Influence of land use and host species on parasite richness, prevalence and co-infection patterns

Highlights

• Human land use affects blood parasite diversity, prevalence and intra-host parasite interactions.

• Plasmodium and Haemoproteus parasites are more abundant in degraded habitats.

• Leucocytozoon parasites are preferentially found in natural forested areas.

• We found large variations in parasite diversity and prevalence among bird species.

Abstract

Tropical forests are experiencing increasing impacts from a multitude of anthropogenic activities such as logging and conversion to agricultural use. These perturbations are expected to have strong impacts on ecological interactions and on the transmission dynamics of infectious diseases. To date, no clear picture of the effects of deforestation on vector-borne disease transmission has emerged. This is associated with the challenge of studying complex systems where many vertebrate hosts and vectors co-exist. To overcome this problem, we focused on an innately simplified system – a small oceanic island (São Tomé, Gulf of Guinea). We analyzed the impacts of human land-use on host-parasite interactions by sampling the bird community (1735 samples from 30 species) in natural and anthropogenic land use at different elevations, and screened individuals for haemosporidian parasites from three genera (Plasmodium, Haemoproteus, Leucocytozoon). Overall, Plasmodium had the highest richness but the lowest prevalence, while Leucocytozoon diversity was the lowest despite having the highest prevalence. Interestingly, co-infections (i.e. intra-host diversity) involved primarily Leucocytozoon lineages (95%). We also found marked differences between bird species and habitats. Some bird species showed low prevalence but harbored high diversity of parasites, while others showed high prevalence but were infected with fewer lineages. These infection dynamics are most likely driven by host specificity of parasites and intrinsic characteristics of hosts. In addition, Plasmodium was more abundant in disturbed habitats and at lower elevations, while Leucocytozoon was more prevalent in forest areas and at higher elevations. These results likely reflect the ecological requirements of their vectors: mosquitoes and black flies, respectively.

Introduction

In a very short time span, human-led changes such as habitat destruction and climate change have caused large disturbances in most natural ecosystems across the planet (Creutzig et al., 2019, He and Silliman, 2019). Tropical forests, which account for one-third of land-surface productivity and evapotranspiration (Malhi, 2012), are estimated to host over half of all global terrestrial biodiversity (Pimm and Raven, 2000) and are key to ensure many of the planet’s vital functions. Since the second half of the past century, these ecosystems have been experiencing increasing impacts of multiple human activities such as conversion to agricultural use (Laurance et al., 2014), logging and other resource extraction (Malhi et al., 2014), fires (Silva et al., 2018), introduction of invasive species (Liebhold et al., 2017), and hunting and wildlife trade (Symes et al., 2018). These activities are expected to disrupt ecological interactions and greatly influence the dynamics of transmission of infectious diseases (Sehgal, 2010, Faust et al., 2018).

There have been many examples of direct effects of deforestation on human pathogens (e.g. malaria: Yanoviak et al., 2006, Yasuoka and Levins, 2007, Burkett-Cadena and Vittor, 2018) and wildlife parasites (Bonneaud et al., 2009, Chasar et al., 2009, Tchoumbou et al., 2020). Increased levels of sunlight, associated with open spaces, have been correlated with increased mosquito densities and feeding rates of mosquitoes on humans, and increased parasite transmission (Vittor et al., 2006, Vittor et al., 2009, Loaiza et al., 2017). However, some studies have also reported no obvious link between deforestation and the incidence of vector-borne diseases (e.g. malaria: Tucker Lima et al., 2017). These conflicting findings are likely attributable to the complexity of vector-borne disease systems, which involve highly diverse groups of vectors, hosts and pathogens, and depend on the study region and associated socioecological factors (MacDonald and Mordecai, 2019). Oceanic islands can therefore be used as simplified study systems to understand these highly complex systems. Compared with their mainland counterparts, islands have lower species diversity and numbers of available habitats (MacArthur and Wilson, 1967), leading to a simplification of many aspects of ecology and making them ideal natural laboratories (Losos and Ricklefs, 2009). Several studies from different archipelagoes across the world have shown lower diversity and prevalence of parasites in insular bird populations compared with the mainland (Beadell et al., 2007, Svensson and Ricklefs, 2009, Pérez-Rodríguez et al., 2013, Loiseau et al., 2017).

Avian malaria and other haemosporidian parasites have been found in all regions except Antarctica and are one of the best-studied groups of blood parasites (Valkiūnas, 2005, LaPointe et al., 2012). They have been used as a system to study the numerous abiotic and biotic factors that affect the distribution and transmission of pathogens among different geographical locations, from the tropics to the poles (Loiseau et al., 2012, Ellis et al., 2019, Harvey and Voelker, 2019). Environmental degradations such as urbanization (Carbó-Ramírez et al., 2017, Hernández-Lara et al., 2017, Ferraguti et al., 2018), habitat fragmentation (Belo et al., 2011, Laurance et al., 2013, Pérez-Rodríguez et al., 2018) or deforestation (Chasar et al., 2009, Tchoumbou et al., 2020) have contrasting effects, showing strong to mild impacts on avian haemosporidian prevalence, diversity or specificity. Besides the effects of environmental degradation, several studies have also suggested that elevation (LaPointe et al., 2010, Zamora-Vilchis et al., 2012, Harrigan et al., 2014, Illera et al., 2017), temperature and precipitation (Paaijmans et al., 2009, Gonzalez-Quevedo et al., 2014) have a major effect on the development, abundance and dispersion of parasites, and on the development of vectors (Lardeux et al., 2008, Beck-Johnson et al., 2013). Avian haemosporidians require an arthropod vector to complete their life cycles (Valkiūnas, 2005): Plasmodium is transmitted by mosquitoes (Culicidae), Haemoproteus by biting midges and louse flies (Ceratopogonidae and Hippoboscidae, respectively) and Leucocytozoon by black flies (Simulidae). Each of these vectors has different ecological requirements and is likely to have distinct responses to environmental changes. As an example, a recent study found that biting midges were more abundant in forested areas than in disturbed areas, while mosquitoes showed the opposite pattern (Loaiza et al., 2017), but mosquito abundance has also been found to be greater in natural and rural areas than in urban areas (Ferraguti et al., 2016).

In this study, the main objective was to investigate how anthropogenic forest degradation, mostly linked to agricultural activity, influences the prevalence, distribution, richness and co-infections of avian haemosporidian parasites, using an oceanic island as a study model: São Tomé Island, in the Gulf of Guinea, Central Africa. São Tomé harbors biodiversity of global conservation importance, mostly due to a high degree of endemism across many taxa (Jones, 1994). The country’s natural parks have been classified as the 17th most irreplaceable site, from a list of over 175,000 protected areas worldwide (Le Saout et al., 2013). The Gulf of Guinea oceanic islands have the planet’s highest concentration of endemic bird species: São Tomé alone having 17 single island endemic species and three endemic species shared with nearby oceanic islands (Stattersfield et al., 1998; Melo, 2007. Bird speciation in the Gulf of Guinea Island system. PhD Thesis, University of Edinburgh, Edinburgh, UK). As a result, the forests of São Tomé have been identified as the third most important forests for the conservation of birds (Buchanan et al., 2011). There is a gradient of anthropogenic disturbance within the island, from native forests, most of which are in the protected area and under little human influence, to human-dominated agricultural land use. One-third of the island is still covered by native forests, another third by secondary forests, resulting mostly from abandonment of agricultural use, and the remaining third by agricultural land (Soares et al., 2020). The most altered areas are non-forested lands such as the artificial open savannas in the drier north and the recently expanded oil palm monoculture in the south. Previous studies on São Tomé Island, supported by a long-term bird monitoring scheme, created extensive knowledge of the avian blood parasite communities, which showed that despite being an oceanic island, it has an important diversity of Plasmodium, Haemoproteus and Leucocytozoon haemosporidian parasites (Loiseau et al., 2017), in contrast to other islands and archipelagoes such as the Hawaiian Islands, Lesser Antilles and Macaronesia (Hellgren et al., 2011, Spurgin et al., 2012, Clark et al., 2014, Ricklefs et al., 2016, Padilla et al., 2017). Having simplified ecosystems, a gradient of human influence, ongoing long-term bird monitoring efforts and a solid baseline knowledge of the parasite communities, makes São Tomé Island an ideal system to study ecological interactions between hosts and parasites, and how those interactions are affected by human land use.

At the bird community level, but also at the species level for the five most abundant bird species, we evaluated: (i) parasite richness of Plasmodium, Haemoproteus and Leucocytozoon, and their phylogenetic relationships; (ii) associations between infection status (uninfected versus infected), and land use and elevation, (iii) the influence of land use and elevation on co-infections (i.e. intra-host parasite richness). Co-infections are rarely studied in this parasite system because double peaks on chromatograms hamper identification (Loiseau et al., 2010, Harrigan et al., 2014), but can critically impact infection dynamics, virulence and host fitness (Marzal et al., 2008, Palinauskas et al., 2011). In our study, we devoted particular efforts to disentangle co-infections, to understand how they are distributed in the function of forest degradation and if some genera or lineages are more often found in positive or negative association with other parasites, to evaluate facilitation or competition dynamics (Eswarappa et al., 2012). We predicted that habitat and elevation variables would have marked impacts on the avian parasite community, inducing a distinct response for each heamosporidian genus (Chasar et al., 2009). Specifically, we predicted that Plasmodium diversity and prevalence should be higher in degraded habitats, while Haemoproteus and Leucocytozoon diversity and prevalence should be higher in forested areas (Loiseau et al., 2017, Loaiza et al., 2019). At the species level, we predicted that birds exhibiting high parasite diversity would also show higher prevalence, as we could expect both parameters to be influenced by characteristics that are intrinsic to each bird species, such as their immune system and parasite resistance (Ilgūnas et al., 2019), and the probability of encounter with vectors and parasites (Malmqvist et al., 2004, Kilpatrick et al., 2006).