From plants to smartphones: how I wandered through science

Posted on February 17, 2026 by lembrechtsjonas

New paper out! –> Decorte et al. in PLOS Digital Health

This paper might be the furthest I have ever ventured from my core scientific discipline. And honestly? That feels very much right.

I started my academic life as an ecologist, fascinated by where plants live, why they live there, and how small-scale environmental variation shapes those patterns. That curiosity gradually pulled me toward microclimate: the fine-grained temperature, moisture, and radiation conditions organisms actually experience. From forests and mountains, it was only a small step to start asking similar questions in cities. And from there, to realize that if we really want dense, high-resolution environmental data, we cannot do it alone.

Alpine plants – where it all started for me

Enter citizen science.

Projects like CurieuzeNeuzen in de Tuin showed me how powerful large-scale public participation can be for environmental monitoring. Thousands of gardens suddenly became part of a distributed sensor network.

From ecology to microclimate monitoring, to citizen science – largely with those ‘garden daggers’, as our citizen participants called the TOMST microclimate loggers

This hard-core citizen science project is what eventually brought me to De Oorzaak, a similarly large citizen science project on environmental noise. While noise itself is an environmental variable, the project deliberately went much further – into perception, experience, and health. Working in this space meant exciting collaborations with psychologists, communication scientists, and health researchers, and learning new ways of thinking about data, causality, and impact.

Installing environmental sensors with citizens – the clear connection between my microclimate work and De Oorzaak.

Perhaps not a surprise: much of this scientific “wandering” happened during my postdoc years – a phase that is, for many young(is) scientists, defined by short contracts, shifting funding opportunities, and a constant need to adapt. Each new position came with its own thematic focus, and rather than resisting that, I embraced it. And it’s only slowly that I started to realise: the uncertainties around funding did not just shape where I worked, but also what I worked on. Looking back, this period pushed me to become far more interdisciplinary than I would ever have planned on paper.

It also means that I now find myself as a co-author on a paper in PLOS Digital Health.

Smartphones, sleep, and a more nuanced story

Smartphones are often portrayed as the villains of modern sleep. We have all heard the narrative: screens keep us awake, notifications fragment our nights, and scrolling in bed equals bad sleep.

What I find refreshing about this study is that it steps away from that simple storyline.

Smartphone use per day (left) and in bed per day (right) among our participants, as obtained objectively from their iPhone and Apple Watch tracking data

Using donated data from participants’ own iPhones and Apple Watches, the ‘De Oorzaak’-team followed 68 participants across 14 consecutive days, tracking:

  • Total sleep duration
  • Sleep stages (REM, core, deep)
  • Total smartphone use
  • Smartphone use while in bed

And the patterns that emerged were nicely nuanced.

More total smartphone use during the day predicted more smartphone use in bed – no big surprise there for anyone, I guess. But more in-bed smartphone use was associated with slightly more total sleep that same night. That does not mean scrolling causes better sleep. But it does challenge the automatic assumption that phone use in bed is always harmful.

One interpretation is that some people may use their phone as part of a wind-down routine. Another is that longer sleep following in-bed phone use reflects compensation after poorer sleep earlier. This, we can’t say based on the limited data we have. What we can say, however, is the following: the relationship is more complex than “phones ruin sleep.”

Perhaps even more important, the strongest effects were not day-to-day fluctuations but stable differences between people. In other words, habits matter. Some individuals consistently combine higher phone use with particular sleep patterns, and understanding those habitual profiles may be more informative than focusing only on nightly variation.

Back in ecology – but not the same ecology

Today, I am firmly back in the ecology camp. I once again spend most of my time thinking about biodiversity, microclimate, and how organisms experience their environment.

But I am doing so with a very different toolbox than when I started.

Those postdoc detours into citizen science, environmental monitoring, and human-centered data have fundamentally reshaped how I approach ecology. I now see interdisciplinarity not as a side quest, but as core infrastructure, which is very helpful to make the necessary change happen to save that precious biodiversity I’m working on.

So yes, I now co-author a paper about sleep and smartphones.

And I see all this not as a detour from my scientific trajectory, but as one of the stepping stones that made me a better (perhaps more practical) ecologist.

Posted in Belgium, General | Tagged , , , , | Leave a comment

De Oorzaak’s final report

Posted on February 17, 2026 by lembrechtsjonas

We’ve written before on The 3D Lab about De Oorzaak as an ambitious citizen science effort to understand urban soundscapes in Flanders. This week marks an important milestone – the cherry on the cake: the project’s integrated scientific final report is finally out! It brings together years of work into a single synthesis that combines environmental measurements, citizen experiences, and health research.

You can find the mastodont of a report – of which we’re super proud – here: https://www.uantwerpen.be/nl/projecten/de-oorzaak/ (it’s in Dutch, though).

De Oorzaak – for those who missed it so far – is a collaboration between Universiteit Antwerpen, Universitair Ziekenhuis Antwerpen (UZA) and De Morgen, and has grown into the largest citizen science project on environmental noise ever conducted in Flanders. Hundreds of smart sensors, deployed across more than a thousand locations, were paired with ten thousand questionnaire responses to move beyond a simple “how loud is it?” question. Instead, the report asks what kinds of sound environments people actually inhabit, and how these relate to sleep, fatigue, stress, and well-being.

A consistent pattern emerges: higher noise annoyance is associated with poorer sleep quality, more fatigue, and higher stress levels. At the same time, these relationships are not strictly one-way. Being stressed or fatigued for other reasons can also lower tolerance to sound, highlighting that cities function as coupled human–environment systems rather than as collections of isolated stressors.

Reported noise annoyance (subjective) correlated significantly with measured decibels (here: Lden) – one of the many relationships between soundscape experience and sound itself we could unearth in the report.

What resonates most with me in this report is – of course – the strong signal around the role of nature in shaping soundscapes. Natural sounds such as birdsong, rustling leaves, or flowing water are systematically evaluated as more pleasant than mechanical or technical sounds. Crucially, nature does not need to drastically reduce decibel levels to have a restorative effect. Improving the quality of the soundscape already matters.

The report also aims to translate science into action. One of the most concrete ideas perhaps is the proposal for an acoustic label for housing, analogous to the energy performance certificate, alongside recommendations on acoustic renovations, building design, coordinated reporting of noise nuisance and, of course, more nature! Together, these suggestions treat acoustic quality as a fundamental component of healthy living environments rather than an afterthought.

We hope this report can really support and create actual change. Of course, such a citizen science project is about much more than just doing science, we actually want to put the topic front and center in our collective mind, and as such accelerate change.

So, give it a read, and get doing!

Posted in General | Tagged , , , , , , | Leave a comment

A hell of a nuisance

Posted on January 24, 2026 by lembrechtsjonas

Non-native species can be, pardon my words, a hell of a nuisance sometimes.

Case in point: invasive pine species in the Southern Hemisphere. Did you know that there are no (none!) native pine species south of the equator? Given how common pines are here in Western Europe, I always found that hard to fathom. But given how common they are in the Southern Hemisphere, it becomes downright mind-boggling.

For over a century, pines have been among the most widely used genera in plantation forestry across many Southern Hemisphere countries, including Chile and Argentina, but also South Africa and New Zealand. The problem is: once they’re there, they start spreading. And many types of native vegetation – such as the iconic Araucaria forests of Chilean Patagonia, or the grasslands of the Patagonian steppe – are highly vulnerable to pine invasion. You can almost see the invasion front creep forward.

The forefront of the Pine invasion, with the Lonquimay-volcano in the background

Even worse, once they’re established, these pines are devastating for native biodiversity. They grow into complex, tangled forests that are extremely hard to get through (“unleashed” pines don’t need straight lines anymore – let alone straight stems). They acidify the soils and smother out the light, resulting in near-deserts underneath their needly branches.

A dense stand of expanding lodgepole pines, with stems growing in all directions, creating an understory that is almost impossible to pass

For nature conservation in places like Patagonia, this is a nightmare. And to make matters worse, it’s pretty tough to get rid of them once they’re established. When I had the opportunity to visit Malalcahuello National Reserve in central Chile (home to the famous ‘monkey puzzle tree’ (Araucaria araucana) forests), I got to see the trouble in action. Pine removal is simply a lot of manual labour, and the end result is far from pretty.

But when they’re gone, they’re gone, right?
…Right?

Experimental sites in the Patagonian steppe before (left) and after (right) pine removal. The left picture shows the density and complexity of the pine canopy, the right picture the ‘desert’-like emptiness that remains here after pine removal.

That’s exactly what we set out to test in a paper just published in the Journal of Vegetation Science, led by our colleagues from the University of Concepción in Chile. From previous work, we already knew how bad pines are for native ecosystems: they significantly reduce the richness and abundance of native species, and cause major changes in microclimatic conditions (air and soil temperature) and soil properties (reductions in nitrogen, potassium, and pH). The big question was: do those systems recover after pine removal?

First, a little good news. Yes, we did see a recovery of microclimatic conditions to levels close to uninvaded control sites, driven by the reduction in pine canopy cover and litter. But… that’s where the good news ended.

Strong effects of pine removal (green) on microclimate conditions, as compared with pre-removal conditions (orange). Where pine canopy reduced maximum temperatures significantly, its removal resulted in consistently high maxima.

We also looked at how native understory diversity responded to pine removal, two years after the intervention. The result? It didn’t do shit. The desert remained just as deserted after pine removal as it was before – especially in the sites that had been most heavily invaded.

Understory plant species diversity parameters as a function of the ton per hectare of pine biomass that was removed, before (orange) versus after (green) pine removal for the Araucania forest site. The graph shows the complete lack of change in understory vegetation, despite the substantial change in pine canopy.

This tells us that the legacy effects of pine invasion are strong, at least up to two years after removal. If anything bounced back at all, it was pine seedlings. Native species barely benefited from the improved microclimate conditions.

Pine seedling

So is this a gloomy story?
Yeah – maybe this time it is.

But that doesn’t make it any less important. It’s crucial to know that some conservation problems are simply a pain in the ass. At least now we know, and we can keep searching for better solutions. Our paper suggests that effective management of invasive conifers must move beyond tree removal alone, and include complementary restoration actions that address persistent abiotic and biotic legacies.

Scattered pine trees in between a few old and persistent Araucaria trees

And perhaps this is, once again, a warning: if you can prevent those pines from establishing in the first place, it’s a whole lot cheaper (and far less annoying) than trying to get rid of them later.
But that’s a different story altogether. Because first we need to know which species to act on before they become annoying… and that’s simply not how we humans tend to work.

An impressive Araucaria-tree, looking out at the volcano

Reference

Fuentes-Lillo et al. (2026) Beyond Control: Short-Term Legacy Effects of Invasive Nonnative Trees May Halt Biodiversity Recovery. Journal of Vegetation Science. https://onlinelibrary.wiley.com/doi/10.1111/jvs.70110

Posted in General | Tagged , , , , , , | Leave a comment

Don’t mess with physics

Posted on December 22, 2025 by lembrechtsjonas

Gather ‘round, my friends, because this one is important! A few years ago, we released our Global maps of soil temperature – a project that brought climate data much closer to the conditions that actually matter for the organisms living in and on the soil. However, and this is the important bit: some of our maps are physically impossible.

While defying physics sounds like a big no, I’d say it’s not truly a surprise (I already wrote about similar issues in a warning here). The first version of these maps was always meant as a stepping stone, a workable product that could get our data into the hands of ecologists worldwide, and would be a step up – but not yet a leap – from what we were using up till then. But now, thanks to some careful detective work by Tomas Uxa, we can actually quantify just how our maps break one of the basic rules of heat transfer in soils. His opinion paper on the matter in Global Change Biology is thus a must-read.

Here’s the deal: heat in soils behaves predictably. One key rule is that temperature fluctuations get smaller as you go deeper. Simple, right? Not (always) so in our maps. When comparing two soil depths, an average of 26% of the grid cells – and up to 46% for certain bioclimatic variables – showed reversed patterns. In other words, deeper soils sometimes had bigger temperature swings than shallower ones. Physically impossible (and also only present in less than 5% of the raw data).

Breakdown of the physically impossible differences between the two soil layers of our global mpas, for the relevant bioclimatic variables

Why did this happen? To create our maps, we trained independent machine learning models for each soil depth. Separate models. Separate datasets. And while ML is amazing, it doesn’t inherently respect the laws of physics. The result: maps that are mostly useful, but occasionally rebellious.

Global breakdown of the number of bioclimatic variables per grid cell that is physically impossible. Note that the areas of ‘worst offence’ don’t match with the areas that are suggested to be excluded based on the extrapolation outside of the environmental space of the training data (b-c)

Uxa’s recommendation is practical: the maps are still incredibly useful, but when working with multiple depths, use each depth separately and keep this caution in mind. Any analysis that relies on comparing the two depths directly may produce physically impossible results and should be avoided.

Regional distribution of the proportion of suspicious grid cells among the seven considered soil bioclimatic variables (a) in the European Alps and (b) on the southern margin of the Congolian rainforest. Note that the mismatch closely relates to environmental complexity and steep environmental gradients, such as distinct topography changes or sharp biome transitions. The white areas represent glaciers and water bodies.

And the timing couldn’t be better. We’ve just in earnest started the work on the follow-up: Global maps of soil temperature 2.0, a new version that will hopefully solve many of the problems of its predecessor. The new version will incorporate over three times the data we had before, cover environmental variation better, and – crucially – move toward a one-model approach across depths and months. This should align our predictions more closely with physical reality. We’re also planning higher spatial resolution thanks to the explosion of computing power – so ecologists can finally get the detail that matters at scales that matter.

Up till then, a note of caution: handle our first-generation global soil maps (and really, any global maps) with care, and read Uxa’s piece to understand the quirks. And stay tuned… 2.0 promises to be a big leap forward.

Posted in General | Tagged , , , | Leave a comment

Twelve years of blogging (and counting)

Posted on December 18, 2025 by lembrechtsjonas

With the end of the year in sight, I always reflect a bit on this website and what it has achieved. Quite a shock to realize now that I’ve been blogging consistently for more than twelve years! Even better, the visitor statistics are still showing a steady increase. And yes, that genuinely makes me happy. It feels like quiet confirmation that all those hours spent writing and reflecting were not entirely in vain.

Visitor stats of The 3D Lab since its foundation in 2013 at the start of my PhD.

That feels like a good excuse for some reflection: what does it actually mean to blog as a scientist, and would I recommend it to others?

When I started the blog, I did so more or less on day one of my PhD. Unsurprisingly, there wasn’t much research output to talk about yet – because, well, the research still had to happen. In those early years, the blog had much more of a travel-magazine vibe. It was full of fieldwork stories and photos from the mountains, documenting places and adventures rather than results.

Slowly but surely, those travels turned into datasets, and those datasets turned into papers. At that point, the blog evolved with them. I made it a habit – also from day one – to write a blogpost about every paper on which I was a co-author. I often get the question where I find the time to do that, and honestly: it is often hard to squeeze into an already full schedule. But my thoughts have always been simple: if I can’t find two hours or so to summarize the key findings of a paper, why did I even bother contributing to it in the first place?

For me, blogging about published papers thus always – foolishly perhaps – trumps writing new ones. It’s a way of celebrating past wins before chasing the next deadline. Otherwise, it’s very easy to get stuck in a rat race where papers pile up but joy may disappear.

As my network grew, the audience of the blog grew with it. And I think providing this extra layer on top of a scientific paper – where I highlight what I find important – helps get the message across more clearly and, I hope, more convincingly. It reduces the risk of papers being forgotten and buried under the ever-growing avalanche of scientific output. Just as importantly, it helps reinforce the idea that each individual paper has value – to me, and to others – rather than being just another brick in a publication wall.

The blog has also given me space to tell stories that don’t fit neatly into the standard scientific paper format (like this one). Several of those posts have been among the most read. Early on in the ChatGPT hype, for instance, I wrote a post about how I perceived its pros and cons for research. Another well-visited piece dives into the surprisingly non-trivial question of which microclimate sensor to choose, a post I still occassionally share with people that ask me that question.

It’s also been an excellent platform to publicize new global research networks and invite people to join. That worked particularly well for initiatives like SoilTemp (now MEB) and EcoFracNet. One post I still hear about is the one where we offered free microclimate sensors to be installed in as remote places as possible, to help fill gaps in our global database. That kind of outreach simply doesn’t fit anywhere else as naturally.

And yes, the blog still occasionally features fieldwork photos and travel reports – although I’m definitely less of a nature photographer now than I was during my PhD.

So, would I recommend blogging to other scientists?

In general: yes, if you’re comfortable with it being a slightly slower medium. Blogging works best when combined with other forms of outreach; otherwise it’s hard to get the word out. Twitter used to be ideal for that, and while I never really managed to recreate that dynamic elsewhere, LinkedIn has become the most workable alternative for now. It’s not the same, though.

One big advantage is that writing blogposts is surprisingly therapeutic. It forces you to slow down and really think about your published work. Even without a large audience, that process alone helps sharpen your understanding of the bigger picture you’re contributing to. You could post the same insights directly on social media, but to me that always feels fleeting.

So yes, I’ll keep on blogging. If only to keep forcing myself to think about what I want to tell people, and why I’m doing what I’m doing in the first place. And I hope you’ll stick around, keep reading, and occasionally stumble upon a useful nugget or two along the way.

Posted in General | Tagged , , , , | Leave a comment

Sarane rocks!

Posted on December 18, 2025 by lembrechtsjonas

You wouldn’t guess it from the amount of work she has already put into getting the global MIREN Rocks network off the ground and moving forward, but Sarane is only at the very beginning of her scientific trajectory. That trajectory just received a major boost: Sarane has been awarded an FNRS FRIA doctoral grant, allowing her to start a PhD on her beloved rock cliffs and their ecology.

This is a fantastic achievement for her, and wonderful news for us as a team. It gives us the certainty that MIREN Rocks can now deliver much more in-depth science than would have been possible without her central role.

In her PhD, co-supervised by Alain Vanderpoorten (University of Liège) and myself (Utrecht University), and in close collaboration with Koenraad Van Meerbeek (KU Leuven), Sarane will work with the global vegetation database collected by MIREN Rocks partners worldwide. In parallel, she will zoom in on the mechanisms shaping rock vegetation in the Meuse valley (Wallonia). There, we will develop high-resolution microclimate models using in-situ sensors and drone-derived digital surface models, and set up a split-plot experiment to assess the role of physical (vegetation removal) and chemical (climbing chalk) disturbance on cliff vegetation.

And this is only the beginning – there is much more in the pipeline for this PhD and for the global MIREN Rocks network. So if you love cliffs even one tenth as much as Sarane does, I strongly recommend following her work.

The best place to do so? Instagram, via @mirenrocks.

Congrats again, Sarane, and looking forward to work together further the coming years!

Posted in Belgium | Tagged , , | Leave a comment