Summary – On the organization of ecosystems: ecological autocatalysis in African savannas

One of the biggest challenges humans, especially ecologists, face is to understand the functioning of the natural world. This is of crucial importance, whereas changes in land use, biodiversity and climate, as a result of human impact, set the remaining natural areas under pressure.  Increased understanding in the functioning of ecosystems will help to manage, protect and restore nature, which will be very important for creating a sustainable future for human beings on planet earth.

Ecosystem organization

The structure and functioning of nature can be captured in three key elements. First, chemical elements like carbon, nitrogen and phosphorous: the building blocks of all existence. Second, energy, that enables us to grow, maintain ourselves and reproduce. Third, there is a certain structure or organization that describes how these building blocks and energy move through space and time. One could compare this with a house, where bricks represent the building blocks, and construction workers build the house, maintain it and maybe break it down some day. The blueprint represents the organization of the house. When we then scale up to a village or city, certain patterns can be observed. You might find small houses at some place, tall houses at another. A church in the centre. Neighbourhoods are build, others disappear. The villages is extending or shrinking. Such patterns can be described as the organization of a village and can change through time. In a similar way, ecosystems are organized in specific configurations. At some places you find lots of trees, at other place mainly grasses or herbs. Sometimes you find many herbivores, sometimes large colonies of birds. This “ecosystem organization” is the subject of this thesis. I ask myself the question which factors determine the configuration of ecosystems and whether external factors, like climate, are most important, or whether organisms themselves control the ecosystem configuration and processes. This sounds very ambitious, and I won’t deny that it is. Therefore, this thesis should not be seen as a discussion of a demarcated subject that is intensively investigated and analysed, and results in clear conclusions and possible solutions. In contrast, this thesis should be seen as an exploration of a general concept that describes and explains the organization and functioning of ecosystems, that is currently lacking. It connects knowledge on (interactions between) species with our knowledge on (components of) ecosystems. Autocatalysis is a key element of this concept.

Ecological autocatalysis

Autocatalysis is a self-reinforcing process, well studied in disciplines as chemistry and physics, but relatively rare in ecology.  Ecological autocatalysis is a positive interaction between organisms in a circular fashion, through which a species positively affects itself through positive effects on other species. A classic example is bladderwort, Utricularia, a carnivorous plant that occurs in freshwater systems and acts as a substrate for algae and cyanobacteria. These algae and cyanobacteria are eaten by zooplankton, which subsequently is eaten by Utricularia itself. Therefore, bladderworts positively affect algal growth that increases the amount of zooplankton, which subsequently stimulates bladderworts. Altogether they thus from a group of organisms that positively affects each other in a loop and we therefore describe this as ecological autocatalysis.

Positive interactions between species have long been underestimated in ecology, where since Darwin proposed his theory of evolution by natural selection, negative interactions have received most attention, whereas the struggle for existence results in competitive (negative) interactions between species. This has changed in the last decennia and processes like facilitation (often within trophic levels) are more and more investigated. Also, mutualism (often between trophic levels), a strong direct positive interaction, receives much attention, often to investigate whether and why cheating is important.  Indirect, often weak, positive interactions between trophic levels are hardly studied. Ecological autocatalysis is one of such indirect positive feedbacks and I develop this concept further in chapter 2 towards a general principle within ecosystems using some general examples. Furthermore, I argue why this concept is useful in explain the pattern we observe at the ecosystem scale. Last, I describe in the synthesis (chapter 9) how we can apply this concept to African savanna ecosystems. The chapters in between are used to gather additional evidence for the concept.

Ecological autocatalysis in African savannas

Three distinct autocatalytic loops can be described in African savannas (see chapter 9, figure 1). 1) The first consists of lawn grasses, of which most biomass is consumed by large herbivores. Herbivores subsequently produce dung that is used by dung beetles and termites and further decomposed by soil fungi and bacteria. Nutrients that are released can be taken up by lawn grasses again. 2) Bunch grasses are generally avoided by large herbivores because of their low nutritional value. When they become dormant at the onset of the dry season they function as an excellent fuel for fire that “consumes” these bunch grasses. The nutrients in the ash subsequently return to the soil where bunch grasses can take them up again. 3) Woody vegetation (trees and shrubs) is not so much consumed by herbivores and often outgrow fire. The produced leaves fall down as soon as they senesce and are consumed by soil organisms. Earthworms, isopods, millipedes and termites all forage on the litter, which is further decomposed by soil fungi and bacteria. Nutrients that are released become available to trees and shrubs again.

My thesis

This thesis comprises nine chapters, with chapter 1 as a general introduction on the subject. Chapter 2 describes the concept of ecological autocatalysis and depicts how the concept can be used to describe and explain the organization and functioning of ecosystems.

In chapter 3, I investigate the relationship between large herbivores and grazing lawns. The positive effect of lawn grasses on large grazing herbivores is mainly explained by their high nutritional value and has been thoroughly studied and documented. However, how herbivores positively feedback on lawn grasses is less clear. Current literature mainly describes the increased nutrient concentrations in the soil as a result of dung and urine deposition by herbivores. Lawn grasses profit from this and are better able to withstand grazing than bunch grasses. I investigated another feedback mechanism, in which herbivores change the soil water balance. Trampling by large herbivores results in a compact soil with low water holding capacity. Lawn grasses are adapted to this low water availability, whereas bunch grasses need much more water. Therefore, large herbivores increase the extent of grazing lawns, from which they profit themselves.

In chapters 4, I quantify the spatial redistribution of nutrients by herbivores and dung beetles. I found that large herbivores effectively transport nitrogen from bunch grasslands to grazing lawns, creating long-term nutrient hotspots. Dung beetles showed a smaller but opposite effect and distributed dung deposited by large herbivores on grazing lawns back to bunch grasslands. Nevertheless, the main effect observed was a net export of dung (and therefore nitrogen) from all grasslands towards so-called “middens” by white rhinoceros.

In chapter 5, I study plant nutrient limitations. I am mainly interested whether these limitations are related to large environmental gradients (like rainfall) or determined by local interactions, as a result of the autocatalytic loops that have been described earlier. I found support for the latter. Shrubs, bunch and lawn grasses situated just several meters from each other, show large differences in the limitation of nutrients. In contrast, similar vegetation types separated multiple kilometres from each other are limited by the same nutrient. These results suggest strong biotic feedback mechanism on plant nutrient limitations.

In chapter 6, I further investigate this topic, now focusing on litter decomposition. Here, I also find that organisms, represented by termites, decouple the process of decomposition from environmental variation in rainfall. Free living bacteria and fungi show decreased activity under dry circumstances. However, specialized fungi, living inside termitaria are not affected, whereas the termites create stable conditions (moisture and temperature). Therefore, this mutualistic interaction enables high decomposition rates, even in dry periods.

In chapter 7, I study grassland heterogeneity. Simplified, I investigate why bunch grasses are tall, and lawn grasses are short. Does this reflect differences in productivity, consumption or both? Results show that lawn grasses are mainly less productive, with nicely fits with the findings from chapter 3 (low moisture availability). Remarkably, 75-80% of the lawn grass primary production was consumed by large herbivores, in contrast to ca. 40% of bunch grasses.

In chapter 8, I study a similar question regarding woody species. Why do we observe such a high variation in woody cover? And what determines the spatial patterns of these woody individuals? The main new insight from this chapter is that woody species are more or less evenly distributed under low rainfall conditions, whereas under high rainfall they often aggregate. A possible explanation for these patterns is competition for water (and therefore large distances between trees) at low rainfall, while at high rainfall trees facilitate each other in the protection against fire. Closed canopies increase moisture and decrease grass growth, both effectively increasing the likelihood of excluding fire.
In chapter 9, I synthesize all these new findings and incorporate them in the concept presented in chapter 2. Detailed results on feedbacks (chapter 3 and 4), their effect on ecosystem processes (chapter 5 and 6) and the consequences for ecosystem organization (chapter 7 and 8), are described and substantiated with existing literature.

Altogether, I presented a new concept in this thesis, which has the potential to become a general applicable theory that describes and explains the organization of ecosystems. Furthermore, I collected additional information to substantiate these ideas. Nevertheless, it will take years, maybe decennia to really prove the concept. I will definitely continue with this myself, but mainly hope that this thesis inspires other ecologists and observe their own ecosystem from a different perspective, trying to identify and investigate autocatalytic loops. Especially these indirect positive interactions that might not seem so evident at first sight, could play a much more important role than currently acknowledged.