Friday, April 19, 2024

Jan Perret: Plants stand still but manage to hide

EcologyJan Perret: Plants stand still but manage to hide

2023 HARPER PRIZE SHORTLIST: Throughout March, we are featuring the articles shortlisted for the 2023 Harper Prize. The Harper Prize is an annual award for the best early career research paper published in Journal of Ecology. Jan Perret’s article ‘Plants stand still but hide: Imperfect and heterogeneous detection is the rule when counting plants‘ is one of those shortlisted for the award:

👋 About me

I grew up in western France, in a landscape of bocages and deep forests, with, of course, many types of cheese. As a teenager, I was bitten by the nature bug and my curiosity became a full-blown obsession. I roamed the countryside, unravelling its secrets, from the tiniest critters to the largest trees.  So, naturally, I decided to turn my passion into a career. At 18, I started my Biology-Ecology Bachelors at the University Paris-Saclay. For three lively years, I delved into the world of flora and fauna (a little less into biochemistry). My next stage was a Master’s in Biodiversity Management at the University of Montpellier, a degree that prepares students to work in protected areas with local actors, because, at the time, I aimed to find a job where I could work outside as often as possible. I took a slightly different path, as I now spend most of my time behind a computer doing statistics… but, believe it or not, I love it!

During my studies, I had the luck to do several internships in nature reserves. They were quite a task – my main activities were lugging heavy gear through rugged terrain, and taking part in the reserves’ monitoring programmes, or in other words, counting all sorts of organisms… trees, newts, birds, bats—you name it, I counted it. Oh, and plants? Yeah, I counted more plants than I care to remember. It was during my last internship, knee-deep in a meadow to count the region’s only population of a rare plant species, that I started to have doubts… What if our monitoring methods were missing the mark? The data seemed wonky, and I had the sneaking suspicion that I was missing many individuals, hidden in the meadow… A year later I started a PhD on the improvement of plant population monitoring methods, and the study published last spring in the Journal of Ecology is part of my doctoral work.

Figure 1: Picture of the experiment in progress at a Mediterranean site, near the Pic Saint-Loup, 25km north of Montpellier, France. The participants are making a count of individuals for an ‘easy’ species, the pyramidal orchid, Anacamptis pyramidalis. The question is, did they notice the vegetative individuals hidden in the scrubs?

🔎 About the shortlisted article

Missing individuals during a count (i.e. imperfect detection) results in biased estimates of population sizes and trends. In addition, if detection varies over time, for example because the habitat undergoes succession during the study and the individuals of the target species become harder to detect, it can lead to the detection of non-existing trends. Imperfect detection is the rule in animal studies, and many monitoring programmes now correct this bias by estimating detection probability. Yet this correction remains exceptional in plant studies, suggesting that most plant ecologists implicitly assume they always detect all individuals. To assess whether this assumption is valid, I conducted a field experiment in which I asked botanists with various experience levels to count plants in 1 × 1 m quadrats with three different counting methods. I did this on 30 species of varying visibility, and I chose study sites at different stages in vegetation succession. In total, 158 participants took part in the experiment, resulting in a dataset of 5,024 counts.

The results were stunning! No observer ever detected all the individuals in the 10 quadrats of a field session. Detection varied with the counting method, but more importantly with the species’ visibility and successional stage. Surprisingly, the density of individuals in the quadrat had a strong negative effect on detection, i.e. the more individuals there were in a quadrat, the lower was the proportion of individuals that were detected! I hypothesised that this could result from a fascinating phenomenon that is well documented in human cognition: the Weber–Fechner law! On the other hand, the participant’s experience in botany had little effect on detection.

The results of this experiment show that imperfect and heterogeneous detection is the rule when counting plants. So, my intuition was right; to avoid obtaining biased estimates of population sizes and trends, plant ecologists need to use methods that estimate the detection probability of individuals and not rely on raw counts.

🌱 What’s next?

I currently have a one-year teaching and research position at the centre for functional and evolutionary ecology in Montpellier, France, where I did my PhD. I continue to work on applying methods for taking into account detection errors in studies of plant population dynamics, as these methods have been little used for plants so far. I also collaborate with French plant conservationists to improve their monitoring programmes. The next step, after plant populations, is to develop methods to account for detection errors in studies of plant communities! I have just started to write the project, and hope to find funding to start the study next spring, and go back to the field again!

Find Jan on X, GoogleScholar and ResearchGate.

Read the full list of articles shortlisted for the 2023 Harper Prize here.

Check out our other content

Most Popular Articles