Generating ideas for research questions in ecology & environmental science

Making new links across different dimensions of ecology and beyond

(Updated 24.6.2023, 3.7.2023, 16.4.2024, 8.9.2024; 11.2.2025)

Generating ideas in ecology (image created in leonardo.ai)

Asking new questions is a great way to advance science. But coming up with new, interesting questions to ask about ecology and the environment — or any other topic — is incredibly difficult. This continuously updated post tries to summarize some of what has worked in the past. Here is what I learned from reflecting on some of our lab’s own work or when reading papers that others published.

Creativity typically entails connecting among topics in a way that brings about something novel and useful. Thus, most of the points below are about such connections. Can we use such a catalog of approaches in a systematic way to generate new questions in ecology and environmental science?

Here is the list of approaches that you can try.

• Playing with ecological hierarchies. Many concepts and ideas in ecology are defined with a particular level of the ecological hierarchy in mind: individual, population, (functional group), community, ecosystem, landscape. Can we move concepts from one such level (where it has been defined and studied) to another one and explore the consequences of this ‘transfer’? One example from our work is the idea of not just individuals interacting, for example competing, but entire communities. This led to the idea of microbial community coalescence: the meeting of entire microbial communities.

• Transferring experimental designs from one field to another. Take the experimental approach to disentangle species richness (i.e. number of species) effects from the effects of composition or identity of species. Tilmann pioneered the ‘random sampling of species from a species pool’ approach to achieve this. What if we replace species with global change factors: you can use the same exact design, but apply it in a totally different context: asking about the effects of the number of global change factors.

• Exchanging settings. Settings in ecology can for example be ecosystems types or compartments, for example aquatic versus terrestrial, or above-ground vs. below-ground in terrestrial ecosystems. Often, something has been studies in one setting, but not in another. One example: what about microplastics in terrestrial ecosystems and the soil vs in the oceans.

• Transferring to other organism groups. There are huge differences among different organisms. For example, fungi are linear, modular, weird and awesome. Many things don’t immediately make sense for this type of organism, like what is an individual. If we we apply concepts to this group from other groups that are unitary, what will happen? For example, what does ‘movement’ mean for fungi? A transfer that is generally promising is that from larger organisms to microbes. A recent example is the idea to transfer the ‘old growth’ concept from forests and grasslands to microbes.

• Transfer among temporal and spatial scales. Moving along temporal and spatial scales is of course an age-old and large unsolved general challenge in ecology. Can we transfer general mechanisms or concepts that have been designed for one spatial or temporal scale and move them to another?

• Completing the matrix. What I mean here is this: find broadly related but distinct concepts, and then see what has been done for one of the concepts, but not the other. An example is to think about the pair “climate change” and “global change”. Many things have been explored for climate change, but maybe not for global change. One example from my work: global change refugia (in analogy to climate change refugia).

• Importing concepts from other areas of biology. Many areas of biology have followed a separate trajectory, for example ecology and evolution. What can be gained by exchanging concepts between such fields, for example between ecology and evolution. See an example by Loreau et al (2023).

• Importing concepts into ecology from other sciences. The examples above mostly stayed within the realm of ecology and the environment, but it is of course possible to import concepts from very different fields of science. This is quite challenging, since it requires knowledge about a very different domain, but also potentially very rewarding. One could import concepts from economics, such as principles of trade, and apply it to the exchange of goods in a mycorrhizal symbiosis, for example.

In terms of connecting two points, of course, typically an idea is more innovative if the ‘distance’ bridged between ideas is greater. Like it may not be as transformative to ‘travel’ the distance from deciduous forests to evergreen forests, or from one soil type to another. In order for the connection to have the potential to really open new opportunities for study, the distance needs to be a rather big one, and the link should be non-obvious.

Perhaps not everything is about making such new links. There are also some other ways to generate ideas. Here are some:

• Examining responses over a wider range of levels. This is an approach that does not deal with a connection of two points. Often in experimental ecology we examine yes/ no or low/high kind of responses. Very interesting insights can be generated by thinking about a driver in terms of being represented by a a lot of levels, permitting an analysis of dose-response type relationships where they are typically not examined. For example, this paper in Nature experimentally manipulated immigration rates over a wide range of values (12 different ones), revealing an interesting pattern.

• Examining a phenomenon by experimenting (or observing) at a different order of magnitude. We typically think of individual experiments at individual sites, but networks like NutNet or DroughtNet or similar distributed experiments have shown that increasing the scope of experimentation, like by an order of magnitude, can lead to new insights. A similar point applied to observational studies, which now often are global in scope, instead of local or regional. Can we play with this idea?

• Making a response variable a treatment. Maybe this is not often possible, but it may be worth a try. Recently, I flipped a readout-variable (response variable) around to make it a treatment: antibiotic resistance genes. This can open completely new opportunities.

Other than the three points above, what other approaches are there that do not necessarily connect among dimensions of ecology?

What other categories or dimensions of ecology are there for making new linkages, for finding new questions? Can you think of any? Can we generate an ecological ‘question engine’ that combines all of the above?