Zoonotic diseases, also known as infectious diseases transmitted from animals to humans [1], are a significant global concern [2]. These diseases result from the transmission of diverse microorganisms through direct contact, consumption of contaminated substances (i.e., food, water), or exposure to animals (vectors, hosts, and reservoirs) [3]. With zoonotic diseases accounting for 75% of known emerging pathogens [4], they have become a pressing issue with significant implications for public health. Moreover, the increasing trends in zoonotic diseases are influenced by various factors, including population growth, deforestation, land use and climate change, and globalization [5,6]. Understanding and effectively addressing zoonotic diseases are paramount, as they not only cause millions of deaths each year [7] but also result in substantial economic damages. Therefore, comprehensive knowledge and robust control measures are vital to mitigate the risks associated with zoonotic disease emergence and transmission [2, 5,6] making this topic of utmost societal importance.
Landscape ecology plays a crucial role in understanding the relationship between landscape characteristics and infectious disease dynamics in specific regions, communities, and species-specific contexts [8,9]. Both composition and configuration aspects can determine host presence and abundance, transmission rates, disease prevalence, and consequently spatiotemporal patterns of diseases [10,11]. Consequently, recent studies have shown how land use type, habitat fragmentation, and hydrological aspects affect disease dynamics and transmission [12,13]. For instance, habitat fragmentation and human encroachment resulting from the loss of forested areas in Africa have been associated with increased opportunities for the transmission of the Ebola virus from animal reservoirs to humans [14,15]. Similarly, in both the Amazon Basin and Southeast Asia, the loss of natural forest and their conversion to agricultural plantations have been linked to the exacerbated transmission of vectorborne diseases such as malaria and dengue fever [16,17]. This happens because deforestation, land use change, and habitat simplification promote generalist and host species, increasing their abundance and the risk of zoonotic pathogen transmission [6,[18][19][20][21]. However, regional context and species-specific factors further shape this landscape-disease relationship [12,22,23]. Understanding these relationships through a landscape ecology lens contributes to more effective disease management strategies [24,25] and allow better land use planning, to ensure the establishment of landscapes with a low pathogenicity [26].
The Tropical America, spanning from Mexico to southern South America [27], is recognized as the most diverse region in the world, and consequently, harbors a high diversity of pathogens [28]. It also faces the higher rates of land use change (i.e., loss of native vegetation) of the globe [29], which makes several parts of this region hotspots for future emerging infectious diseases [30,31]. Understanding how landscape changes influence the diversity, trends, and patterns of zoonoses is of paramount importance due to their potential economic, environmental, and public health implications globally [5,32]. However, our current knowledge about the mechanisms through which land-use change influences host and pathogen communities and the risk of cross-species transmissions is limited, especially in the Tropical America region [33]. These gaps in knowledge need to be identified and are critical as they hinder our ability to develop effective strategies for disease prevention and control [32,33]. A focused investigation into the mechanisms underlying the impact of land-use change on zoonotic disease dynamics in the Tropical America is essential to fill these knowledge gaps and pave the way for informed decision-making and targeted interventions [9,32,33].
In this synthesis we aim to fill this research gap and assess the knowledge about vector-borne and zoonotic diseases and landscape structure in the Tropical America region. To achieve this, we create a conceptual model for the Tropical America region summarizing the main relationships and effects found in a landscape ecology perspective; Subsequently, we identify knowledge gaps to propose an agenda regarding landscape ecology and vector-borne and zoonotic disease for the next five years of research. Our results will enhance our understanding of disease dynamics in this region, contribute to improved public health outcomes [9,19,32,33], and allow for better strategies to fill current knowledge gaps.
To assess the general knowledge about zoonotic/vectorborne diseases and landscape structure in the Tropical America region, a syntehsis of the literature was carried out using five steps, based on Arksey and O'Malley [34] and Levac et al., [35]: (1) Identifying the goal of the synthesis: the main objective of this systematic map was to evaluate the existing knowledge on landscape dynamics and disease transmission in the Tropical America, by answering the question (i) what is the relationship between landscape structure and the emergence of vectorborne and zoonotic diseases in the Tropical America? (2) Identification of relevant articles: a scientific literature search was conducted using the litsearchr package [36], which creates a pool of possible keywords relevant to a field of study. First, we ran a search on 14 March 2023 on Scopus database using the "naïve keywords" (neotropic* OR forest OR tropical* OR landscape) AND (zoonotic diseases OR zoonos* OR infectious diseas* OR diseas* OR vector* OR virus*), resulting in 1901 articles. Results were then imported into R, and by using the litsearchr package [36], important keywords were identified in a keyword co-occurrence network: (tropical disease OR tropical region OR neotropics OR forest) AND (infectious disease OR public health OR risk factor OR zoonotic disease OR borne disease OR neglected tropical OR zoonotic pathogen). A new search using these keywords was performed on 15 March 2023 on Scopus, and resulted in 54 articles. To ensure we were covering as many studies as possible, we used the results from both searches (1930 unique articles) and conducted a snowball procedure with the selected articles. The searches were conducted in English and with no restriction on year (as disease outbreaks can happen in isolated points in time, we did not restrict the years of search to capture the largest number of studies possible). We choose english instead of other languages, because the results can be replicated by any researcher of the world. Code with litsearchr search, papers downloaded using the naive search, and papers resulting from the lisearchr search are all available at (https:// github. com/ paula prist/ Curre ntLan dscape. git).
To decide on the inclusion of articles, we use a multistep process: (i) five reviewers performed a preliminary scan of the titles and abstracts, in order to discard those that were not related to our objectives. As a result, from 1930 unique articles, 103 were included for full reading and were then further evaluated by two reviewers, who jointly decided on their inclusion or exclusion. From the 103, 55 more articles were selected for reading after the snowball procedure, regardless of whether they were an empirical, modeling or review study. (3) Article selection: an article was only included in the review if it was written in English and addressed some aspect of landscape ecology and zoonotic/vector-borne diseases. Our inclusion criterion for the landscape aspect was if quantitative or qualitative measurements of any landscape structure element were performed (i.e., both composition and/or configuration), which would allow a comparison to be made between the results found and the different elements of the landscape. For vector-borne and zoonotic diseases, we included articles that assessed the presence or abundance of vectors, hosts, or reservoirs, prevalence of infection in them, or human cases of associated diseases. (4) Data management: to extract and summarize the data we created a spreadsheet, including authors, year of publication, title, geographical location, disease evaluated, landscape feature evaluated, response found. (5) Analyzing, summarizing, and reporting the results: we conducted a qualitative analysis synthesizing the main patterns found in the literature.
We carefully read 158 articles relating landscape variables and zoonotic/vector-borne diseases, and found that 74 met our criteria, entering the final analysis (Table 1). Most of the studies were empirical (64 studies), followed by modeling studies (eight), and meta analysis (two studies). The selected articles were classified into three main categories referring to the landscape variables investigated: landscape composition (i.e., land use and land cover composition and their relative proportion), landscape configuration (i.e., the spatial arrangement of the landscape units) and landscape structure (i.e., landscape composition and configuration together). Studies that accessed vegetation indexes such as NDVI were classified as landscape composition studies. Similarly, studies that evaluated disturbance (i.e., urbanization, agricultural areas, edge habitats) were also classified as landscape composition studies.
The earliest knowledge found in our synthesis for the Tropical America region regarding landscape structure and zoonotic/vector-borne diseases was from 2002, with a low number of studies being published per year, until it increased in 2016 (Fig. 1). Despite a boom in the number of articles in 2021, there is only a small growth in the number of articles published per year. Even considering that the total number almost doubled after 2013, we still see less than 10 studies being published every year (with the exception of 2021). This shows that this is still an under-explored topic in the scientific community of this region.
Likewise, 83% of the studies found are concentrated in five countries: 55% in Brazil (n = 41), 11% in Panama (n = 8), 6% in Colombia (n = 5), and 5% in Argentina and Peru (n = 4 each). Mexico presented three studies, while Paraguay two. All the other countries had only one study each. Two studies were performed in more than one country (Brazil and Argentina and in the entire Tropical America region). This shows that there is a large knowledge gap about landscape-zoonosis dynamics in 15 of the 27 countries that compose the Tropical America region (Fig. 2).
A total of 16 zoonotic diseases/vector related-diseases have been studied in the Tropical America region 87 times (considering that one single article sometimes have analyzed more than one disease), but again, with a trend of many studies looking at a low number of diseases. Five diseases represent 68% of the total current knowledgemalaria (25% n = 22), Chagas disease and Leishmaniasis (13%, n = 11 each), yellow fever (9%, n = 8), and hantavirus (8%, n = 7). Spotted fevers, Arboviruses (diseases specifically transmitted by Culicidae and Aedes aegypti), and mosquito-borne diseases (studies evaluating mosquitoes communities) accounted for 13.9% of the studies found Table 1 (continued) ID Reference Country of study Type of study Disease (s) measured Landscape classification Landscape metrics measured Effect on disease risk Outcome 59 Thies SF, Bronzoni RVM, Michalsky ÉM, Santos ESD, Silva DJFD, Dias ES, Damazo AS (2018) Aspects on the ecology of phlebotomine sand flies and natural infection by Leishmania hertigi in the Southeastern Amazon Basin of Brazil. Acta Trop. 177:37-43. https:// doi. org/ 10. 1016/j. actat ropica. 2017. 09. 023 Brazil Empirical Leishmaniasis Composition Level of urbanization (with forest areas) Positive Higher sand flies diversity and frequency in permanent forest preservation areas 60 Tirera S, de Thoisy B, Donato D, Bouchier C, Lacoste V, Franc A, et al. (2021) The Influence of Habitat on Viral Diversity in Neotropical Rodent Hosts. Viruses 13:1690 French Guiana Empirical Rodent borne viral diseases Composition Disturbed and pristine forests, savannahs, and peri-urban habitats Positive (from peri-urban, disturbed, and pristine habitats) Viral diversities were greater in pristine habitats compared with disturbed ones, and lowest in peri-urban areas 61 Travi BL, Adler GH, Lozano M, Cadena H, Montoya-Lerma J (2002) Impact of habitat degradation on phlebotominae (Diptera: Psychodidae) of tropical dry forests in Northern Colombia. Journal of Medical Entomology, 39(3),.451-456 Colombia Empirical Leishmaniasis Composition Habitat degradation Negative Habitat degradation negatively affected sand fly communities, decreasing vector abundance 62 Valero NNH, Prist P, Uriarte M (2021) Environmental and socioeconomic risk factors for visceral and cutaneous leishmaniasis in São Paulo, Brazil. Science of The Total Environment. 797:148,960 Brazil Empirical Leishmaniasis Composition Native vegetation cover Positive Higher probability of leishmaniosis occurrence in municipalities with high native forest cover 63 Valle D, Clark J (2013) Conservation Efforts May Increase Malaria Burden in the Brazilian Amazon. PLOS ONE 8(3): e57519. https:// doi. org/ 10. 1371/ journ al. pone. 00575 19 Brazil Empirical Malaria Composition Forest cover Positive Malaria incidence was related to greater forest cover Table 1 (continued) ID Reference Country of study Type of study Disease (s) measured Landscape classification Landscape metrics measured Effect on disease risk Outcome 64 Vaz VC, D'Andrea PS, Jansen AM (2007) Effects of habitat fragmentation on wild mammal infection by Trypanosoma cruzi. Parasitology. 134(Pt 12):1785-93. https:// doi. org/ 10. 1017/ S0031 18200 70032 3X Brazil Empirical Chagas disease Configuration small (< 10 ha), medium (10-40 ha), and large (> 40 ha) fragments, continuous forest Negative (from fragmented to continuous forest) Seroprevalence was higher in the fragmented habitat than in the continuous forest 65 Vieira CJSP., Andrade CD, Kubiszeski JR, Silva DJF, Barreto ES, Massey AL, Canale GR, São Bernardo CS, Levi T, Peres CA, Bronzoni RVM (2019) Detection of Ilheus virus in mosquitoes from southeast Amazon, Brazil. Transactions of the Royal Society of Tropical Medicine and Hygiene, 113(7), 424-427. https:// doi. org/ 10. 1093/ trstmh/ trz031 Brazil Empirical Arboviruses (culicidae) Composition Different land use types-urban, forest fragments and agricultural areas Positive for forest areas Mosquito diversity and abundance was higher in forest areas. Urban areas were dominated by Culex species 66 Vieira CJDSP, Steiner São Bernardo C, Ferreira da Silva DJ, Rigotti Kubiszeski J, Serpa Barreto E, de Oliveira Monteiro HA, … & Vieira de Morais Bronzoni, R (2022) Landuse effects on mosquito biodiversity and potential arbovirus emergence in the Southern Amazon, Brazil. Transboundary and Emerging Diseases, 69(4), 1770-1781 Brazil Empirical Arboviruses (culicidae) Configuration Forest edge density, forest size and shape Positive Arbovirus vectors' richness and abundance were associated with small size of forest remnants with more irregular shape and higher edge density 67 Vittor AY, Gilman RH, Tielsch J, Glass G, Shields T.I.M., Lozano WS, Pinedo-Cancino V, Patz JA, (2006) The effect of deforestation on the human-biting rate of Anopheles darlingi, the primary vector of falciparum malaria in the Peruvian Amazon. American Journal of Tropical Medicine and Hygiene, 74(1), 3-11 Peru Empirical Malaria Composition Forest cover in a 1km2 radius Negative Deforestation (areas with < 20% of forest cover) increase A. darlingi biting rate in 278 times when compared to non-deforested areas (areas with > 70% of forest cover) Table 1 (continued) ID Reference Country of study Type of study Disease (s) measured Landscape classification Landscape metrics measured Effect on disease risk Outcome 68 Vittor AY, Pan W, Gilman RH, Tielsch J, Glass G, Shields T, Sánchez-Lozano W, Pinedo VV, Salas-Cobos E, Flores S, Patz JA (2009) Linking deforestation to malaria in the Amazon: characterization of the breeding habitat of the principal malaria vector, Anopheles darlingi. Am J Trop Med Hyg. 81(1):5-12 Peru Empirical Malaria Composition Forest cover in a 1km2 radius Negative Anopheles darlingi larvae were most frequently found in sites with < 20% forest cover 69 Vittor AY, Armien B, Gonzalez P, Carrera JP, Dominguez C, et al. (2016) Epidemiology of Emergent Madariaga Encephalitis in a Region with Endemic Venezuelan Equine Encephalitis: Initial Host Studies and Human Cross-Sectional Study in Darien, Panama. PLOS Neglected Tropical Diseases 10(4): e0004554. https:// doi. org/ 10. 1371/ journ al. pntd. 00045 54 Panama Empirical Vector-borne alphaviruses (Madariaga virus-MADV and Venezuelan equine encephalitis-VEEV) Composition Habitat type (pasture, farms, shrubs, forests) Proximity to human dwellings Activities (cattle ranching, farming, fishing) Positive No effect Positive MADV-positivelly associated with pasture (cattle and horses) and farms (rice, cassava and watermelon). Inverselly correlated with the presence of shrubs within 10 m from residence. VEEV-positivelly associated with farms (rice, sugarcane, watermelon and yam), forests (hunting and logging activities), activities in rivers and pastures (cattle). The prefered habitat of vector rodents coincided with areas associated with human infection risk (Zygodontomys brevicauda-sugarcane, Transandinomys bolivaris-forest) 70 Wayant NM, Maldonado D, Rojas de Arias A, Cousiño B, Goodin DG (2010) Correlation between normalized difference vegetation index and malaria in a subtropical rain forest undergoing rapid anthropogenic alteration. Geospat Health. 4(2):179-90. https:// doi. org/ 10. 4081/ gh. 2010. 199 Paraguay Empirical Malaria Composition NDVI Negative Number of malaria cases increased in modified areas (from forest to non-forest) Table 1 (continued) ID Reference Country of study Type of study Disease (s) measured Landscape classification Landscape metrics measured Effect on disease risk Outcome 71 Wilk-da-Silva R, Mucci LF, Ceretti-Junior W, Duarte AMRC, Marrelli MT, Medeiros-Sousa AR (2020) Influence of landscape composition and configuration on the richness and abundance of potential sylvatic yellow fever vectors in a remnant of Atlantic Forest in the city of São Paulo, Brazil. Acta Trop. 204:105,385. https:// doi. org/ 10. 1016/j. actat ropica. 2020. 105385 Brazil Empirical Yellow fever Structure Forest area with intermediate and high degrees of conservation (FA) Consolidated urban area (CUA) Anthropic area associated with forest area (AAF) Lake areas Edge between FA and CUA Edge between FA and AAF Negative for YF vectors No effect No effect No effect No effect Positive for YF vectors Landscapes with higher amounts of forest edge areas had increased richness and abundance of YF vector species 72 Wilk-da-Silva R, Medeiros-Sousa AR, Laporta GZ, Mucci LF, Prist PR, Marrelli MT. (2022) The influence of landscape structure on the dispersal pattern of yellow fever virus in the state of São Paulo. Acta Trop.228:106,333. https:// doi. org/ 10. 1016/j. actat ropica. 2022. 106333 Brazil Empirical Yellow fever Structure Forest Formation and its interface with: Water Human-impacted area (no-vegetation) Urban area Non-forest formation (native bushy or herbaceous vegetation) Forestry (Eucalyptus or Pinus spp.) plantations Negative Negative No effect Negative Positive Positive Yellow fever virus was influenced by forest edges in interface with agricultural areas 73 Winck GR, Raimundo RLG, Fernandes-Ferreira H, Bueno MG, D'Andrea PS, Rocha FL, Cruz GLT, Vilar EM, Brandão M, Cordeiro JLP, Andreazzi CS (2022) Socioecological vulnerability and the risk of zoonotic disease emergence in Brazil. Sci Adv. 8(26): eabo5774. https:// doi. org/ 10. 1126/ sciadv. abo57 74 Brazil Modelling Chagas disease, yellow fever, spotted fever, skin and visceral leishmaniasis, hantavirus, leptospirosis, malaria, and rabies Composition Proportion of natural vegetation cover City remoteness Negative Positive The mean zoonotic diseases was negatively affected by vegetation cover and positively affected by city remoteness
(with four studies each), while rabies and leptospirosis 6.8%. Cryptosporidium, Giardia spp., Bertiella infection, and helminths form 9% (n = 5). Apparently, this bias is related to the countries where the studies were conducted; for instance, all of the top five most studied diseases are common infections in Brazil, the country with the largest number of studies. With respect to landscape parameters, most of the knowledge focuses on understanding how landscape composition affects the transmission risk of these diseases (50.67%, n = 56), with 19% (n = 14) of studies assessing both composition and configuration; and ~ 5.5% (n = 4) focusing only on the configuration of native vegetation areas.
Most of the studies found a significant relationship between landscape aspects and disease risk, with more conserved habitats (higher forest amounts, NDVI and connectivity) showing higher pathogen and host diversity [37][38][39], but with lower prevalence and infection intensities [37,39]. Consequently, these habitats also present a low spillover risk, which corroborates with the dilution effect hypothesis [40]. For yellow fever virus, it was also found that more conserved areas can barrier the movement of the virus, decreasing transmission risks for this disease [41]. Habitat loss was appointed as an important driver of the Brazilian spotted fever [42], leishmaniasis, malaria [43,44], and Chagas disease [45]. Likewise, disturbed/impacted areas presented lower pathogen and host diversity, but with hosts exhibiting higher abundances and pathogen prevalences and infection intensities [37,39] and a higher spillover risk. However, some studies also found an opposite effect-large amounts of forest areas, and the closer an individual is to the forest or dense vegetation areas, higher are the risk of contracting leishmaniasis and malaria [46,47]. However, none of these studies had a landscape ecology design, where some of the potential cofounding factors are controlled. In addition, none of them (but see [48]) assessed non linear responses (i.e., for example, leishmaniasis reaches a maximum value at intermediate levels of forest cover and then decreases as forest cover increases [48]) and potential forest cover thresholds, where sharp increases in transmission risk can occur when habitat loss reaches a certain value. But in summary, with the existing knowledge to date, landscape composition has a positive association with most of the diseases studied, with habitat loss and degradation (i.e., urbanization, anthropogenic land use types) presenting the greatest risk to human health.
Diseases, with Habitat Fragmentation and Increases in the Amount of Forest Edges Being the Most Important Drivers Habitat fragmentation has lead to an increase in the infection rate of Anopheles mosquitoes [49], and consequently boosted malaria transmission risk, especially if these
Table 1 (continued) ID Reference Country of study Type of study Disease (s) measured Landscape classification Landscape metrics measured Effect on disease risk Outcome 74 Yamada K, Valderrama A, Gottdenker N, Cerezo L, Minakawa N, Saldaña A, Calzada JE, Chaves LF (2016) Macroecological patterns of American Cutaneous Leishmaniasis transmission across the health areas of Panamá (1980-2012). Parasite Epidemiol Control. 18;1(2):42-55. https:// doi. org/ 10. 1016/j. parepi. 2016. 03. 003 Panama Empirical Leishmaniasis Composition Proposrtion of forest cover Both positive and negative Leishmaniasis reaches a maximum value at intermediate levels of forest cover and then decreases as forest cover increases
forest patches are immersed in an agro-pastoral matrix [50,51]. For yellow fever virus, a similar response was observed, with highly fragmented habitats composed by large amounts of forest edges in interface with agricultural areas, and forest roads increasing the movement and potentially transmission risks [41,52,53]. For other diseases, like the Brazilian spotted fever, not only habitat loss and fragmentation are important but also isolation between these remnants [42], while for hantavirus and leishmaniasis, an important landscape feature that seems to determine transmission risk is the amount of sugarcane, coffee and maize present [54][55][56][57].
Despite these indications, no studies have tested effects of habitat configuration controlling the amount of forest cover in the landscape, which could potentially affect the results obtained.
From 74 studies evaluated, four found no significant relationships between landscape changes and disease risk [58,59]. In all of them, the effects of land use types and degradation on arboviruses, Chagas disease, Cryptosporidium, Giardia spp., and rabies risk were tested. However, these studies did not test or control for important landscape factors, such as percent remaining cover between sampled points, which may affect the observed outcomes. Nevertheless, this pattern shows that in 95% of the knowledge found, landscape features affect the transmission risk of diseases in the Tropical America.
Based on the knowledge compiled, we built a conceptual model summarizing the most common relationships between landscape change and disease risk (Fig. 3). Natural and preserved landscapes (box 1) can harbor a high diversity of hosts, vectors and consequently, of pathogens. However, in these areas the abundance of vectors and reservoirs is low, and consequently pathogen prevalence (box 2). These factors, coupled with low human-wildlife contact rates, lead to a decrease in spillover risk (grey arrows, box 3). When human actions alter this natural environment through habitat loss (box 4), two possible pathways can alter the distribution of species and hence the risk of diseases in semi-natural and anthropogenic landscapes-land use changes from natural to urban (box 5), and from natural to agricultural environments (box 6).
Urbanization leads to a complete alteration of the fauna communities through the loss of sylvatic species (i.e., species that require natural habitats to survive; boxes 7 and 8), and the consequent increase in the abundance of synanthropic species (box 9). The propensity of these species to live exclusively or occasionally within or near human habitations (synanthropy) has long been recognized to increase the transmission risk of important zoonoses that threaten public health [60]. Therefore, in urban environments we normally see an increase in the transmission risk (red arrow) of zoonoses and vector-borne diseases linked to these species (i.e., dengue, zika virus, leptospirosis). However, at the same time, other diseases will have a reduced risk in highly urbanized areas (black arrow), because main hosts and vectors will not find suitable habitat for their survival in these locations (e.g., hantavirus, malaria, Chagas disease, leishmaniasis).
When the native vegetation loss gives space to agricultural areas (box 6), such as crops and pasture, for example, spillover risk will be dependent not only on the amount of area lost but also on the spatial arrangement of the remaining habitat areas (boxes 10 and 11). If vegetation loss leads to tina; BR, Brazil; CO, Colombia; CR, Costa Rica; EC, Ecuador; FG, French Guiana; GU, Guatemala; ME, Mexico; PE, Peru; PG, Paraguay; PN, Panama; VE, Venezuela
landscapes with a high fragmentation state (box 11), spillover risk can be enhanced. This may happen because specialist species normally are extinct in these landscapes, giving space to generalist species, which normally act as reservoir and hosts for zoonotic diseases. These species not only adapat but also thrive in fragmented landscapes, increasing their abundances and consequently the infection prevalence of the pathgens they carry. In addition, highly fragmented habitats normally have increased forest edges, which boost contact between humans-wildlife, increasing spillover risk (red arrow). This is a favorable context to the circulation of yellow fever virus, for example, as small fragments surrounded by agro-pastoral activities may support viable populations of non-human primates and vector species (Alouatta and Haemagogus spp., respectively), making people who use the vicinity of the fragments especially vulnerable [50,61]. However, these effects may be dependent on the amount of habitat remaining, a factor not yet tested for zoonotic diseases (see the "Future of the Disease Ecology in the Tropical America" section for more details).
If the remaining vegetation is present in a low fragmented state, with reduced amounts of edge density, spillover risk can potentially decrease. This happens because in less fragmented landscapes the abundance of specialist species are maintained, with a consequent reduction in the densities of generalist species (i.e., zoonotic disease hosts and reservoirs, [19]), which may contribute to the reduction of contact between generalist-reservoir species and humans [12] reducing spillover risk.
Based on these findings, we also expect that when forest restoration is promoted (box 20), it results in an increase of native habitat amount, which decreases the amount of edge density, and the interface between humans and wildlife (box 21), consequently decreasing transmission risk (dashed Fig. 3 Conceptual model showing the main relationships between landscape changes and zoonotic spillover risk based on the reviewed data for the Tropical America. Processes that can increase or decrease spillover risk are shown as gray arrows; processes that decrease spillover risk are shown as black processes that increase spillover risk are shown as red arrows. Images designed by freepik.com grey arrow). However, there is no evidence to date to support this hypothesis.
Based on the body of empirical evidence found through the literature synthesis, we discuss some points that have been little or not at all explored, and that are important to be studied in the coming years to increase the knowledge about processes and mechanisms relevant to the effects of landscape structure on the emergence of zoonoses in the Tropical America. These gaps vary from investigating the nexus between simple landscape metrics and prevalence of infection in hosts, to more complex investigation of mechanisms associated with infectious disease emergence in socioecological systems.
There are gaps in Regional and Disease Knowledge that Should be Considered in Future Research Our results showed that a low number of diseases comprise most of the knowledge that exists today. Likewise, most of the countries that compose the Tropical America region have no studies on this topic. Given the great biodiversity of this region, coupled with the great diversity of pathogens and the high deforestation rates and land use change, it is essential that these knowledge gaps are filled. Not only studies on how landscape structure affects the risk of these diseases, but also how host and vector populations are affected by these changes, as well as the viral communities they carry, are essential to form a knowledge base and prevent human action from increasing not only the incidence of these diseases but also of pandemic potential viruses.
Dynamics: Effects of Time lag on Responses of Vectors, Hosts, and Reservoirs Many studies have demonstrated the impact of native vegetation loss on zoonotic diseases outbreaks [6,20]. In contrast, the effects of landscape dynamics, which refer to the changes that occur in the landscape structure over time, are rarely evaluated (see [62]). These changes can include alterations in the size, shape, and connectivity of habitat patches, as well as the fragmentation or expansion of different land use types [63]. However, to gain a more comprehensive understanding of which factors are leading to an increased risk of these diseases, it is essential to analyze the effects of landscape structure dynamics on vector, host, and reservoir populations. In addition, it is important to consider that there is a time required for these species to adjust to the new landscape conditions. These time lag responses refer to the delay between changes in the landscape and the subsequent effects on disease dynamics [64]. Understanding the relationship between landscape dynamics and the time lag responses of vectors, hosts, and reservoirs is crucial for predicting and managing disease outbreaks. Therefore, we emphasize the importance of conducting long-term studies with different landscape dynamic aspects to gain a deeper understanding of this topic. These studies are essential for implementing appropriate land management strategies that consider the potential health impacts of landscape modifications.
There is a substantial body of research that focuses on the relationship between deforestation and the emergence of zoonotic diseases. However, the impact of forest gain or other nature-based solutions on zoonotic disease risks in the Tropical America has not been empirically investigated. Only two studies, employing modeling and review approaches respectively, have examined this topic in this region (see [41,61])-while a hypothetical restoration scenario presented positive effects in reducing the abundance of hantavirus reservoir rodents [61], a recent conceptual framework has shown that the effects of forest restoration on zoonotic diseases may depend on the existing landscape's context [26]. Understanding the potential effects of forest gain on zoonotic disease risk is crucial for the development of effective strategies for disease prevention and control, as well as for promoting ecosystem health, including naturebased solutions.
Contexts Urban landscapes, with their intricate interplay of green spaces and gray infrastructure can serve as complex ecosystems that influence the presence and transmission of zoonotic and vector-bone diseases. For instance, the configuration and distribution of vegetation within urban areas can directly impact the abundance and diversity of species potentially acting as reservoirs or amplification hosts for zoonotic and vector-borne. Understanding these relationships is crucial for effective disease prevention and control strategies in urban settings. Yet, current research on urban landscapes and zoonotic and vector-borne diseases primarily focuses on other aspects, such as social determinants of health or the impact of natural outdoor environments on human health [65]. Consequently, the specific influence of urban design and landscape structure on zoonotic disease dynamics remains largely unexplored. Moreover, the findings from these studies could inform evidence-based urban planning and design strategies aimed at reducing zoonotic disease risks. Implementing nature-based solutions, such as strategically incorporating green spaces and green roofs [66] could help mitigate the transmission of zoonotic diseases. Additionally, optimizing building designs and spatial layouts to minimize potential disease hotspots and enhance public health measures could contribute to healthier urban environments.
Aspects Based on our findings, there has been limited research assessing the impact of landscape configuration on the transmission risk of zoonotic and vector-borne diseases. The majority of the existing studies focus on evaluating the effects of landscape composition, particularly forest loss or the extent of remaining forest cover on zoonotic diseases, while disregarding the crucial aspects, such as arrangement of both habitat [67] and matrix [68]. Studies have demonstrated that these factors have a significant impact on landscape permeability and species interactions [69], and they may be more important than remaining coverage [70] particularly in disease transmission. However, the effects of habitat configuration are landscape context-dependent and species-specific, making predictions and generalizations difficult [71]. In relation to the landscape context, the size, shape, and proximity of habitat patches can interact with species traits to shape their responses to fragmentation [71]. We emphasize the significance of studies on zoonotic diseases, go beyond the conventional metrics of landscape composition, and offer a more nuanced understanding of the role of habitat configuration.
Climate is a key factor limiting the distribution and establishment rate of vector populations [72], human-wildlife contact [73], and other risk factors that can modify zoonotic risk [74]. Temperature and precipitation play a key role in limiting the distribution of vectors and hosts, with temperature increases due to climate change creating favorable conditions for the expansion of the distribution range of these two groups of species [75,76]. Increases in temperature can also lead to an acceleration of the extrinsic incubation period (i.e., time interval between infection and the vector's transmission capacity) [77] increasing the rate of replication of the pathogen in the vector, which becomes infectious more quickly, favoring prevalence and intensity of infection [77,78]. These conditions are exacerbated by land use changes, and natural environment fragmentation-deforestation can heat a local area by as much as 4.5 ℃, and can even raise temperatures in undisturbed forests up to 6 km away [79]. These climate changes, even on a small scale, can favor the spread and transmission of diseases, making deforested and fragmented areas more at risk due to microclimatic conditions that favor the infectivity of the vectors present-one study in Asia, for example, found that the conversion from lowland rainforest to plantations increases suitability for Aedes albopictus development by 10.8% [80], while another one in Costa Rica found that the effects of the El Niño Southern Oscillation on the incidence of Leishmaniasis is exacerbated in places with greater deforestation [81]. These relationships between landscape, climate, and diseases are still little explored, especially in the Tropical America region, and are essential to better comprise the mechanisms behind spillover risk.
Landscape Effects are Scale-Dependent and it is Essential to Understand this in Order to Propose Appropriate Management Strategies Both landscape composition and configuration parameters have the potential to impact the dynamics of vector-borne and zoonotic diseases. However, these effects can be scale-dependent. A study in Southeast Asia, for example, found distinct effects from different spatial scales analyzed for zoonotic malaria occurrence [82]. Transmission dynamics of zoonotic malaria occurs among humans who live in settlements in deforested sites where the malarial parasite, Plasmodium knowlesi, is harbored by macaques (Macaca) and transmitted by anopheline vectors. Forest cover loss in the previous year influenced P. knowlesi occurrence only at small scales (within 0.5 km of households, or 0.78-km 2 ), while fragmentation had higher levels of influence at larger scales (5 km, or 78-km 2 ) [82]. Another study found that temperature and human population density at 1 km scale were important drivers of Hendra virus spillover, while forest cover and pasture were only predictors if considered in a 100-km radius scale [83]. Understanding these dynamics and their scales of effect are essential to propose adequate management and mitigation strategies.
There is a Large gap in Knowledge Between Trade-Offs and Synergies of Different Zoonotic/Vector-Borne Diseases, Which Should be Assessed Before Proposing Landscape Management Strategies It is reasonable to assume that each vectorborne and zoonotic disease has its own signature of scale, composition, and configuration in the human-modified landscape. Therefore, a landscape that can decrease the abundance of hosts and/or vectors of one pathogen may also lead to increases in the abundance of hosts of another pathogen. One study tried to understand these trade offs and synergies for three different diseases (malaria, leishmaniasis, and Chagas disease) in the Brazilian Amazon, using the number of human cases as response variable. They identify that there are trade-offs between the spatial and temporal aggregation of these diseases, with few municipalities being considered critical for more than one disease at the same time [84]. At the same time synergies between landscape features that can affect both malaria and cutaneous leishmaniasis were found-municipalities at risk should be considered those with more than 50% of forest cover and that are experiencing deforestation. For Chagas disease, the at-risk landscapes have continuous forests immersed in a matrix of non-pasture (i.e., crops of Açai trees) [84]. Understanding these trade-offs are of extreme importance to define the most cost-effective landscape management strategies. By knowing that a landscape management can decrease one zoonosis risk, while increasing another one, actions and control strategies that aim to change human risk behavior can be prioritized.
Human Behavior and Socioeconomic Aspects are Important Drivers in Determining Disease Risk and can also be Determined by Landscape Parameters The risk of zoonotic disease transmission is dependent on interacting ecological and human behavioral and socioeconomic factors [85,86]. Humans are especially vulnerable for zoonotic pathogens when slaughtering, butchering, or consuming raw meat from wild animals [87], for example. In addition, human behaviors or costums that can lead to human-wildlife interactions can also lead to an increased risk of disease spillover, and these behaviors will change according to the landscape features close to these populations. Hunting is normally increased in fragmented landscapes, putting more people at risk for zoonotic spillover if compared with intact forest areas. Similarly, human cases of Marburg virus in Kenya were associated with visits to caves that house infected bats [88], while drinking untreated water can result in exposure to diverse pathogens, such as Leptospira leptospirosa spp. [89]. Despite its extreme importance, human risk behaviors are rarely taken into account in landscape epidemiology studies. Similarly, socioeconomic factors, which are also directly linked to zoonotic [90] are understudied in a landscape ecology perspective. The lack of access to nutritious food, for example, is associated with an increased risk for multiple health conditions [91], while housing conditions can also be important factors to prevent or favor contact between humans, animal hosts, and zoonotic pathogens [92]. Identifying populations at risk [56], risky behaviors and customs and socioeconomic factors that lead to increased exposure of people to pathogens, and understanding how landscape features affect these drivers may be a key element in mitigating the risk of zoonotic disease transmission. Even if we manage landscapes by keeping host and vector density controlled, as well as the chance of contact between humans and these animals, risky human behavior can lead to the spread of pathogens and increased risk of transmission even in low pathogenicity landscapes. For instance, effective longterm community projects around natural areas have proven to reduce illegal resource extraction activities inside protected areas in Uganda through social change [93]. This type project likely decreases the contact between people and wildlife and as a consequence, reduces disease risk. At the same time, higher landscape accessibility (measured as combining terrain ruggedness, elevation, friction, slope, aspect, and tree cover) lead to more illegal resource extraction in the park, an important threat to biodiversity conservation. This example shows how health and biodiversity conservation is connected in fragmented landscapes and supports the idea that the environment needs to be managed under a One Health perspective, thinking about the interconnectedness between people's health, animal health, and environmental health [94]. Therefore, it is essential to better understand interconnectedness and behaviors and socioeconomic factors that lead to increased disease transmission risk in order to prevent spillovers from occurring in Tropical America landscapes.
Our synthesis has shown that the scientific production linking landscape change to zoonoses in the Tropical America still has been incipient and mainly focuses on malaria, yellow fever, Hantavirus, and Chagas disease. There is an urgent need for research connecting landscape ecology and disease ecology, so stakeholders can make science-informed decisions for land use management, and public health plans can be developed considering landscape pathogenicity. Despite significant knowledge gaps, our results point out that landscape structure affects dynamic processes of disease emergence, leading to consequences for ecosystem services and public health. Research on complex topics largely relies on capacity building and investment in research and development in the Tropical America and more broadly in the Global South. We plead for the enforcement of support in research on these topics in the Tropical America, which can be guided through interdisciplinary approaches such as One Health and Nature Based solutions.