Posted by Cléber Rodrigo de Souza

Every year Methods in Ecology and Evolution rewards Robert May Award in the best paper published in the journal by an author early in their career. Ten Early Career Researchers entered the shortlist for this year’s award, including Cléber Rodrigo de Souza who is a data scientist at the Federal University of Lavra in Brazil. In this interview, Cléber shares knowledge about their letter ‘Structure and diversity of tropical forests: a comparison of methodological choices‘.

Tell us about your career stage, research, hobbies and interests

My name is Cléber Rodrigo de Souza, I am a Brazilian forest engineer and I recently completed my PhD in forest sciences. Throughout my academic career, I have worked to understand the role of the environment and historical heritage in forest structure, dynamics, and phytogeography, with a focus on often-overlooked arid tropical forests.

I am still working to address this issue, while also working as a data scientist in an intelligence and innovation agency at the Federal University of Lavra (Zetta Agency). The Innovation Agency provides services to Brazilian and foreign government agencies by providing technological solutions to a range of problems. Research is an important part of my life and is closely linked to my daily routine and hobbies. In my free time, I am usually in touch with nature, through walking, cycling and swimming.

Cléber in the mountains where some of the data were collected.

How would you send your article to someone if you only had 30 seconds in the elevator?

Measuring the diameter of a tree in a semi-deciduous forest.

Tropical forests play a crucial role in mitigating current and projected climate change. However, to fully appreciate the importance of tropical forests in the global carbon cycle, it is vital to assess carbon gains, reserves and species diversity within a given area. Addressing this issue is particularly important in bypassed seasonal rainforests (deciduous and semi-deciduous) which are often found in drier environments than rainforests.

I found that the application of widely used protocols for assessing the function and diversity of tropical forests were ineffective in estimating the same values ​​in arid tropical forests and yielded inaccurate data. This may be because many species in these more seasonal forests do not grow to the same size as the species used in tropical forest sampling procedures. Our findings encourage other researchers to consider key ecosystem factors when applying standard procedures to assess forest structure, dynamics, and diversity.

Where did the idea to develop this method come from?

A multi-stemmed tree in a semi-deciduous forest.

The idea for this work was born in the most natural place for forest ecologists to conceive ideas… during fieldwork (especially when monitoring a seasonal semi-deciduous forest). It is important to note that the research team was undergoing significant changes in the way we handled data collection, largely driven by the need to include our plot data in a collaborative research network called So at this point in time, we were adapting our local methods of collecting, recording, and storing data to meet collaborative networking standards.

As there were no written protocols for seasonal dry forests, it was recommended that we use the general procedure for tropical forests. We implemented sampling procedures from grid standards (such as the amount of lianas and estimating the position of the tree canopy) with our local standards (such as minimum inclusion size and inclusion method). However, we noticed discrepancies between the two data sets. For example, a tree with 25 isolated stems all with a diameter <5 cm at chest height would be included in our local methodology but not in the collaborative network methodology. This was a problem because a tree of this size is ecologically representative of the area in which many of the trees were smaller in diameter with a larger number of stalks.

At this point, we talked about all the possible data lost using widely adopted tropical forest methodologies in a dry forest ecosystem, where many trees are generally smaller and with many stems. Most importantly, we wondered how this data loss affected our understanding of vital ecosystem functions. It was then that we first began to discuss how to adapt the widely used rainforest protocols to better measure the structure and dynamics of arid forests.

What were the main challenges in developing this method? How did you overcome this?

Once we decided to conduct this study, we began to search the literature for the basic basis of the methodological choices used in data collection in tropical forests. For example, we wanted to understand the reasons for using a certain criterion, such as the widely accepted minimum stalk diameter of 10 cm in tropical forests. We wonder what were the specific reasons for using this widely applied threshold? Why for example 10 cm in diameter and not 9 or 11 cm?

At this point we faced the first difficulty, it was a very limited job dealing with this specific issue. Thus, the definition of methodologies was more the result of a repetition of methods used by researchers and pioneer groups throughout history, rather than following strong quantitative criteria and methods. In this sense, developing our analysis with limited theoretical foundations was indeed a challenge.

Cléber and supervisor Prof. Rubens Santos in a dry forest above the limestone

We also faced difficulties in developing a quantitative approach related to methodological choices, which are considered discreetly and inconsistently. For example, the minimum size 10 or 5 cm and the method of inclusion according to the tree or stalk. What are the consequences of these choices and should we avoid using a simple category comparison approach? It was then that we decided to simulate scenarios of these distinct methodological approaches, simulating two variations of the inclusion method and a wide range of minimum inclusion size to quantify the variables of tropical forest function and diversity.

Who will benefit from your method?

These methods have direct benefits for research projects in arid tropical forests, informing their data collection and leading to methodological choices that are more appropriate to the ecology of these arid ecosystems. Our results have helped modify the collection techniques used by our partner groups in Brazil, who also work with this type of forest, and we expect this to extend to research groups working with dry forests around the world.

This will give a broader understanding of the diversity of these forests, which can help conservation policies and strategies that are more appropriate for their ecological space. Thus, the ecosystems of arid tropical forests and the many endangered species living within them will benefit from our study.

Cléber from a Cavanillesia umbrella tree

If you could travel back in time, would you add or change anything about your method?

We used an innovative approach to develop our technique, which was challenged by the inadequacy of the accompanying literature. However, if we were to repeat the study, it would be interesting to include a larger number of sites and expand to other very important types of vegetation, such as fire-induced tropical savannas. So if I could use a time machine, I would try to find collaborators working with data from other countries and types of vegetation.

You can read Cléber’s full letter here

and learn more about the Robert May 2022 shortlist here.

Leave a Reply

Your email address will not be published.