Tag Archives: species distribution modelling

On Dispersal, Connectivity and the Will of the Fish

Image Credit: Dennis Jarvis, CC BY-SA 2.0, Image Cropped

Integrating dispersal along freshwater systems in species distribution models (2020) Perrin et. al., Diversity & Distributions, https://doi.org/10.1111/ddi.13112

The Crux

Trying to figure out where a species can comfortably live is one thing, but figuring out which habitats they can actually access is another. I like to think most marsupials would do quite well in South America or Africa, but the fact is that they’re not dispersing across the Atlantic or Pacific anytime soon. However a Species Distribution Model (a statistical model that can be used to predict the likelihood of a species being found somewhere) often requires a more nuanced approach than “big ocean separating these two habitats”.

To integrate a species’ ability to actually access an area into a Species Distributions Model (SDM), we often use the concept of connectivity. Often, this means simply measuring the distance between two populations. But sometimes a species ability to disperse might not reflect something as simple as how far it needs to go. A perfectly good habitat might be only 100 metres away, but cut off by a raging great cliff. Or a road.

In this study, we wanted to see whether we could relate connectivity parameters used in an SDM to the actual ability of the species to disperse.

What We Did

We used two separate study systems here. One consisted of roughly 300 lakes within Northern Norway housed within a single catchment, or watershed, whereby a single path between each lake could be traced. Here we had presence-absence records for two species, the northern pike (Esox lucius) and the European perch (Perca fluviatilis). Both are native to the region, but they are starting to expand into more lakes and have a more severe effect as the climate warms. We used an SDM to investigate which factors determined species presence, including connectivity variables like the length of the rivers between each lake and a downstream population, and the average slope of those rivers.

The second ecosystem was a series of lakes in Sweden which pike and perch had previously occupied, but had been removed from in the 60s and 70s through the use of rotenone, a chemical dumped in small lakes which wipes out fish populations. These were useful, as we knew that the lakes were otherwise suitable for the species given their presence beforehand. As such, here we used a much simpler model to focus on dispersal ability, simply comparing whether or not the species were able to access and then recolonise the lakes from which they had been removed. We compared successful recolonisation from the nearest downstream lake to the same connectivity parameters as in the larger model.

Did You Know: Island Biogeography & Lakes

They obviously don’t look it, but when it comes to biogeography, lakes are essentially a special type of island. Most of the rules of island biogeography apply to them (for fish anyway); larger lakes are more likely to have more species, lakes close to the ocean or other large lakes (the ‘mainland’) are more likely to have those species as well. The big difference between regular islands and lakes is that we can mark pathways between them much more easily. You’d think that would make it easy for us to stop fish spreading into new lakes as the climate warms, but the problem is as always people – people often spread fish from lake to lake, and the rules of island biogeography don’t apply in quite the same way to someone with a car.

What We Found

The slope of the river was a much more important factor in determining a species presence than the actual distance between populations. This makes sense, as a steep slope could make it difficult for a fish to swim up, or could indicate the presence of a waterfall. Furthermore, adding connectivity parameters to our SDM in our first study system did improve our models, but did it represent dispersal accurately?

For pike, the effect of slope was pretty consistent across the two study systems, indicating that the effects of connectivity in a large SDM can mirror a species dispersal ability. However for perch there was some inconsistency across the two study systems, indicating that perhaps there was some other aspect of the rivers between populations that had a larger effect on dispersal.

While European perch might be native to parts of Scandinavia, it is alien to others. If it’s able to freely disperse between lakes, it could be a serious problem as the climate warms (Image Credit: Christa Rohrbach, CC BY-SA 2.0)


This study suffers from the same “lab vs. field” pitfalls as any other experiment that compares a complex study system to a smaller, ‘simpler’ one. Here, time is a factor. Our first study system looks at populations that have had centuries, in some cases millenia, to establish, whereas the second one looks at short-term re-establishments. It’s possible that given enough time, pike or perch could have eventually recolonised some of those lakes.

So What?

Having an idea of the effect of how different slope measurements can affect the dispersal of species is a great help, as it lets us know which lakes are protected by natural dispersal barriers, and which are likely to be invaded by species moving from downstream. However the fact that for perch, slope parameters varied in their effects across the study systems is a stern reminder that we need to always be mindful of how connectivity parameters actually relate to dispersal ability.

Sam Perrin is a freshwater ecologist currently completing his PhD at the Norwegian University of Science and Technology who is now completely done with this paper and never wants to look at it again. You can read more about his research and the rest of the Ecology for the Masses writers here, see more of his work at Ecology for the Masses here, or follow him on Twitter here.

Ecological Modelling, the Coronavirus, and Why They’re Not A Perfect Match

Image Credit: Pharexia, Ratherous, AKS471883, Source Data from  Johns Hopkins University CSSEThe Centers for Disease Control and PreventionNew York TimesCNBC.

As it quickly became clear in late February and early March that COVID-19 was not going away anytime soon, attention turned to trying to figure out when and where the virus would spread. Epidemiologists and virologists have had their work cut out for them, trying to simultaneously reassure and warn people the world over about the dangers, the nature and the potential timeline of the virus.

So it came as somewhat of a surprise to see ecologists try and tip their hat into the ring. Early on in the pandemic, teams of ecologists sprang up, trying to use Species Distribution Models to predict the spread of the virus. And whilst this might sound helpful, many of these studies lacked collaboration with epidemiologists, and their predictions very quickly fell flat. Some studies suggested that areas like Brazil and Central Africa would be largely spared by the virus, which quickly turned out not to be the case. Flaws in the studies were spotted quite quickly by concerned members of both the ecological and epidemiological communities alike, and a few teams got started on responses.

Read more