m@rble ELectronic conference on MARine Biodiversity in Europe

summaries theme 1

Summaries of theme 1, as they are posted here, are compiled by the session's chairman, Carlos Duarte.

Main issues in marine biodiversity research

What follows is an overview summary of the discussions of session 1, prepared by the session's chairman, Carlos Duarte. The individual messages, as posted by the participants, will remain available for some time.

European scientists are more and more joining forces in order to study biodiversity patterns in Europe, by sharing logistics and exchanging expertise. The BIOMARE concerted action, based on the Implementation Plan of Marine Biodiversity Research in Europe of the Marine Board, will eventually propose a mechanism to achieve at least the large scale and hopefully the long-term research as well. Its implementation will require a vast effort of the marine community in Europe. This can only be achieved if a sense of purpose and an agreement on the main scientific issues is created. Although a number of documents are now in the making, especially the ESF Marine Science Plan, the definition and solution of the main issues and special problems in marine biodiversity research require this concertation..

The discussion conveyed a large number of registered scientists (270), who maintained a very lively transit of messages along the entire duration of the session. A total of 107 messages were posted.

Partial summaries for the first couple of days are still available:

• day 1
• day 2
• day 3
• day 4

1. Large-scale patterns of marine biodiversity

What are the large-scale patterns of marine biodiversity in Europe?

Global patterns on species diversity have been already elucidated, such as clear cline of increasing species richness from Arctic to ca 20oN, and the fact that tropics are not necessarily richest. The pattern in the southern hemisphere is far less clear as we have relatively little data. There is also little doubt that Indonesia has the highest species richness with radial declines from there, but there are "hot-spots" of high richness that occur elsewhere and need explanation. Yet, we are lacking comparable patterns within basins or sub-basins. The reason for that is that the differences are not as great as those present globally, requiring, therefore, high precision data to resolve them. Examinations at such large scales likely account for only a small fraction of the variance in biodiversity, so that examinations must resolve a variety of scales. Whereas large-scale patterns may be considered quasi-static, those at smaller scales are subject to temporal changes. Coastal areas may also show distinct large-scale patterns.

Evolutionary patterns in marine biodiversity are only evident at the largest scales, for these patterns are blurred by dispersal events (only a few extreme events may be sufficient to lead to established populations), so that within the biogeographic provinces patterns seem to correlate with environmental variability. The potential of analyses of large databases derived from existing sources to ascertain these correlations should be explored, although the patterns may be confounded by inter-calibration problems. These are ever more important as the capacities to professionally identify specimens becomes more rare. Perhaps technological tools may help to assist identification for some taxa. The myriad of metrics used to quantify beta diversity makes comparisons amongst studies cumbersome. Standard, agreed metrics must be devised is comparative analyses between estimates derived in different studies are to be used. This requires, most likely, an exercise to examine the extent of redundancy between metrics and to explore the statistical behaviour of the different metrics. The close relationship between species richness and effort also implies that, with the current rate of description, robust large scale patterns would require considerable effort.

The elucidation of large-scale patterns of marine biodiversity requires a proper appreciation of the relevant sampling scales and the samples required to adequately represent marine biodiversity at any one site. Adequate representation of sampling scales, including very large scales as to unravel patterns derived from evolutionary processes from those reflecting shorter-term processes, such as disturbance. The appropriate grain or scale of efforts to elucidate large-scale patterns must be determined, and this may differ across taxa. Exploratory analyses of the scale-dependence of marine biodiversity, across the broadest possible range of scales, is necessary as a support to plan and interpret further research.

The search for patterns on global marine biodiversity will be challenged by an uneven capacity to elucidate species membership across taxa. In addition, estimates of biodiversity, as species richness, are dependent on effort. Approaches that reflect taxonomic breadth or relatedness of species, seem to be better for broad-scale geographical comparisons than absolute estimates of species richness.

Provided these limitations, efforts to establish large-scale patterns in marine biodiversity must be operational, and the desire to describe synoptic large-scale patterns for all of marine biodiversity seems still a distant goal. The use of Functional groups may prove convenient. The relationship between species richness and the functional diversity of communities need be examined, as also stressed elsewhere in the discussion.

Taxonomic distinctness may be difficult to compare across taxonomically heterogeneous groups. These limitations can be circumvented if we use a consistent taxonomy in each case.

In addition, effort to establish large scale patterns should be driven by the formulation, and testing, of a priori expectations on the drivers of large-scale patterns (e.g. the existence of biogeographic boundaries, strong, permanent water fronts as barriers for dispersal, etc.), which could then be tested a range of organismal groups. This approach will be complementary to a strictly explorative one. Availability of suitable habitat is an important control on biodiversity, which may confound other existing patterns. This suggests that it ought to be, therefore, simpler to find large-scale patterns in the distribution of planktonic organisms. For instance, macro-Zooplankton has distinct latitudinal patterns of species richness, and, unlike benthos patterns, they are rarely endemic of high latitude provinces, although high latitude endemism is more frequent in the southern than in the northern latitudes.

The concerted action BIOMARE has proposed a pragmatic approach towards the elucidation of large-scale patterns in European marine biodiversity that may inspire additional efforts elsewhere. A set of Primary Reference sites throughout Europe has been proposed to establish comprehensive All-Taxon Biodiversity Inventories (ATBIs), which might enable us to calibrate the relationship between overall biodiversity and the diversity of key taxa or biodiversity indicators. Using these biodiversity 'surrogates' or 'indicators', we may then be able to establish patterns of biodiversity on a finer spatial scale at a larger number of reference sites. This initiative is an important step forward to render a complex question operational: by selecting protected areas as the target there may be hope to separate the ‘background’ large scale patterns of large scale biodiversity from effects derived from local anthropogenic forcing. Once the ‘pristine’ (i.e. as pristine as possible) conditions have been described we may investigate how these may have been affected by ‘point source’ anthropogenic effects.

2. Biogeochemical cycling and ecosystem functioning

What is the function of biodiversity in biogeochemical cycling and ecosystem functioning?

The evaluation of the relationship between biodiversity and ecosystem function and biogeochemical cycling has proved cumbersome. What communities would be most convenient as subjects to test this notion? Planktonic communities, where synthetic communities can be easily constructed from cultured organisms, may be ideal for ‘species addition’ experiments, whereas benthic communities may be most convenient for ‘species removal’ experiments. Meiobenthic assemblages in sediments may be good candidates to test the relationship between biodiversity and ecosystem functioning, since natural assemblages are much easier to maintain in replicated manipulative microcosm experiments in the laboratory than planktonic or macrobenthic assemblages.

Field experiments are complementary of (replicated) laboratory experiments carried out on experimental (or manipulated) communities. The respective advantages and drawbacks of field and laboratory experiments are well known and this is probably not the place to have this discussion once again. Laboratory experiments are definitely needed in order to unravel the relationships between diversity and ecosystem functioning. However, in this particular context, and due to numerous and complex interactions between compartments, we should consider problems associated to the time scale of our experiments. In other words, what is the exact meaning of manipulating a community to quasi immediately measure one of its functions? Should not we let this ‘new community’ reach a new equilibrium before carrying out experiments? Do we presently have the experimental infrastructure to do that? If not this should probably be one of our first priorities.

In this context, field experiments may constitute a sound complementary approach provided that ‘confounding factors’ are reduced. This is probably the case if strong heterogeneity in the composition of the community occurs at small spatial scale. There are cases of such communities for the macrobenthos.

Besides experimental ecology, there is also natural history. And historians do not make experiments. Ecology is an historical science. I concur that in ecology it is possible to perform experiments because it is ethically sound. Bin Laden, if it's him who made it, made an historical experiment, perturbing a system, he made predictions too. But he is not a historian, he is like the hurricane disturbing the coral reef. So, yes, we ecologists can perform experiments in our historical systems, but very small ones. Little questions with little answers. What about the rest? Nothing, because it cannot be tested?

Trophic interactions between organisms, which are sometimes highly species-specific (e.g. parasites such as virus), are strong determinants of the fate of populations and, thereby, bulk biogeochemical fluxes. Trophic interactions among organisms explain blooms and population outbreaks, which have been reported to have major effects on both benthic (e.g. sea urchin­kelps) and planktonic (virus-bacteria) populations. Moreover, these trophic interactions, as represented in simple models using Lotka-Volterra- like expressions to link highly selective and less-selective predation with the competitive ability of yield linkages between biogeochemical cycles and diversity than are normally ignored in the literature.

Biofilms and microbial mats are old relicts where biodiversity and biogeochemical diversity are confined within a thin layer. They can be good models to study the link between biodiversity and biogeochemical functioning. Biogeochemical processes are the result of species’ activity, and, in particular, bacterial activity. Yet, our understanding of the responses of bacterial biodiversity to perturbation is rather meagre. Yet, particular bacterial groups are responsible for very specific functions in biogeochemical cycling of the elements.

Which functional parameters should we use? One of the main difficulties in relating biodiversity and ecosystem function relies in the fact both of these parameters are resulting from a great number of complex interactions. Therefore, it is quite unlikely that the relationships between these two kinds of parameters will be simple. An alternative sound approach consist in focussing on function carried out at the level of functional groups (e.g., particle filtration for suspension-feeders.....).

Are biogeochemical cycles disconnected from food webs and from life cycles?

These topics are studied by different people in a very reductionistic way. Isn't it time to build complex theories putting things together? Is mathematics the formal language of complex systems? Isn't it that such issues are not tackled because the number of variables makes them intractable from a mathematical and an experimental point of view? After having split ecological systems in a myriad of sub units, isn't it time to start to put them together and find a conceptual continuity linking all these disconnected pieces? Ecology is the science of interaction, we tend to forget it for ease of analysis. Reductionism is alright only if it is followed by a synthesis.

3. Taxonomical and functional biodiversity

What are the links between taxonomical and functional biodiversity?

The biodiversity is very interesting not only itself. It is very interesting both as natural phenomenon influencing the activity of ecosystems, and as a consequence of ecosystem function. The assessment of the link between biodiversity and stability is a difficult one, and need be challenged using perturbation experiments.

Are rare species only rare at some spatial or temporal scales but ready to become abundant as the environment changes? Can a currently rare species be a potential key-species in the future?

Recent terrestrial and a few marine data sets show a linear relationship between local scale and regional scale species richness. Local species richness is likely to vary from place to place within the region depending on environmental conditions (both natural and anthropogenic), which must add noise to the possible relationship to regional relationship. The investigation of such patterns is fundamental to our understanding of how species richness is controlled. If this is the case, however, the use of species richness as an indicator of local environmental degradation may be difficult.

4. Species redundancy

Is there species redundancy? If so, what is its role?

Species redundancy is an old topic in plankton ecology (i.e. the plankton paradox), which was resolved by invoking non-equilibrium conditions to account for the co-existence of species with similar resource requirements. While the presence of ‘functionally redundant’ species can be explained, the question as to how much biodiversity can be lost before any effects on ecosystem functions is detected is still unresolved, both for marine ecology as for terrestrial ecosystems, where the experimental demonstration of a link between species diversity and ecosystem functions has proved elusive.

Functional redundancy cannot be readily tested. The concept of functional redundancy implies a near-perfect functional match in all relevant functions. This is impossible to test, since it can be always argued that there may be important functions that were missed by any one test.

The issue of ‘functional redundancy’ in animals that feed on particles (as opposed to phytoplankton that absorb nutrients in solution) is problematic. Just because there are several deposit-feeders present in a community does not mean that they are competing for resources or that there is functional redundancy. Most smaller species, at least, are very selective about what they eat.

To what extent do apparently redundant species contribute to the resistance and resilience of ecosystem functions to change and disturbance?

There are questions on this topic that can be addressed using planktonic organisms, such as shifts in the relative abundance of silicifiers versus calcifiers in ecological and geological time scales. Diatoms are large, tough, highly diverse and well suited to (physical) disturbance (mixing); coccolithophores are, on the other hand, small, and only outnumber diatoms in conditions of stability (stratified waters). Also, these organisms are available in the geological record (silicate/calcium carbonate sediments seem to be associated with glacial/interglacial periods). In this sense, we may argue that these may be good model organisms to elucidate questions such as (1) why are some marine (phytoplankton) groups more diverse than others in the modern oceans; and (2) how can we explain the lower degree of (genetic) diversity in freshwater ecosystems compared to marine ecosystems (diatom scenario)?

5. Life cycles

What is the importance of life cycles in the evaluation of biodiversity?

What is the relative importance of the dispersal (e.g. planktonic larval stages) vs. adult stages in determining the large-scale biodiversity patterns of sessile organisms?

What is the contribution of settling diversity to community diversity?

There is no evidence that benthic assemblages dominated by species with planktonic larval stages are any more or less diverse than those with direct development. There may also be a feed-back between the diversity of the community and that of settlers.

6. Comparison with freshwater and terrestrial ecosystems

What are the differences between marine, freshwater and terrestrial ecosystems?

Are large-scale patterns in marine biodiversity driven by similar gradients (elevation-depth, latitude) than those of land and freshwater biota?

Freshwater habitats are clearly more ephemeral (except deep lakes as Baikal etc.) than marine habitats, so isolation processes and species turnover in freshwater biogeographic regions may be expected to be enhanced compared to the marine.

How does biodiversity change along salinity gradients? There are excellent data sets that should help find any general patterns that may exist.

7. Rate of recovery following disturbance

What are the rates of biodiversity recovery following disturbance? What role does biodiversity has in setting recovery rates?

Establishing rates of biodiversity recovery is a key goal. There may be sufficient data as to conduct a meta-analysis for some particular taxa, however, it is difficult to standardise the disturbances as to render the recovery rates comparable. Patterns of recovery in macrobenthic communities are well established. In contrast, at local scales, rates of recovery seem to be much more variable and depend on e.g. hydrography, season, extent and nature of disturbance.

Are we able to predict rates of recovery?

Examination of case studies suggests that there may be no general answer to these questions. The decline of biodiversity following disturbances caused by physical disruption, pollution, over-fishing and introduction of exotic species differ in their mechanisms, and so will the possible recovery, taking into account differing spatial and temporal scales. There is an urgent need to establish the ‘point of no return’ for those disturbances in different communities - when is the change too extensive for a turnaround.

8. Ecosystem engineers and key-species

What are the ecosystem engineers and key-species?

The concept of keystone species is identical with intermediate disturbance. These concepts are probably important to understand species roles, but cannot be used to decide priorities in conservation. Advice to manage biodiversity may be premature before we have solid scientific evidence to base our advice upon. Engineer species modify habitat conditions and, therefore, possibly the pool of species to be encountered therein. These effects may be either positive or negative, but it is clear that such species (e.g. mangroves, seagrasses, kelps, some corals, etc.) hold a prominent position on the biotic control of biodiversity, and their loss or appearance may lead to important - albeit still poorly understood - changes. They should be particular targets for management once these interactions are understood

A link between species richness and stability link has been assumed for decades, but there is no robust evidence, and it is yet to be demonstrated. This should be one of the main questions to address experimentally (comparative analyses are much to confounded with other co-variates to be useful as a test of the concept).

Do engineer species really have such a prominent position, or is this merely an artefact of the restricted scales we study? Obviously they often sustain a high species density. But often they take a small area only on a regional (a few 100 km) scale. Therefore, in many species the majority of individuals may be scattered in the surroundings and the 'hot spot' of species richness may not be very important for regional population endurance. So, what happens if an ecosystem engineer is removed? Is there any lasting effect on the other species, or is it readily replaced by formerly rare (possibly outcompeted) species within a few decades (what would appear as a sudden change on an evolutionary time scale). Finally, if the environment changes dramatically, wouldn't conservation of an established engineer prevent natural succession?

Humans can also be considered as engineer species in coastal ecosystems. Do artificial structures enhance coastal biodiversity? The public rarely has a ‘direct view’ of marine species richness, except for habitats such as coral reefs and shallow rocky reefs. As a consequence there is a tendency by coastal managers to ‘enhance the quality’ of coastal habitats by using artificial structures. Artificial structures are also claimed to contribute to habitat conservation and restoration. This human activity often acts at large spatial scales on a very short temporal scale. Moreover, if little is known about species engineers, often acting at a very slow rate, even less is known about question such as: "What are the effects of artificial structures on coastal ecosystems?" "How do they affect biodiversity?" Deployment of artificial substrata on coastal sandy bottoms obviously increases the number of species in the area and makes it more ‘economically valuable’. However, alterations of ecosystem functioning (e.g. trophic webs, energy fluxes etc.) and of the abiotic features (e.g. hydrodynamic regimes, sedimentation rates etc.) need a careful assessment before planning creation of artificial rocky island.

Whereas preserving ecosystem engineer species would preserve species that are of ecological importance, species that contribute most to preserving the evolutionary history of a taxonomic group need particular attention too. Species that diverge close to the base of a phylogenetic or taxonomic tree and have few close relatives will preserve more evolutionary history than those that diverge further up and have more congeners. There is quite a big literature on the use of Taxonomic distinctiveness in the selection of species for conservation, but little evidence for its application in practice.

9. Non-equilibrium conditions

The larger the spatial scale, the more we may expect diversity to be a reflection of the steady-state conditions. So, what is the spatial scale of a 'particular marine community' (respectively, the spatial scale we should use in a study on diversity)?

There may be an inverse relationship between disturbance frequency and spatial scale, which may support the proposed likelihood that large-scale patterns reflect some sort of steady-state, or average, conditions, whereas local biodiversity patterns may more often reflect non-equilibrium situations.