Ecology is not a dirty word

Last year I wrote about the Insect Armageddon story – an important paper that received some exaggerated media hype.

A new paper just published in PNAS adds another twist to the insect declines saga…clearly, this story is far from over.

Lister & Garcia analysed data collected in the Luqillo Experimental Forest in Puerto Rico. This area of tropical rainforest is not a ‘pristine’ untouched wilderness, as some media reports are claiming – no place on Earth is untouched by humans! The site has been a long-term research location for decades, going back to the early 1900s, with a focus on experiments to understand the effects of disturbances of all kinds. Many important experimental research projects involving human disturbances (like this one) have happened in the Luqillo forest.

This study is important for a few reasons.

Same exact location, same methods, same data collector. This reduces the risk of confounding factors affecting the data, which can happen when comparing data from different locations, or data that have been collected with slightly different methods or by a different person. This means we can be sure the data from the 1970s is comparative with the data from recent years.

Not all insects have the same ecology. This is a critical fact to understand. Some insects live on land and some in water; some eat poo, some flowers, some rotting meat, and some leaves; some fly, some crawl. Any study of ‘insects’ needs to include multiple sampling methods, because the way you collect insects will determine what insects you actually collect. For example, most moths fly at night, so sampling moths only in daylight is useless; different insect groups have different colour preferences, so some coloured traps will only attract particular groups of insects. The authors used sticky traps on the ground and sweep nets to collect data – both are standard methods for collecting particular groups of insects (but not all insects). The graphs show lower biomass in recent years of some taxonomic groups, particularly Homoptera (leafhoppers, aphids), Diptera (flies), Coleoptera (beetles), Orthoptera (grasshoppers, katydids, crickets) and Areneida (orb spiders).

Multiple species and interactions. This study also collects data on insectivorous lizards, frogs and birds, again using the same methods in the same locations. They only focus on one genus of lizards (Anolis spp.), one species of frog, and small birds that can be caught in mist nets (see point above, same goes for other animals). But this is a neat addition to the story, as it illustrates just how connected ecosystems are. Interacting species affect each other’s populations, so focusing only on one species tells us only one small part of the story. If insects decline, so do the populations of other animals that rely on them for food.

Long-term climate data. The authors used long-term climate data from two stations in the forest itself. This increases the likelihood of the predicted relationship between observations and local conditions (compared to if they had used broader-scale data from the region that may not reflect the local conditions).

Accurate analysis. I’ve seen some people claiming the study is just showing correlations…the implication being the interpretation isn’t valid. This isn’t true. Sure, there was no direct observation of lizards rolling over en masse as insects rained from the sky… But the authors have used rigorous statistical methods to provide the most accurate analysis of their data. They also take care not to misuse causal language throughout. So we can be confident that the results are exactly as the authors describe: they are not claiming a causal relationship, but they are showing strong support for the hypothesis that climate change has contributed to insect declines.

This doesn’t mean that other factors aren’t at play. Nature is made up of complex interactions, many of which we don’t know about or aren’t able to predict. There are always going to be multiple factors affecting any observed pattern in natural systems. Science isn’t made on a single study. No study can prove anything beyond the methodological context of that study. But, as individual studies asking similar questions build up, we understand more and more.

The insect decline story isn’t over. This study isn’t ‘hyperalarming’, as some media outlets are claiming…and which isn’t even a real word. But it is a red flag. It adds to the mounting evidence that insect populations around the world are changing and we don’t really know why.

Insects are keystone species in many systems and provide multiple ecosystem services. Clearly, the ways we manage our local patches are having global repercussions on insect populations. The knock-on effects to the ecosystem services we take for granted are still unknown. We need many more studies like this to understand the implications. Investing in more long-term monitoring locations, training more entomologists, insect taxonomists and ecologists, and building greater knowledge of ecosystem services should be priorities.

© Manu Saunders 2018