PhD Graduation

Predicting phytoplankton community dynamics in detail seems an overwhelming task as there are so many species, and a myriad of combinations of potential conditioning factors. Furthermore, even with full knowledge of all aspects of species biology intrinsic chaos in communities may make detailed prediction fundamentally impossible. Aggregated estimators of phytoplankton communities may work to predict overall community responses to varying environmental conditions. However, phytoplankton species differ strongly in their effect on ecosystem functioning and ecosystem services. Therefore, it is important to consider community composition rather than just biomass.

This thesis focuses on the question whether species might be clustered in groups that are reasonably homogeneous in a functional sense, and might be better predictable from environmental conditions than individual species. To answer this question we first explored the factors that affect richness and biomass at the species level and then evaluated how well trait-based groups of species capture function and may be predicted from environmental conditions. We used a large data base including more than 700 species from 200 lakes in different climate zones and continents.

Firstly we evaluate which are the main factors that appear to determine the number of species in phytoplankton communities of a group of subtropical shallow lakes in relation to diversity of other groups of organisms found in these lakes. Our results indicated that, as found for temperate systems, submerged plant cover and transparency promote higher species richness in several groups, including phytoplankton.

Then we analyze differences in predictability of individual species from commonly measured environmental variables such as nutrient levels and zooplankton abundance. The presence or absence of species could in general be relatively well predicted. By contrast, biomass of most species appeared to have little relation at all with environmental variables, with the exception of a small group of species from diverse phylogenetic and functional groups that appeared to be relatively predictable. Such predictable species tended to reach a high biomass, and occurred relatively more in situations where competition for resources seems less severe.

Moving forward from individual species we propose a functional classification of phytoplankton species based exclusively on organismic morphology. We first showed that morphological characteristics are systematically correlated to functional properties, such as growth rate and sinking rate, and also to the population size and biomass attained in the field. Then we used cluster analysis to define seven morphology-based functional groups (MBFG) based on the selected morphological traits. Functional traits and demographic parameters not used for the classification differed significantly among the clusters, suggesting that they may indeed represent meaningful functional groups.

We then explored how well the aggregated biovolume of groups of species can be predicted from environmental variables using three different classification approaches: MBFG, phylogenetic groups, and functional groups proposed by Reynolds. Groups from all classifications were more closely related to environmental conditions than individual species on average. However, results indicated that MBFG can be predicted better from environmental conditions than groups based on the other classification methods. This suggests that morphology captures ecological function of phytoplankton well, and that functional groups based on morphology may be most suitable to focus on if we aim at predicting the composition of communities.

Finally, a simple model was constructed to simulate dynamics of morphology based functional groups using information on physiological rates obtained from literature. A trade-off emerged between maximum growth rate and nutrient competitive ability for most MBFG groups. Furthermore, model predictions were in line with patterns in field data and with ecological characteristics typically associated to the morphological features of the different groups.

In the synthesis I argue that my results suggest that we might think of phytoplankton communities as subsets of an omnipresent pool of cosmopolitan species selected by local environmental conditions. I speculated that the selection process would basically work on functional groups, whereas the relative importance of particular species within such groups could be essential random, as those species are functionally equivalent and therefore interchangeable.

Promotor: Prof.Dr. Marten Scheffer