Why fieldwork is still crucial for science research

There are some things it's impossible to discern without ground truthing

Cassie Freund


Wake Forest University

The summer fieldwork season is upon us, and I’ve spent the majority of of the past couple of months camped on the side of a mountain in the Peruvian Andes. My research team and I are here collecting data on plant communities that are slowly recolonizing the hundreds of landslides dotting my study site in and around the mega-diverse Manu National Park.

Our work is adventurous and exhilarating. It's also physically and mentally demanding, as we work from dawn to dusk at high altitudes and in all weather conditions. Cooking over a small camp stove gets pretty old after a week. After about two, dry socks become the ultimate creature comfort. 

I'm not complaining: knowing that each year I get to spend several months exploring tropical jungles is part of why I became an ecologist and conservation biologist.

Conservation scientists are an increasingly diverse group of people (although, we still have work to do on that front) and it's impossible to neatly summarize why we are so driven to get our hands dirty in the field. But I feel comfortable positing one reason: the species and ecosystems that we study, which have immeasurable intrinsic, ecological, and even economic value, are rapidly disappearing.

Despite the fact that humans have set foot on almost every terrestrial corner of this planet, we still lack basic information about the range of other life forms that share it with us. We still don't even know to the nearest million how many species there are. The only way to answer this question is further exploration, and the good news is that there's no shortage of scientists eager to plumb the depths of the ocean, wade through swamps, and drop into rainforest canyons to find out.

This motivates my own research in the vast forests of the Peruvian Andes, where I study how landslides affect these mountain jungles. My research team and I explore landslides at different stages of recovery, painstakingly identifying and measuring each individual plant that we capture in our study plots. It's tedious business, but half of my dissertation depends on this detailed data collected on the ground. 

But these days we can also gather incredibly detailed information about Earth's ecosystems remotely, extract information using machine learning, and analyze all of it using increasingly complex statistical methods. These advancements are revolutionizing biodiversity conservation – I'm slowly reconciling my passion for getting my hands dirty in the rainforest with the realization that fieldwork is no longer the only cornerstone holding up the house of inquiry in my field. ​

This is why a ​Global Ecology and Conservation study published in April, led by Dr. Carlos Antonio Ríos-Saldaña, director of a Mexico-based biodiversity nonprofit called BioCorima, caught my attention. The authors asked whether fieldwork-based studies were losing their status among conservation scientists in favor of modelling and data synthesis studies. Their research was motivated by the fact that fine-scale data on where plants and animals live and the impact of human activity on their populations - the type of data that we get from fieldwork - is integral for environmental policy makers. This is especially true in rapidly developing countries, many of which are exceptionally rich in biodiversity and actively wrestling with issues of sustainable development.

Ríos-Saldaña and colleagues were concerned by what their analysis showed: not only is field research becoming less popular among conservation scientists than desk-based modelling and data synthesis work; it is also less likely to be published in the most prestigious conservation journals, relegated instead to smaller journals that people are less likely to read. The latter point is particularly disheartening for field biologists. Publications in high-profile journals are – for better or worse – still the ultimate scientific currency, and a rejected article stings especially sharply after you have shed actual blood, sweat, and tears to collect your data.

According to their analysis of published research from 43 conservation journals, Ríos-Saldaña and his colleagues found that the frequency of papers based on field studies has decreased by 20 percent since 1980, while the prevalence of modelling and data synthesis studies has jumped by 600 and 800 percent, respectively. They also found that just 55 percent of the papers they analyzed from top-tier conservation journals, like Conservation Biology and Conservation Letters, included a fieldwork component, compared to 93 percent of papers from lower-impact journals such as Southeastern Naturalist and the Caribbean Journal of Science.

But I'm not convinced by their results. The team's conclusions are based on a random selection of 386 papers from a possible 59,575 articles published during the study period of 1980-2014. That's just 11 papers per year, not a sample size that I would be comfortable drawing such major conclusions from. I'm not even sure that counting publications is the best way to quantify how many scientists are doing fieldwork. And, it's not clear to me how studies that used multiple approaches (e.g. both fieldwork and modelling) were categorized. These methods aren't mutually exclusive: often researchers will collect data on some phenomenon of interest in the field to build a mathematical model that explains it, or vice versa.

Ultimately, I’m not sure this distinction between field and modelling work is even relevant: it’s plain to me that the future of conservation science integrates them both. There are some scientific questions that can't be answered with field data alone. My dissertation research is a good example of this: to understand the role of natural disturbances in mountain forests across the Peruvian Andes, I need a way to look at the entire landscape. So, in addition to data collected in the field, a substantial portion of my dissertation hinges on information gathered with a small drone, satellites, and airplane-mounted lasers (a mapping technology called LiDAR). From these remotely sensed data, I plan to calculate the amount of stored carbon dioxide that is lost from the forest when trees are swept away in landslides, as well as how quickly the trees – and therefore a large part of the carbon storage of the ecosystem – recover from these catastrophes.

These same data-gathering methods are also incredibly powerful tools for conservation. For example, it is now possible to detect and monitor tropical deforestation in real time, a much-needed capability given that 2017 was the second-worst year for tropical forests in recorded history. I hope to use the results of my research to model the impact of landslides on forest carbon storage across the Andes to better predict what might happen to this ecosystem service under climate change. 

But remote sensing tools can never replace fieldwork. We need to look at ecosystems with our own eyes – what scientists call ground truthing – to make sense of what we see from the air, to appreciate the intricacies of plant and animal life that satellites simply cannot capture, and to detect early warning signs of biodiversity loss.

Although I question whether Ríos-Saldaña and colleagues convincingly showed that fieldwork-based studies are declining, they raise an important argument: funding and publishing such research should still be a priority for the scientific community. Without field science, we would never know that there is a species of ant in Brunei that literally explodes when threatened, or that hippos sometimes cause massive fish die-offs in East Africa. We also would never have discovered life-saving anti-cancer drugs or the fossils that shed light on the origin and history of our species.

I've come to see that the most important question about research methods in conservation science isn't necessarily whether fieldwork is on the downswing, or if we have too many modelling studies. This just distracts us from our mission. Nature is rapidly disappearing, and we must take advantage of every method at our disposal to save it. The question we should all be concerned with answering is whether the methods that we employ to study and protect nature are appropriate for our goals. 

Biodiversity conservation is an enormous and vital scientific field. Every bit of data matters in the fight to protect species from extinction and maintain Earth's magnificent ecosystems, which is why I’m confident that all of the effort my team and I are putting into our work is worth it. There’s no better feeling - especially when it is paired with copious amounts of chocolate, Andean cheese, and a stockpile of dry socks - to fuel a few more weeks in my tent on the side of a mountain.

Comment Peer Commentary

We ask other scientists from our Consortium to respond to articles with commentary from their expert perspective.

Julia York

Polar Science

University of Texas at Austin

Fieldwork is also extremely important when studying animals. Studying an animal in its environment really should be the primary way studies are conducted. Only when you need a controlled environment to answer a specific question should the animal be brought into the lab. If I want to study you and how you live your life I’m going to get very different answers if I observe you in your house rather than bringing you to my lab and keeping you in a box. That is not different for other animals.

If fieldwork studies are declining, I think it is because of the cost and risk, which makes many PIs afraid to invest their limited grant money in them. We should reward field work studies by publishing them in high-impact journals to balance the risk with reward.