Good luck to all! Bob

Bob: great to see you here! It's been too long. Your answer what I would expect and, of course, mine was what you would expect from me. My concern is simple — one can have great ideas, or one can have abundant data, and only by considerable effort can one bring ideas and data together. So, yes, I've seen work on "health" — but worry, as I should, about whether there will be enough information to do what's necessary to make statements about it in practice.

Thank you all for your contributions to this discussion. What I wonder is also to what extend we may make the European Union evolve. For the moment, most of the discussion is focussed on top predators and fish based food web indicators. What I hope is to make them evolve toward more holistic descriptors. For exemple, by proposing to use stable isotopes, that may be less expensive that what is actually proposed.

Nathalie: I would encourage you to contact Joanna on this point . Stuart: Great to hear from you as well! You're right, of course. Matters are still tenuous, but that's why we need field work like Joanna's. Both: Ideas on system health have evolved somewhat since the references I gave yesterday. In particular, the notion of a "window of vitality" is gaining attention. See, for example, Figure 7 on p1890 of . Furthermore, this apparent balance between constraint and flexibility can be used for management. See, for example, the remediation of the eutrophic situation represented in Figure 14 on p52 of . The best to both of you! Bob