How do deforestation and invasive species alter food webs in tropical rainforests? Insights from diet DNA metabarcoding on Sao Tome and Principe Islands.


Tropical biodiversity is deeply threatened by land-use changes caused by human activities [1]. The loss and degradation of forest is a powerful predictor of local species loss in many tropical forests [1]. Agricultural practices, such as production of palm oil, cacao and coffee, drive much of this loss, as they fundamentally alter forest structure [2]. Changes in abiotic conditions such as light intensity, temperature and humidity can also profoundly alter the habitat available to native species [3].

Forest loss and degradation often lead to the homogenisation of vegetation, which can in turn affect the temporal and spatial availability of resources (e.g., flowers, fruits, leaves and seeds) for the associated fauna, often resulting in the disappearance of the more specialist and sensitive native species, leading to further biotic homogenisation [5]. At the same time, human activities associated with deforestation and habitat fragmentation introduce exotic and often invasive plant and animal species. Invasive species may lead to additional erosion of native species, either directly, by for instance increasing predation pressure, or indirectly by competing for shared resources, which could further contribute to changes in functional diversity. Invasive species are also a burden for local economies, such as rats and pest insect damage on shade cocoa farms [6].

Despite important progress in understanding the influence of deforestation and biological invasion on biodiversity loss, comparatively little is known about how these processes impact the interactions between different species, and the implications such altered interactions may have for biodiversity and ecosystem function [8]. Historically, the challenge of building accurate ecological networks has been the difficulty of obtaining detailed datasets across different trophic layers; without such data, the effect of invasive species and forest loss/degradation on ecosystems cannot be properly understood [8]. However, the recent development of high-throughput DNA metabarcoding has overcome this issue [9], and this technique is being increasingly deployed in biodiversity network analyses [10].
This project will assess the role that native and introduced fauna have in controlling invasive pests, particularly in commercial plantations such as cocoa farms. The project will take place in Sao Tome and Principe, two close tropical oceanic islands in central Africa. These islands hold an extraordinarily high number of endemic species across several taxonomic groups, but also many introduced species, notably birds, mammals and plants [11]. Here, rats, palm civets, mona monkeys and other introduced species have profoundly changed a native ecosystem historically poor on terrestrial mammals. However, the relationships between native and introduced species, and the effects that such relationships may have on biodiversity and local economies, are poorly understood.
The project is a collaboration between Dr Davide Dominoni (Glasgow), an ecologist interested in the effects of human disturbance and urbanisation on animal behavior and biodiversity, Dr Andreanna Welch (Durham), a molecular ecologist with ample experience on DNA metabarcoding and with a team of researchers in Portugal (Dr Luke L. Powell, Dr Martim Melo and Dr Ricardo F de Lima) with strong roots in the study of tropical biodiversity in Africa. They have long-standing and ongoing links with Sao Tome and Principe, including on-ground support during field work.


The project will combine intensive field work along disturbance gradients in the islands of Sao Tome and Principe with advanced DNA sequencing approaches to answer three fundamental questions.

1. How does biodiversity of plants, arthropods, birds and mammals change along gradients of human disturbance?

(a) Conduct field work to collect data on the abundance and diversity of these key taxa.
(b) Using state-of-the-art statistical analyses, model the relationship between human disturbance, forest loss and diversity of such taxa.

2. How do introduced species affect ecological networks?

(a) Using plant, arthropod and vertebrate samples (esp. bird and mammal faeces), perform DNA metabarcoding to understand the diet of vertebrates along disturbance gradients.
(b) Using DNA metabarcoding data, build ecological community network models to understand how human disturbance and invasive species affect ecological relationships.
(c) Perform sensitivity analyses to understand the consequences of losing or adding an exotic or native animal from the network.

3. What vertebrate species control the population of invasive pests of crop plantations?

(a) DNA metabarcoding of mammal and bird samples can also be used to detect which species may control the population dynamics of invasive pests, such as rats and insects.
(b) We will investigate whether this is a bottom-up process (vegetation and habitat quality control rats and other pests) or a top-down process (introduced civets and native owls predate and control pests).

Project Timeline

Year 1

Kick-off meeting, literature review, field work in Sao Tome and Principe to collect plant, arthropod and vertebrate samples, attendance of workshops on advanced statistics and bioinformatics. Lab work for DNA metabarcoding

Year 2

Field work as in year 1, analysis and write up for objective 1, attend courses on ecological network modelling and scientific writing; Lab work for DNA metabarcoding and initiate analyses for objectives 2 and 3.

Year 3

Finalise analyses and write up of papers for objectives 2 and 3, attendance of scientific meeting.

Year 3.5

Attendance of international scientific meeting, completion/revision of manuscripts and submission of thesis.

& Skills

The training provided by this project will cover a broad suite of important skills ideal for the future ecologist. It will provide the student with the opportunity to conduct field work in an area of outstanding natural diversity, learning various field techniques such as transects, point counts, trapping of birds and mammals, vegetation surveys and arthropod sampling. In the lab, he or she will learn cutting-edge DNA sequencing skills that can provide a strong foundation for a research career in ecological and evolutionary genetics, but also for careers in environmental testing, molecular biology, agricultural biotechnology, and medical testing/research. Parallel to intensive field work and lab work, the student will also be exposed to quantitative (ecological network modelling) and computational approaches (bioinformatics for DNA metabarcoding data) and their application. This will provide the student with an extremely strong basis for pursuing independent research in their field(s) of interest or for a transition to roles in growth areas such as data science and ecological consulting.

The scholar will be based within IBAHCM under the supervision of Dr Davide Dominoni. At IBAHCM the student will be exposed to a vibrant research environment that takes pride in its strong collaborative approach to scientific research. Several special interest groups are present in the institute that will help the student develop theoretical knowledge and practical skills, including the Avian Biology group, the Spatial Ecology group, the Diversity group and the Statistical Ecology group.

In year 1, the student will receive training on GIS, management of large datasets, advanced used of R and Bayesian statistics, by following dedicated courses offered within IBAHCM, by attending external specific workshops and courses, and through the support of the supervisors (DD and LP in particular). In year 2, the student will develop skills in DNA metabarcoding by spending time with co-supervisor Welch in Durham, and on ecological network and biodiversity modelling by spending time with collaborator Powell in Portugal. The student will also join retreat sessions on scientific writing and will start to lay out the first manuscript. In year 3, through participation in Institute seminars and national and international conferences, they will also develop presentation and communication skills. In year 3.5 they will receive training in grant-writing and interview skills in preparation for job applications.

References & further reading

[1] Alroy, J. Effects of habitat disturbance on tropical forest biodiversity. Proc. Natl Acad. Sci., 114, 6056– 6061 (2017).

[2] Edwards DP, et al., Wildlife-friendly oil palm plantations fail to protect biodiversity effectively. Conserv. Lett. 3, 236 (2010).

[3] Murcia, C. Edge effects in fragmented forests: implications for conservation. Trends Ecol. Evol. 10, 58–62 (1995).

[4] Marconi L and Armengot L. Complex agroforestry systems against biotic homogenization: the case of plants in the herbaceous stratum of cocoa production systems. Agric. Ecosyst. Environ., 287 (2020).

[5] Menezes Pinto Í, Emer C, Cazetta E and Morante-Filho JC. Deforestation Simplifies Understory Bird Seed-Dispersal Networks in Human-Modified Landscapes. Front. Ecol. Evol. 9:640210 (2021).

[6] Thibault M, Vidal E, Potter MA, Sanchez T, Brescia
F. The invasive Red-vented bulbul (Pycnonotus cafer)
outcompetes native birds in a tropical biodiversity
hotspot. PLoS ONE 13(2): e0192249 (2018).

[7] Heleno, RH, Mendes F, Coelho AP, Ramos JA,
Palmeirim JM, Rainho A, de Lima, RF. The upsizing of
the São Tomé seed dispersal network by introduced
animals. Oikos 00: 1–12 (2021).

[8] Hagen M, Kissling WD, Rasmussen C, Marcus
AM, De Aguiar MAM, Brown LE et al. (2012)
Biodiversity, species interactions and ecological
networks in a fragmented world. Adv. Ecol. Res, 46,
89– 210 (2012).

[9] Vacher C et al. Learning ecological networks from
next-generation sequencing data. Adv. Ecol. Res., 54
pp. 1-39 (2016).

[10] Mollot G, Duyck P-F, Lefeuvre P, Lescourret F,
Martin J-F, et al. Cover Cropping Alters the
Diet of Arthropods in a Banana Plantation: A
Metabarcoding Approach. PLoS ONE 9(4): e93740

[11] Soares et al Land-use intensification promotes
non-native species in a tropical island bird assemblage
Anim. Conserv 23 573-584 (2020).

Further Information

This project is in competition with others for funding, and success will depend on the quality of applicants. Funding includes tuition fee waiver for Glasgow University, a competitive stipend, and research support. To express interest please first contact Dr Davide Dominoni ( by early January 2022.

Apply Now