Pathogen and pest communities in agroecosystems across climate gradients: Anticipating future challenges in the highland tropics

Pathogen and pest communities in agroecosystems across climate gradients: Anticipating future challenges in the highland tropics | bioRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-M677548'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search New Results Pathogen and pest communities in agroecosystems across climate gradients: Anticipating future challenges in the highland tropics View ORCID Profile Romaric A. Mouafo-Tchinda , View ORCID Profile Aaron I. Plex Sulá , View ORCID Profile Berea A. Etherton , View ORCID Profile Joshua S. Okonya , View ORCID Profile Gloria Valentine Nakato , View ORCID Profile Yanru Xing , View ORCID Profile Jacobo Robledo , Ashish Adhikari , View ORCID Profile Guy Blomme , View ORCID Profile Déo Kantungeko , View ORCID Profile Anastase Nduwayezu , View ORCID Profile Jan F. Kreuze , View ORCID Profile Jürgen Kroschel , View ORCID Profile James P. Legg , View ORCID Profile Karen A. Garrett doi: https://doi.org/10.1101/2025.01.08.631994 Romaric A. Mouafo-Tchinda 1 Plant Pathology Department, Institute of Food and Agricultural Sciences, University of Florida , Gainesville, FL, USA 2 Global Food Systems Institute, University of Florida , Gainesville, FL, USA 3 Emerging Pathogens Institute, University of Florida , Gainesville, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Romaric A. Mouafo-Tchinda For correspondence: rmouafotchinda{at}ufl.edu karengarrett{at}ufl.edu Aaron I. Plex Sulá 1 Plant Pathology Department, Institute of Food and Agricultural Sciences, University of Florida , Gainesville, FL, USA 2 Global Food Systems Institute, University of Florida , Gainesville, FL, USA 3 Emerging Pathogens Institute, University of Florida , Gainesville, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Aaron I. Plex Sulá Berea A. Etherton 1 Plant Pathology Department, Institute of Food and Agricultural Sciences, University of Florida , Gainesville, FL, USA 2 Global Food Systems Institute, University of Florida , Gainesville, FL, USA 3 Emerging Pathogens Institute, University of Florida , Gainesville, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Berea A. Etherton Joshua S. Okonya 4 Association for Strengthening Agricultural Research in Eastern and Central Africa , Entebbe, Uganda Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Joshua S. Okonya Gloria Valentine Nakato 5 International Institute of Tropical Agriculture , Arusha, Tanzania Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Gloria Valentine Nakato Yanru Xing 1 Plant Pathology Department, Institute of Food and Agricultural Sciences, University of Florida , Gainesville, FL, USA 2 Global Food Systems Institute, University of Florida , Gainesville, FL, USA 3 Emerging Pathogens Institute, University of Florida , Gainesville, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Yanru Xing Jacobo Robledo 1 Plant Pathology Department, Institute of Food and Agricultural Sciences, University of Florida , Gainesville, FL, USA 2 Global Food Systems Institute, University of Florida , Gainesville, FL, USA 3 Emerging Pathogens Institute, University of Florida , Gainesville, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Jacobo Robledo Ashish Adhikari 1 Plant Pathology Department, Institute of Food and Agricultural Sciences, University of Florida , Gainesville, FL, USA 2 Global Food Systems Institute, University of Florida , Gainesville, FL, USA 3 Emerging Pathogens Institute, University of Florida , Gainesville, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Guy Blomme 6 Bioversity International , Addis Ababa, Ethiopia Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Guy Blomme Déo Kantungeko 7 International Institute of Tropical Agriculture , Bujumbura, Burundi Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Déo Kantungeko Anastase Nduwayezu 8 Rwanda Agriculture Development and Animal Resources Board (RAB) , Ruhengeri, Rwanda Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Anastase Nduwayezu Jan F. Kreuze 9 International Potato Center (CIP) , Lima, Peru Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Jan F. Kreuze Jürgen Kroschel 9 International Potato Center (CIP) , Lima, Peru 10 Hans-Ruthenberg-Institute for Tropical Agricultural Sciences, University of Hohenheim , Stuttgart, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Jürgen Kroschel James P. Legg 11 International Institute of Tropical Agriculture, Dar es Salaam , Tanzania Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for James P. Legg Karen A. Garrett 1 Plant Pathology Department, Institute of Food and Agricultural Sciences, University of Florida , Gainesville, FL, USA 2 Global Food Systems Institute, University of Florida , Gainesville, FL, USA 3 Emerging Pathogens Institute, University of Florida , Gainesville, FL, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Karen A. Garrett For correspondence: rmouafotchinda{at}ufl.edu karengarrett{at}ufl.edu Abstract Full Text Info/History Metrics Preview PDF ABSTRACT CONTEXT Tropical cropping systems must adapt to the current and future geographic distribution of pathogen and pest communities. An important research gap is how climate change may shift the distribution of pathogens and pests in tropical lowlands and highlands. OBJECTIVE We evaluated the current geographic risk of 27 pathogens and pests in the production of four food security crops (banana, cassava, potato, and sweetpotato) in the Great Lakes region of Africa, and the potential future risk under climate change. Models for each pathogen and pest indicate the potential for changes in geographic distribution, with model fit indicating the potential for decision support systems to facilitate management. METHODS First, cropland connectivity analysis identified locations likely important in the spread of crop-specific pathogens and pests, such as locations in Rwanda and Burundi. Second, we surveyed the 27 economically important pathogens and pests in Rwanda and Burundi, mapping the distribution of each across climate gradients and quantifying patterns of association. Third, we used machine learning to develop models of each species as a function of environmental variables, including host landscape variables. We also evaluated the increase in temperature across altitudes under future climate change scenarios in this region. RESULTS AND CONCLUSIONS Among the ten machine-learning algorithms evaluated, random forests and support vector machines generally performed best for predicting severity and infestation. Host landscape variables were useful predictors for some species. Based on climate matching, 44% of the pathogens and pests could become more common with warmer temperatures at higher altitudes, while 17% may become less common. SIGNIFICANCE This study indicates adaptation priorities for crop health in a region with multiple challenges to agricultural sustainability. The models developed here also indicate which species may have more potential and relevance for future development of pathogen and pest forecasts. 1. Introduction 1.1. Impact of climate and crop landscape structure on plant pathogens and pests Agroecosystems in the tropics, such as the Great Lakes region of Africa, are often challenged by climate variability and societal dynamics including demographic pressures, socio-political instability, and resource-related conflict ( Ogot and Niane, 1984 ; Ong’ayo, 2018 ). Climate change is likely to affect agricultural productivity and exacerbate the proliferation of some pathogens and pests in this region ( Ebregt et al ., 2005 ; Ferdu and Solbreck, 2010 ; Okonya and Kroschel, 2013 ; Gallego-Tévar et al ., 2024 ). Ecosystems can be reshaped by new temperature and precipitation patterns, driving pathogens and pests to new locations ( Sparks et al ., 2014 ; Cilas et al ., 2016 ; Garrett et al ., 2022a ; Garrett et al ., 2022b ). The interactions among climate heterogeneity, the geographic structure of agroecosystems, and pathogens and pests in tropical regions such as the Great Lakes region represent a major knowledge gap. Research has examined how climate affects the dynamics of specific pathogens and pests ( Kroschel et al ., 2014 ; Adhikari et al ., 2015 ; Blomme et al ., 2020 ). However, field studies that allow for simultaneous comparisons of many different pathogens and pests along climate gradients are rare, and little is known about how regional factors, such as the geographic structure of agroecoystems in the Great Lakes region, exacerbate risks. It is an open question how pathogen and pest communities are distributed across altitudes in this region, what variables are predictive of pathogen and pest community structure, and how their distribution may shift under climate change. Climate matching studies identify which locations have similar climate conditions, such that they may be equally suitable for a species; ecological communities may move following changes in where suitable climate conditions are found in the future ( Colwell et al ., 2008 ; Singh et al ., 2023 ; Chevalier et al ., 2024 ). For example, pathogen and pest communities now at hotter, lower altitudes may move to what are currently cooler, higher altitudes if there is warming due to climate change. The dynamics of pathogens and pests in response to climate have been widely studied ( Hodkinson, 2005 ; Pautasso et al ., 2012 ; Juroszek et al ., 2022 ; Mishra et al ., 2024 ). Higher altitudes are commonly associated with lower species richness, disease severity, and pest damage ( Hodkinson, 2005 ; Blomme et al ., 2020 ). Since altitude is typically negatively correlated with temperature ( Poveda et al ., 2012 ), altitude can be a useful proxy when evaluating the effect of temperature on plant disease and pest damage. For example, previous studies in South America and East Africa predicted that rising temperatures due to climate change will modify the geographic distribution of pathogens and pests, increasing plant disease severity or pest damage at higher altitudes ( Poveda et al ., 2012 ; Blomme et al ., 2020 ). New climate conditions favorable to certain pathogens and pests are likely to occur at higher altitudes, disrupting ecological balances and exposing higher-altitude plants and ecosystems to new risks. The geographic distribution of crop production can also influence geographic patterns of disease severity and pest damage. Crop landscape structure has been well documented to influence crop damage from diseases and pests ( Thies and Tscharntke, 1999 ; Poveda et al ., 2012 ). Analysis of the potential invasion network for pathogens and pests in crop landscapes, such as cropland connectivity analysis, has been used to identify candidate priority locations for surveillance and management of host-specific pathogens and pests ( Xing et al ., 2020 ; Andersen Onofre et al ., 2021 ; Buddenhagen et al ., 2022 ). An analysis of cropland geographic structure indicates the geographic distribution of potential spread of pathogens and pests. Higher host plant availability generally increases the local risk of host-specific pathogen and pest establishment, while greater host connectivity can also increase the risk of spread and establishment. 1.2. Disaster plant pathology: Vulnerabilities and solutions in the Great Lakes region Many families living in the Great Lakes region rely on food security crops such as banana and plantain, cassava, potato, and sweetpotato. However, many households remain food insecure because of declining agricultural productivity, driven by frequent disasters in the region, including conflicts, health emergencies, and natural disasters that displace populations and disrupt agriculture ( Ochieng et al ., 2014 ; Ochieng et al ., 2015 ; Ong’ayo, 2018 ). These challenges are compounded by the cascading effects of high prevalence of plant pests and diseases, low agricultural input use, and loss of harvest and assets ( Kroschel et al ., 2014 ; McEwan, 2016 ; Almekinders et al ., 2019 ). The Rwandese genocide in 1994, the civil war in Burundi from 1993 to 2005 and also in 2015, and the persistent conflicts in Eastern DR Congo are particularly important humanitarian disasters in the Great Lakes region, which have left an indelible mark ( Ong’ayo, 2018 ). Health emergencies and natural disasters, such as severe droughts in Ethiopia, Kenya, and Uganda (2015-2016) and devastating floods in Rwanda, Uganda, and DR Congo (2020), have severely restricted agricultural practices such as effective surveillance and management of fields, causing considerable damage to infrastructure and cropland ( Ochieng et al ., 2014 ; Ochieng et al ., 2015 ). Disaster plant pathology, integrating concepts from plant pathology and disaster science, addresses the risks and potential impact of plant diseases before, during, and after disaster events that threaten agricultural systems and ecosystems ( Etherton et al ., 2024 ; Mouafo-Tchinda et al ., 2024 ). It focuses on how disturbances modify host-pathogen dynamics and increase damage and the spread of pathogens, and on the development of strategies for better managing plant health. A combination of global and national research efforts has been adopted to manage the impacts that these disasters, pathogens, and pests may have on agroecological systems ( RTB, 2013 ; Science Group Project on Accelerated Varietal Improvement and Seed Systems in Africa, 2024 ). For example, disease control programs which encourage farmers to recognize and destroy diseased plants, and disease-resistant crop breeding programs are currently underway across the Great Lakes region. When models of pathogen and pest responses to environmental variables are available, decision-support tools can be developed to predict where specific disease- and pest-resistant varieties will be needed in the long-run, and in the short-run whether likely economic injury levels motivate management actions by growers ( Kaundal et al ., 2006 ; Carisse and Fall, 2021 ; Fenu and Malloci, 2021 ). Pest and disease forecasting can reduce production costs, pesticide use, and environmental pollution, protecting the health of farmers and consumers. Machine learning can be used to analyze high volume datasets, and improve the predictive power, accuracy, and efficiency for diseases and pests ( Kaundal et al., 2006 ; Garrett et al., 2022a ). Implementing decision support systems in low-resource countries such as those in the Great Lakes region is challenging; in addition to the often-limited availability of high-resolution local weather data as input to models, field observations and georeferenced datasets for the many important pathogen and pest species are often unavailable for constructing predictive models. 1.3. Objectives In this study, we evaluated the geographic distribution and risk factors for 27 pathogens and pests in the production of four food security crops — banana (including plantain), cassava, potato, and sweetpotato — in the Great Lakes region of Africa, including the potential future risk under climate change. The first objective was to analyze the geographic risk of host-specific pathogen and pest spread based on host landscape connectivity for these crops. The second objective was to characterize the seasonal and geographic occurrence and associations of 27 economically important pathogens and pests affecting these crops, through field surveys conducted in two key locations in the region, Rwanda and Burundi. Part of this dataset has been used to characterize the prevalence of potato viruses and vectors ( Okonya et al ., 2021 ) and the occurrence of banana pests and diseases across altitudes ( Nakato et al ., 2023 ). The third objective was to predict pathogen and pest severity/infestation based on climate variables and host landscape structure using machine learning, and to assess which pathogens and pests are likely to experience altitudinal range shifts under projected changes in temperature. The field surveys were implemented along environmental gradients to compare lower- and higher-altitude pathogen and pest communities, where lower-altitude communities have the potential to move toward higher altitudes in the future. The simultaneous study of these 27 pathogens and pests allows a unique direct comparison of their associations with environmental predictors. These analyses can also support decision-making by (i) national agricultural agencies that plan geographic surveillance strategies for economically important pathogens and pests, (ii) local agricultural research communities and international partners as they prioritize efforts to improve current and future mitigation responses to pathogens and pests, and (iii) farmer communities that need to be prepared for changes in pathogen and pest dynamics due to natural or human-driven disasters, including changing climates. 2. Materials and methods 2.1. Landscape structures influencing pathogen and pest risk 2.1.1. Identifying locations with high cropland connectivity The spatial pattern of host availability is a key ecological trait that influences the spread of host-specific pathogens and pests. We analyzed the connectivity of host landscapes in the Great Lakes region to map the potential establishment and regional spread of host-specific pathogens and pests, to identify candidate priority locations for spatially targeted surveillance in four food security crops — banana (including plantain), cassava, potato, and sweetpotato. This cropland connectivity analysis assesses how crop production is connected in a landscape, based on host availability and physical proximity, in a gravity model ( Jongejans et al ., 2015 ). Cropland connectivity analysis identifies potential habitat networks through which pests and pathogens can spread, a first approximation to epidemic or invasion networks, which can be refined as more data become available ( Margosian et al ., 2009 ; Xing et al ., 2020 ; Andersen Onofre et al ., 2021 ). For each crop, we used publicly available estimates of the cropland harvested area (or cropland density) across sub-Saharan Africa in 2017 ( International Food Policy Research Institute, 2020 ) as a proxy for host availability to host-specific pathogens and pests. Maps of cropland density, aggregated at a spatial resolution of 10-min (or grid cells of image 9.27 km × 9.27 km), were used to evaluate the potential epidemic or invasion network using a gravity model ( Xing et al ., 2020 ). Here the gravity model incorporated two pathogen or pest dispersal kernels: for the inverse power law model, and 𝑐 𝑖 𝑐 𝑗 𝑒 − 𝛾𝑑𝑖𝑗 for the negative exponential model. In each dispersal model, 𝑑 𝑖𝑗 is the Vincenty ellipsoid distance between a pair of cropland nodes (locations) 𝑖 and 𝑗, 𝑐 𝑖 is the cropland density for node 𝑖, and 𝛽 and 𝛾 are dispersal parameters describing the relative likelihood of pathogen or pest movement across 𝑑 𝑖𝑗 ( Xing et al ., 2020 ; Andersen Onofre et al ., 2021 ). We calculated a cropland connectivity risk index (CCRI) ( Xing et al ., 2020 ) for each location with crop production in the region as a measure of its likely importance in epidemic or invasion networks. The CCRI was the weighted mean of four network metrics (node centralities), evaluated using the geohabnet package v2.0.0 ( Keshav et al ., 2024 ) in R v4.4.0 ( R Core Team, 2024 ), averaged across the results from a sensitivity analyses for a range of dispersal parameters, as described by Mouafo-Tchinda et al . (2024) . Two CCRI formulations were used to capture key facets of node connectivity. The first CCRI considered (“CCRI-betweenness”) was the mean of betweenness centrality weighted by 1/2, node strength weighted by 1/6, eigenvector centrality weighted by 1/6, and the sum of nearest neighbors’ degrees weighted by 1/6. The second, “CCRI-closeness”, considered 1/2 closeness centrality instead of betweenness centrality, with the same weights for the other metrics. We emphasized betweenness or closeness centrality to account for, respectively, the importance of a node as a bridge between parts of the network and the relative facility with which a location can be reached from all other locations in the network. Habitat connectivity analysis using the geohabnet package ( Keshav et al ., 2024 ) is part of the R2M toolbox for rapid risk assessment supporting mitigation of pathogens and pests ( garrettlab.com/r2m ). Other examples of applications of habitat connectivity and R2M tools, in addition to those discussed elsewhere in this paper, illustrate how these tools can be used in an integrated analysis ( Andersen Onofre et al ., 2019 ; Garrett, 2021 ; Etherton et al ., 2023 ; Mouafo-Tchinda et al ., 2024 ; Etherton et al ., 2025 ). 2.1.2. Associations among altitude, crop landscape structure, and climate We characterized associations among altitude, crop landscape structure, and climate in the Great Lakes region. The map of altitudes was retrieved from the Shuttle Radar Topography Mission (SRTM) database, using the geodata package v0.5-8 ( Hijmans et al ., 2023 ) in R. We also retrieved the 19 bioclimatic variables for the region from Worldclim database v2.1 ( Fick and Hijmans, 2017 ) using the geodata package version 0.5-8 ( Hijmans et al ., 2023 ) and the terra package v1.8-0 ( Hijmans, 2024 ) in R. For crop landscape structure variables, we used the raster maps of cropland density and connectivity (CCRI-closeness) generated in section 2.1.1 , using the terra package ( Hijmans et al ., 2022 ). We assessed the associations between altitude, bioclimatic variables, and cropland structure by calculating the Pearson correlations between these variables using the metan package v1.18.0 ( Olivoto and Lúcio, 2020 ) in R. 2.2. Pathogen and pest communities 2.2.1. Field sampling and spatiotemporal structure of pathogen and pest communities For each crop species, we identified fields for sampling along four survey transects in Rwanda and Burundi, each transect representing an altitude gradient. In Burundi, the field surveys included cropland around Lake Tanganyika in the Rusizi watershed, representing the Bubanza, Bujumbura Rural, Cibitoke, and Muramvya provinces ( Fig. 1C ). In Rwanda, the field surveys were in major cropland around Lake Kivu in the Ruhengeri watershed, which includes part of the Northern and Western provinces ( Fig. 1C ). The survey locations represented major crop landscapes in both countries, encompassing a mixture of cropping systems and a range in altitude from 900 to 2800 m.a.s.l. ( Fig. 1a-b ) ( Kroschel et al ., 2014 ). Fields for sampling were selected at regular intervals along the transects regardless of whether diseases or pest damage were apparent. The number of fields sampled in Burundi and Rwanda was 189 and 103 for banana, 78 and 116 for potato, and 107 and 104 for sweetpotato, respectively. For cassava, sampling was only carried out in 50 fields in Burundi, with the second Burundi survey not conducted due to civil unrest, and Rwanda was not sampled because few farmers grew cassava in the Rwanda study area. Download figure Open in new tab Fig. 1. Location of fields of banana, cassava, potato, and sweetpotato where pathogen and pest surveys were carried out in Africa (A), in the Great Lakes region (B), in transects across altitude gradients, in Rwanda and Burundi (C). White lines in (C) indicate the province boundaries of Rwanda and yellow lines indicate the province boundaries of Burundi. Crosses and circles represent locations surveyed during the dry and rainy season, respectively. The complete surveillance effort included 412 crop fields. Sampling was conducted in each watershed during two cropping seasons, dry and rainy. In both countries, two rainy seasons occur: one from March to May and the other from September to December (WorldData.info, 2024). The dry seasons occur from June to August and January to February (WorldData.info, 2024). For each field sampled, we evaluated crop-specific pest or disease intensity ( Table 2 ) based on prevalence in the region, disease severity, infestation rate and intensity, and/or the number of pests, using methods as described by Nakato et al . (2023) , Okonya et al . (2021) , and Legg et al . (2009). A total of 27 economically important pathogens and pests in the region were included in this study, including 6 for banana, 6 for cassava, 7 for potato and 8 for sweetpotato ( Table 2 ). Here prevalence refers to the presence or absence of a pest or disease in a field, while disease severity, pest infestation intensity or infestation rate is the percentage of leaves or plants damaged by a disease or pest. For infestation rate, infestation intensity, and severity, twenty plants per field, selected in a systematic pattern, were assessed for visual symptoms of damage by pests and diseases. The percentage of plants infested was noted as the infestation rate. The percentage of symptomatic leaves per plant was noted as infestation intensity for pests and severity for diseases. For damage by some virus diseases and pests such as cassava mosaic disease (CMD) and cassava green mite (CGM), crop damage was scored from 1 (representing no symptoms) to 5 (the most severe) ( Sseruwagi et al ., 2004 ; Legg et al ., 2009 ; Okonya et al ., 2021 ). The disease or pest damage score measures the severity or infestation intensity in a field, based on visual assessments and a predefined scale. Thirty plants were randomly sampled at regular intervals along the season and damage was assessed by scoring each plant on a scale of 1 to 5 as described by Legg et al . (2009) . View this table: View inline View popup Download powerpoint Table 1. Support vector machine (SVM) and random forest (RF) performance for predicting pathogen and pest levels in smallholder farms in Burundi and Rwanda, for selected species with higher model performance. The response variable was severity for the pathogens and infestation intensity for pests, except for infestation rate for sweetpotato aphids. View this table: View inline View popup Table 2. Correlation between altitude and pathogen and pest prevalence, severity, and infestation in Burundi and Rwanda (PRV: prevalence, SEV: severity, Rate: infestation rate, INF: infestation intensity, NUM: number of pests). For banana, fields were surveyed in Burundi in March 2015 (during the rainy season) and in July 2016 (dry season), while fields were surveyed in Rwanda in July 2015 (dry season) and in November 2016 (rainy season). Cassava fields were surveyed only once in Burundi in March 2015 (rainy season), for reasons described above. For potato, field sampling was conducted in Burundi in July 2016 (dry season) and December 2017 (rainy season), and in Rwanda in June-July 2015 (dry season) and November-December 2017 (rainy season). For sweetpotato, the field surveys took place in Burundi in November-December 2015 (rainy season) and August 2016 (dry season), and in Rwanda in July 2015 (dry season) and November 2017 (rainy season). The irregular dates for these crop-field surveys were due in part to the civil unrest prevailing in the region at the time the study was conducted. 2.2.2. Quantifying patterns of field-scale association of crop pathogens and pests We evaluated the patterns of association of pathogens and pests based on prevalence and severity/infestation using non-metric multidimensional scaling (NMDS) ( Kruskal, 1964 ) and heatmaps ( Pryke et al ., 2007 ) for visualization. We initially calculated five measures of dissimilarity (Bray, Manhattan, Jaccard, Chao, and robust-Aitchison) in the NMDS analysis using the ggvegan package v0.1.999 ( Simpson, 2020 ) in R. We selected the Bray distance in the final analysis due to its highest goodness of fit. A heatmap was prepared using the metan package v1.18.0 ( Olivoto and Lúcio, 2020 ) in R to map the association of pathogens and pests based on prevalence and severity/infestation rates. 2.3. Prediction of pathogen and pest damage and geographic range shifts 2.3.1. Machine learning to predict pathogen and pest levels Weather, climate, and geographic variables were used to predict severity or infestation rates/intensity, comparing machine learning algorithms. We retrieved weather variables (monthly, quarterly, and annual temperature, precipitation, and relative humidity) from the NASAPOWER database using the nasapower package v4.0.10 ( Sparks, 2018 ) in R. The climate variables included the 19 bioclimatic variables, solar radiation (kJ m-2 day-1), wind speed (m s-1), and water vapor pressure (kPa) described, along with the altitude variable, in section 2.1.2 . The CCRIs and cropland density were assembled as described in section 2.1.1 . Data were preprocessed to exclude collinearity between variables. A stratified split was then used to divide the data into training and test sets in an 8:2 ratio. We then trained ten machine-learning algorithms: logistic regression (GLM and GLMNET), support vector machine (SVM, with linear and radial basis function), neural network (NN), k-nearest neighbors (KNN), stochastic gradient boosting (GBM, generalized boosted modeling), bagged trees, classification and regression trees (CART), and random forest (RF). We compared the performance of these algorithms to predict the severity or infestation of each pathogen and pest using the caret package v6.0-94 ( Kuhn, 2008 ) in R. We selected SVM (with radial basis function) and RF for further analyses based on their having the highest performance in terms of MAE (mean absolute error), RMSE (root mean squared error) and R-squared ( Kaundal et al ., 2006 ). In the next stage using the SVM and RF models, we executed a 10-fold cross-validation with ten repetitions to prevent overfitting the training model and used a grid search for hyperparameter tuning. The tuning process was driven by trainControl, with RMSE-based model evaluation to prioritize accuracy. Hyperparameter tuning for SVM included C (regularization parameter) and Gamma (kernel coefficient), both with a tuneLength of 10. Hyperparameter tuning for RF included mtry (the number of predictors randomly selected at each division) with a tuneLength of 10. We trained the model using the train function in the caret package and estimated the performance parameters MAE, RMSE, and R-squared. We used the resulting model to predict severity or infestation rate/intensity for the test data and estimated the performance parameters (RMSE and R-squared). 2.3.2. Pathogen and pest communities across altitudinal gradients To study the geographic range of pathogen and pest communities associated with the four crops in the sample transects in Rwanda and Burundi, we evaluated how pathogen and pest levels (prevalence, disease severity or infestation rate/intensity) changed with altitude across the sampling locations. We also evaluated the pathogen and pest richness (number of pathogens and pests present from those included in the sampling) in communities during the dry and rainy seasons, across the altitude gradients. The associations between prevalence, infestation/severity and richness of pathogens and pests across altitude gradients were evaluated based on Pearson’s correlation. 2.3.3. Climate change: impact on regional temperature patterns We assessed potential climate shifts along the altitude gradients in the Great Lakes region, using current and future temperature data from WorldClim database v2.1 ( Fick and Hijmans, 2017 ). We used future climatic variables from three global circulation models (GCMs: GISS-E2-1-G, MIROC6 and CNRM-CM6-1), two Shared Socioeconomic Pathways (SSP1-2.6 and SSP5-8.5), and four future time-periods: 2021-2040, 2041-2060, 2061-2080, and 2081-2100. We selected these GCM modeling groups because of their equilibrium climate sensitivity (ECS) ( Hausfather, 2019 ). ECS indicates the severity of future warming impacts. CNRM-CM6-1 assumes a high value of ECS, 4.9 C, while GISS-E2-1-G and MIROC6 assumes low values of ECS, 2.7 and 2.6 respectively ( Hausfather, 2019 ). The two SSPs used in this study represent different emission pathways by the year 2100: sustainability (SSP1-2.6) and fossil-fuel development (SSP5-8.5) emission scenarios featured by the radiation forcing of 2.6 and 8.5 W/m², respectively ( Pinnegar et al ., 2021 ). We calculated the average annual temperature across the three selected GCMs for each future time-period and SSP combination. We used the resulting density plots to assess potential shifts in the distribution of annual mean temperature at low (1500 m.a.s.l.) altitudes, using the ggplot2 package v3.5.0 ( Wickham, 2016 ) in R. 3. Results 3.1. Landscape structures influencing pathogen and pest risk 3.1.1. Identifying locations with high cropland connectivity The analysis of cropland connectivity in the Great Lakes region (i) quantifies how the geographic distribution of crop production could influence the potential spread of crop-specific pathogens and pests and (ii) identifies candidate priority locations for pathogen and pest surveillance for each food security crop ( Fig. 2 ). These maps of cropland connectivity highlight locations in the region that are likely important for understanding and managing the risk of crop-specific pathogens and pests. Download figure Open in new tab Fig. 2. Maps of a cropland connectivity risk index (CCRI-betweenness) for four food security crops in (A) the Great Lakes region of Africa and (B) in a focused geographical extent in Rwanda and Burundi. The CCRI indicates how important a location is likely to be for potential spread of a host-specific pathogen or pest, based on the network of host availability. The CCRI is a weighted mean of four network metrics, including betweenness centrality in the results shown here. In (A), dark blue lines indicate national boundaries. In (B), blue lines indicate the provinces of Rwanda and green lines indicate the provinces of Burundi. In these maps of cropland connectivity, a higher CCRI-betweenness ( Fig. 2A ) or CCRI-closeness indicates the potential to play an important role in the establishment and spread of crop-specific pathogens and pests affecting these crops. Locations with high CCRI are candidate priorities for surveillance and mitigation programs. Although CCRI-betweenness and CCRI-closeness may not yield identical CCRI values, they tended to highlight the same potential locations for prioritization ( Fig. 2A ). For banana, locations in Rwanda, Burundi, the DR Congo, Uganda, Tanzania, Ethiopia, Malawi, and Mozambique have a high CCRI ( Fig. 2 ). Locations in the southern, eastern, and northern provinces of Rwanda have the highest CCRI-betweenness ( Fig. 2A and 2B). For cassava, several locations across the region have a high CCRI. In Burundi, locations in the Cibitoke, Bubanza, Kirundo, Ngozi, Bujumbura Rural and Gitega provinces have the highest CCRI ( Fig. 2A and B ). For potato, locations in Rwanda, Burundi, the DR Congo, Uganda, Tanzania, Ethiopia, and Malawi have high CCRI-betweenness ( Fig. 2 ). These maps also show that locations in the northern, western, and southern provinces of Rwanda have the highest CCRI-betweenness ( Fig. 2A and B ). For sweetpotato, locations in Rwanda, Burundi, Uganda, Tanzania, Malawi, and Ethiopia have a high CCRI ( Fig. 2 ). These maps indicate the important role that Rwanda and Burundi play for understanding and mitigating the risks associated with the spread of pathogens and pests of these crops. Maps of cropland connectivity ( Fig. 2B ) also illustrate the potential roles of the locations we selected for field sampling crop pathogens and pests in the Ruhengeri watershed in Rwanda and in the Rusizi watershed in Burundi. The connectivity of these landscapes varies from crop to crop. Some sampled locations had high cropland connectivity (e.g., cassava and sweetpotato in Burundi transects and potato in Rwanda transects) compared to other locations. Overall, the surveyed locations around Lake Kivu in Rwanda and Lake Tanganyika in Burundi, as well as other locations in the two countries, are likely to be important candidate locations for surveillance for the establishment and spread of banana, cassava, potato, and sweetpotato pathogens and pests in both countries, based on cropland connectivity ( Fig. 2B ). 3.1.2. Associations among altitude, landscape structure and climate The associations among (a) altitude, (b) crop landscape structure, and (c) climate variables such as temperature and precipitation, were analyzed for the Great Lakes region. Higher temperatures were generally associated with lower altitudes. There was a strong negative correlation between altitude and most temperature-related bioclimatic variables (Tem, Tmin, TColQ, TDriQ, TWetQ, TWarQ, Tmax, TColdQ). In contrast, altitude was weakly positively correlated with some precipitation-related bioclimatic variables such as PrecS, PWetQ, and PWetM, and negatively correlated with others, including Prec, PDriQ, PWarQ, and PColQ ( Fig. 3 ). In this region, higher cropland density and cropland connectivity of banana, cassava, potato, and sweetpotato ( Fig. 3 ) were at higher altitudes (except cassava density). Cropland connectivity of these crop species had a stronger positive correlation with altitude compared to cropland density. Cropland density and cropland connectivity for each of these four crops were negatively correlated with most temperature-related bioclimatic variables and positively correlated with most precipitation-related bioclimatic variables ( Fig. 3 ). Download figure Open in new tab Fig. 3 Associations between altitude, bioclimatic variables, and crop landscape structure (crop density and cropland connectivity) in the Great Lakes region of Africa, based on Pearson’s correlation. Strong positive correlations are in dark blue, while strong negative correlations are in dark red. White boxes without labels indicate no association. 3.2. Pathogen and pest communities 3.2.1. Spatiotemporal structure of pathogen and pest communities Most of the targeted pathogens and pests were found in Burundi and Rwanda in both crop-growing seasons ( Fig. 4 ). Sweetpotato mealybugs, sweetpotato butterfly ( Acraea acerata ) and sweetpotato aphids were observed only in Burundi ( Fig. 4 ). The observed prevalence of each pest or disease often varied by country, crop, and season. In both countries and seasons, the highest prevalence was observed for Ralstonia solanacearum and virus symptoms in potato, virus symptoms in sweetpotato, and Cosmopolites sordidus in banana. Download figure Open in new tab Fig. 4. Seasonal distribution of observed pathogens and pests in smallholder production of banana, potato and sweetpotato in Burundi and Rwanda. (Cassava data are not represented here). The length of the bars represents the prevalence for both seasons (where a total of 2 indicates presence in every field in both seasons). Abbreviations: Banana bunchy top virus (BBTV), Cosmopolites sordidus (CS), Ralstonia solanacearum (RS), sweetpotato virus (SPV), and Xanthomonas wilt of banana (XW). 3.2.2. Associations among pathogens and pests The NMDS analysis identified a limited set of pathogens/pests whose levels were associated, although some were common enough that they frequently co-occurred with others ( Fig. 5 ). A few pathogens and pests were associated in terms of prevalence and severity/infestation in the crop fields. Among banana diseases/pests, Fusarium wilt, Xanthomonas wilt and plant parasitic nematodes were associated in terms of both prevalence and severity ( Fig. 5 ). In cassava, all pathogens and pests except B. tabaci were associated in terms of prevalence. However, B. tabaci was slightly associated with CMV in terms of infestation, which is expected as B. tabaci vectors CMV ( Fig. 5 ). In potato, viruses and R. solanacearum were associated in terms of both prevalence and severity/infestation ( Fig. 5 ). In sweetpotato fields, B . tabaci and sweetpotato aphids were associated, having similar infestation levels across fields. Download figure Open in new tab Fig. 5. Relationships among economically important pathogens and pests and altitude in four major crops in Burundi and Rwanda, using non-metric multidimensional scaling (NMDS) based on Bray’s distance dissimilarities of disease severity or infestation intensity. Each dot represents a surveyed crop field. A line (or vector) indicates whether (line direction) and how strongly (line length) the pathogen or pest is associated with others. Nearly parallel lines in the same quadrant indicate pathogens and pests often co-occur in a crop. (Cassava data was collected only in Burundi for one season). Abbreviations: Acraea acerata (AA), Alternaria bataticola (AB), Banana bunchy top virus (BBTV), Bemisia tabaci (cassava, potato or sweetpotato whitefly; BT), cassava bacterial blight (CBB), Cassava brown streak virus (CBSV), cassava green mite (CGM), cassava mealybug (CM), Cassava mosaic virus (CMV), Cylas brunneus (CB), Cylas puncticollis (CP), Cosmopolites sordidus (CS), Fusarium wilt of banana (FW), leaf miner fly (LMF), Myzus persicae (MP), Phytophthora infestans (PI), plant parasitic nematodes (PPN), Pentalonia nigronervosa (PN), Phthorimaea operculella (PO), potato viruses (PV), Ralstonia solanacearum (RS), Synanthedon sp. (clearwing moth; Ssp), sweetpotato mealybug (SPM), sweetpotato virus (SPV), sweetpotato weevils (SPW), and Xanthomonas wilt of banana (XW). Analyses of associations between pairs of pathogens/pests provide a more detailed perspective on which syndromes are common. In banana, some pairs like C. sordidus and P. nigronervosa , plant parasitic nematodes and Fusarium wilt, plant parasitic nematodes and P. nigronervosa , P. nigronervosa and Xanthomonas wilt, and Fusarium wilt and Xanthomonas wilt were positively correlated (p < 0.05) in terms of both prevalence and severity/infestation, while other pairs were only positively correlated (p < 0.05) in terms of prevalence: C. sordidus and plant parasitic nematodes, plant parasitic nematodes and Xanthomonas wilt, plant parasitic nematodes and BBTV, P. nigronervosa and BBTV, and BBTV and Xanthomonas wilt ( Fig. 6 ). Download figure Open in new tab Fig. 6. Association of pathogens and pests in banana, cassava, potato, and sweetpotato in smallholder farms in Burundi and Rwanda, based on Pearson’s correlation of prevalence. (Cassava data was collected only in Burundi for one season). Abbreviations: Acraea acerata (AA), Alternaria bataticola (AB), Banana bunchy top virus (BBTV), Bemisia tabaci (cassava, potato or sweetpotato whitefly; BT), cassava bacterial blight (CBB), Cassava brown streak virus (CBSV), cassava green mite (CGM), cassava mealybug (CM), Cassava mosaic virus (CMV), Cylas brunneus (CB), Cylas puncticollis (CP), Cosmopolites sordidus (CS), Fusarium wilt of banana (FW), leaf miner fly (LMF), Myzus persicae (MP), Phytophthora infestans (PI), plant parasitic nematodes (PPN), Pentalonia nigronervosa (PN), Phthorimaea operculella (PO), potato viruses (PV), Ralstonia solanacearum (RS), Synanthedon sp. (clearwing moth; Ssp), sweetpotato mealybug (SPM), sweetpotato virus (SPV), sweetpotato weevil (SPW), and Xanthomonas wilt of banana (XW). In cassava, CBSV and mealybug were positively correlated (r = 0.69, p < 0.001 in terms of prevalence but negatively correlated (r = -0.44, p < 0.01) in terms of severity/infestation ( Fig. 6 ). Only B. tabaci and CMV were positively correlated in terms of severity/infestation (r = 0.64, p < 0.001) while other pathogens and pests were positively correlated in terms of prevalence (p < 0.05). In potato, viruses and R. solanacearum was positively correlated in terms of prevalence (r = 0.33, p < 0.001) and severity (r = 0.21, p < 0.01), while P. infestans and M. persicae , and leaf miner fly and M. persicae were negatively correlated (p < 0.05) in terms of prevalence and severity/infestation, and leaf miner fly and R. solanacearum were positively correlated (p < 0.05) in terms of prevalence but negatively correlated (p < 0.05) in terms of severity/infestation ( Fig. 6 ). In sweetpotato, C. brunneus and C. puncticollis were positively correlated in terms of prevalence (r = 0.58, p < 0.001) and infestation (r = 0.18, p < 0.01), while A. bataticola and weevils ( Cylas spp.) were negatively correlated in terms of prevalence (r = -0.16, p < 0.05) but positively correlated in terms of severity/infestation (r = 0.18, p < 0.05) ( Fig. 6 ). A. bataticola and mealybug, A. bataticola and aphids, A. bataticola and B. tabaci , A. bataticola and Acraea acerata , and aphids and Synanthedon sp. were negatively correlated in terms of prevalence ( Fig. 6 ). 3.3. Prediction of pathogen and pest damage and geographic range shifts 3.3.1. Machine learning to predict individual pathogen and pest levels We compared machine learning algorithms in terms of their ability to predict severity and infestation in the field samples at monthly, quarterly, and annual intervals ( Table 1 ). SVM and RF had the highest model performance for some pathogens/pests. For the SVM model, the performance characteristics across all pathogens and pests were 3-16% MAE for training, 7-20% and 4-26% RMSE, and 0.20-0.72 and 0.09-0.72 R-squared for training and testing, respectively, compared to the RF model with 3-17% MAE for training, 8-21% and 5-23% RMSE, and 0.23-0.72 and 0.12-0.73 R-squared for training and testing, respectively ( Table 1 ). Among the 27 pathogens and pests tested, the three with the highest predictive performance metrics per crop were reported in Table 1 . The predictive performance was good for certain pathogens and pests, but poor for others ( Table 1 ). As an example, SVM and RF performed well in predicting the monthly, quarterly, and annual infestation intensity of leaf miner fly in this region, explaining 69-71% of the variability in the data (R-squared) with low errors. The RMSE was 17-18% during training and 16-17% during prediction, demonstrating that both models generalized well to unseen data. In addition, the MAE of 11-13% indicated a low average error in training. The tight alignment between training and prediction metrics showed the robustness, stability and reliability of these models to predict leaf miner fly infestation intensity. Pathogens and pests differed in the importance of predictor variables. Contemporary weather variables, such as monthly temperature, precipitation, and relative humidity (extracted through NASAPOWER), generally had high contribution scores. Cropland structure variables also played significant roles, with contribution scores reaching 100 for cropland density, 81 for cropland connectivity, and 69 for altitude. For example, the key variables for predicting leaf miner fly infestation intensity (monthly) using SVM were monthly temperature, precipitation, and relative humidity, precipitation in the driest month, cropland connectivity, cropland density and annual temperature range (i.e. difference between the maximum temperature of warmest month and the minimum temperature of coldest month). 3.3.2. Pathogen and pest communities across altitudinal gradients For each crop, pathogen and pest communities differed at low versus high altitudes ( Table 2 ). Although there was not strong evidence for differences across the altitudinal gradients for 5 out of 27 pathogens and pests (correlations with p > 0.05 in Table 1 ), likely due in some cases to high variability across samples, there was evidence for differences in some pathogens and pests that distinguished the pathogen/pest communities (correlations in Table 2 with lower p-values). For instance, crop production at higher altitudes (>1,500 m.a.s.l) had more Myzus persicae , Phthorimaea operculella , and cassava green mites. At lower altitudes, crop fields had more Fusarium wilt and Xanthomonas wilt of bananas, P. infestans , A. acerata and Bemisia tabaci ( Table 2 ). The distribution of these pathogens and pests across altitudes was supported in the analysis combining observations from Burundi and Rwanda; overall, for banana, potato, and sweetpotato, there was evidence that 44% of the targeted pathogens and pests were more common at low altitudes, and that 17% were more common at higher altitudes. When analyzing data for Rwanda and Burundi separately, differences across altitudes for pathogens and pests were specific to the measure being used (e.g., disease intensity; Table 2 ) or the country. For example, potato fields at high altitudes had a higher prevalence of potato leaf miner fly, while there was not evidence for a difference in the infestation intensity of this pest across the altitudinal gradient. Similarly, there was stronger evidence that altitude was negatively correlated with the prevalence and severity of banana Fusarium wilt in Burundi (p < 0.001) compared to Rwanda (p = 0.069 for severity and p = 0.85 for prevalence). At the community level, potato fields in Rwanda and cassava fields in Burundi had higher pathogen and pest richness at higher altitudes, while richer pathogen and pest communities were found in banana fields at lower altitudes. 3.3.3. Climate change and impact on regional temperature patterns The frequency distribution of annual temperatures across altitudes in the Great Lakes region illustrates the transformative impact of climate change ( Fig. 7 ). A progressive shift in temperature distribution is expected, highlighting how currently cooler climates at higher altitudes are likely to experience much higher temperatures ( Fig. 7 ). A general warming trend is projected at both high and low altitudes. There is a greater temperature increase in the SSP5-8.5 scenario than in the SSP1-2.6 scenario. Currently, the mean annual temperature in the region ranges between 17-30°C at low altitudes, with the mode around 24°C, and between 10-22°C at higher altitudes, with the mode around 17°C. As climate change progresses, and depending on socio-economic pathways, this temperature landscape is likely to change substantially. Under the SSP1-2.6 scenario, by 2090 (2081-2100), temperatures are expected to range between 20-32°C at low altitudes, with the mode around 26°C, and between 10-25°C at high altitudes, with mode 20°C. The high-emission scenario (SSP5-8.5) predicts that temperatures will range between 22-36°C at low altitudes, with the mode around 28°C, and between 12-30°C at higher altitudes, with the mode around 25°C. Download figure Open in new tab Fig. 7. Shift in annual temperature range at lower and higher elevations in the Great Lake region of Africa. Density is the probability density or likelihood of temperature rates occurring at locations in the Great Lakes region of Africa. TP: time period of the projections (while the ‘Current’ reference remains the same). SSP: Shared Socioeconomic Pathway, sustainability (SSP1-2.6) and fossil-fueled development (SSP5-8.5) emission scenarios. Here, high altitude refers to locations above 1500 m.a.s.l. and low altitude those below 1500 m.a.s.l. These projections suggest future climate conditions at higher altitudes could become more favorable for the 44% of pathogens and pests we evaluated that currently are more common at lower altitudes ( section 3.3.2 ). Likewise, pathogens and pests currently more common at higher altitudes (17% of those evaluated in section 3.3.2 ) may become less common as the locations warm in the future. These percentages are based on banana, potato, and sweetpotato, the three crops for which we had the most complete data sets for analysis in predictive models based on field data collected to allow direct comparisons across pathogens and pests. 4. Discussion This study identified environmental predictors for a set of important pathogens and pests in tropical agroecosystems key to food security. We identified candidate surveillance locations based on cropland (host) connectivity, a starting point for understanding geographic spread and establishment of host-specific pathogens and pests. Based on the current geographic and seasonal distribution of 27 economically important pathogens and pests across climate gradients, we evaluated future climate scenarios. The 44% of the pathogens and pests currently more abundant at lower altitudes could become more common at higher altitudes under climate change, while the 17% more abundant at higher altitudes could become less common overall in the future. Identifying differences in pathogen and pest communities and their predictors contributes to a more complete understanding of agroecosystems, which can be translated to inform management strategies, as we discuss below. 4.1. Pathogen and pest risk across landscape structures The cropland connectivity analysis provides a first approximation to the geographic distribution of pathogen and pest risk, based on host availability. This analysis identified candidate priority locations for surveillance and mitigation of the potential spread of host-specific pathogens and pests based on networks of host availability ( Fig. 2 ). Analysis of cropland connectivity provides a baseline for quantifying invasion risks for pathogens and pests at a national or global scale ( Andersen Onofre et al ., 2021 ; Mouafo-Tchinda et al ., 2024 ; Etherton et al ., 2025 ). In the Great Lakes region, locations in Rwanda and Burundi had a high CCRI for all four food-security crops, highlighting their vulnerability to pathogen and pest spread in the absence of phytosanitary measures. Maps of cropland connectivity can inform monitoring of outbreaks of introduced or emerging diseases and pests by national plant protection organizations (NPPOs). Ongoing improvements to understanding other components of pathogen and pest risk – such as the effects of climate change, local trade, genetic resistance deployment, and other management adoption – could be integrated in more complete habitat connectivity analyses as additional resources become available ( Keshav et al ., 2024 ). 4.2. Pathogen and pest communities Analysis of the on-farm distribution of 27 pathogens and pests highlights how these communities change along climate gradients, in terms of seasonal abundance ( Fig. 4 ), community structure ( Fig. 5 ), and interspecific pathogen/pest associations ( Fig. 6 ). Previous studies have typically addressed how individual tropical pathogens or pests are distributed along climate gradients ( Ebregt et al ., 2005 ; Okonya and Kroschel, 2013 ; Nyang’au et al ., 2021 ); a community perspective provides a more realistic scenario for understanding regional pathogen and pest management ( Savary et al ., 2000 ; Were et al ., 2013 ; Makiola et al ., 2021 ). The study addresses this knowledge gap by identifying which pathogens/pests are most prevalent across seasons, which pathogen/pest occurrences are associated, and how climate gradients are likely to influence disease severity and pest infestation. Evaluating the interactions and cumulative effects of multiple pathogens/pests in specific production environments can increase our understanding of injury profiles ( Savary et al ., 2000 ), the cumulative damage to plants caused by the set of biotic stressors including insects, pathogens and weeds. Our findings highlight injury profiles for food security crops across climate gradients. Some associations are already well-known; as expected, BBTV and its vector P. nigronervosa were positively correlated ( Hu et al ., 1996 ; Blomme et al ., 2020 ). BBTV has been present in Africa for decades, its spread continues to have an impact on new regions, and its management remains a major challenge ( Rybicki, 2015 ; Ngatat et al ., 2024 ; Ocimati et al ., 2024 ). We also found strong associations between certain pathogens and pests; Fusarium wilt, Xanthomonas wilt and plant parasitic nematodes were associated in banana, viruses and R. solanacearum in potato, and B. tabaci and aphids in sweetpotato. These pathogen/pest associations underscore the need for integrated management strategies that address the spread and combined impacts of multiple pathogens and pests on plants ( Mouafo-Tchinda et al ., 2022 ; Alcalá Briseño et al ., 2023 ). 4.3. Climate-based predictions of pathogen and pest levels In the Great Lakes region, temperature generally decreased with increasing altitude ( Fig. 3 ), where higher temperatures were associated with an increase in levels of some pathogens and pests and a decrease in others ( Table 2 ). Whiteflies were more common at low altitudes for cassava, potato and sweetpotato. Aphids were more common at low altitudes for banana, but more common at high altitudes for potato. The observed differences in the altitudinal distribution of these pests could be attributed to their species composition, host specificity, and the influence of species-specific ecological adaptations ( Aleuy and Kutz, 2020 ; McCulloch and Waters, 2023 ). Aphid communities included different species, with P. nigronervosa more frequent in low-altitude plants, and M. persicae more frequent in high-altitude plants ( Malumphy et al ., 2017 ; Wells and Clark, 2019 ; Wieczorek et al ., 2019 ; Aleuy and Kutz, 2020 ). Weather variables are well known to be useful for predicting disease severity ( Kaundal et al ., 2006 ; Carisse and Fall, 2021 ; Fenu and Malloci, 2021 ). Cropland geography variables (cropland density, connectivity), altitude, and climate variables (annual mean temperature, temperature annual range, and annual precipitation) were key for most of the pathogens and pests we evaluated. Model performance depends on data availability, the predictor variables, and the biological features of pathogens and pests ( Xiao et al ., 2018 ; Fenu and Malloci, 2021 ). The data from Rwanda and Burundi supported prediction of some pathogens, while others, like BBTV, would require more study. The models effectively predicted leaf miner fly infestation intensity, accounting for 69-71% of the variability with low training and prediction errors. These models could serve as a prototype for predictive models of leaf miner fly infestation intensity in this region ( Table 1 ). Temperature influences the life cycle, reproduction, and migration patterns of pathogens and pests ( Awmack et al ., 1997 ; Yamamura and Kiritani, 1998 ; Mouafo-Tchinda et al ., 2021 ). In response to regional warming, certain pathogens and pests may become more common at specific altitudes ( Poveda et al ., 2012 ; Datta et al ., 2017 ; Blomme et al ., 2020 ), which will require adaptation of pathogen and pest management to future conditions. Several studies suggest that climate change shifts some species ranges to higher altitudes ( Bale et al ., 2002 ; Colwell et al ., 2008 ; Bebber et al ., 2013 ), and the new results from this study clarify the specific effects on individual pathogens and pests in Rwanda and Burundi for key food security crops. Pathogen/pest richness in banana decreased with altitude, while it increased in cassava. In addition, the severity of certain pathogens/diseases such as BBTV, CMV, P. infestans and Fusarium wilt decreased with altitude, while the severity or infestation of A. bataticola and aphids in sweetpotato increased. In general, we found evidence that 44% of the pathogens and pests currently more common at low altitude in banana, potato, and sweetpotato could shift to higher altitudes under climate change, based on climate matching. Specifically, those expected to become more common at higher altitudes are Xanthomonas wilt, Fusarium wilt, BBTV, and P. nigronervosa , in banana; sweetpotato mealybug, A. acerata , sweetpotato weevils in sweetpotato; and B. tabaci in potato and sweetpotato. The 17% of pathogens and pests currently more common at high altitudes could become less common. Cropland connectivity was positively correlated with altitude and was a key predictor of severity/infestation of some pathogens and pests. This suggests that food-security crops at higher altitudes may experience substantial flows of pathogens and pests, amplifying the risk of damage. Customizing risk assessments for these specific pest/pathogen communities across altitudes could make management strategies better adapted to the challenges posed by climate change for each crop in the region. 4.4. Translating agroecosystem understanding of pathogens and pests to support current and future agriculture Historically, global and national research and extension efforts have been employed to manage diseases and pests in the Great Lakes region ( RTB, 2013 ; Kroschel et al ., 2014 ). Integrated management systems combining cultural control, surveillance, awareness creation and involving policy makers and farmers have been adopted ( Legg et al ., 2006 ; Gotor et al ., 2022 ). There has been support for active management programs where clean planting material is provided to farmers to encourage them to destroy diseased plants. Efforts have been made to create awareness and empower farmers to recognize and destroy diseased plants. A challenge for these efforts is limited resources to support ongoing, sustainable plans by governments and national programs, especially after project lifespans are over. The results of this new study can help to inform prioritization and efficient use of available resources. Currently, for many food security crops, efforts are underway to improve seed health, obtain resistance to key pathogens and pests, and adapt these improved technologies and techniques for use in humanitarian settings ( RTB, 2013 ; Andrade-Piedra et al ., 2020 ; Andrade-Piedra et al ., 2023 ). Pathogen and pest risk analysis can help to ensure the economic, social, and environmental sustainability of crop production ( Kaundal et al ., 2006 ; Fall et al ., 2016 ; Carisse and Fall, 2021 ). Image analysis apps such as the PlantVillage Nuru app and the Tumaini app can help smallholder farmers in the region diagnose plant damage caused by pests and diseases, and offer national plant protection organizations and extension staff a tool for surveillance and disease mapping ( Ramcharan et al ., 2019 ; Selvaraj et al ., 2019 ; Kreuze et al ., 2022 ). Predicting the timing of pathogen and pest threats would enable farmers to adjust management strategies in the short run and would support prioritization in research and breeding programs in the long run. Ongoing improvements in predicting pathogen and pest impacts depend on weather data availability, which is often limited in low-income and disaster-prone countries. Satellite platforms like NASA POWER offer a useful alternative ( Sparks, 2018 ), providing near-real-time weather data and allowing researchers in resource-limited settings to develop useful models without the need for extensive ground observations. Another challenge is detection of transient pests such as aphids, which move frequently and rapidly across fields ( Loxdale, 2018 ). Innovative sampling techniques, such as suction traps and environmental DNA (eDNA) analysis, could be helpful for monitoring pest populations more accurately ( Taberlet et al ., 2012 ; Poppinga et al ., 2016 ). However, implementing these advanced approaches in low-income and disaster-prone areas such as the Great Lakes region remains a challenge due to limited resources and technical capacity, requiring scalable and cost-effective solutions. In disaster-prone regions, humanitarian organizations and government agencies must prioritize key areas where abandonment or poor management of staple crops is particularly important to regional spread of pathogens and pests ( Etherton et al ., 2024 ; Mouafo-Tchinda et al ., 2024 ). This study identifies candidate priority locations for mitigation of the establishment and spread of pathogens and pests, characterizes pathogen and pest associations (injury profiles), and evaluates potential altitudinal shifts under the influence of climate change. Future research can build on these results by validating and refining the models for each pathogen/pest, and developing tools that growers and policymakers can easily implement. Ongoing improvement can increase the adaptability and effectiveness of plant health management strategies, a no-regrets adaptation strategy as agricultural systems face increasing challenges due to climate change. Competing interests The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements We appreciate support from the CGIAR Seed Equal Research Initiative, the CGIAR Roots Tubers and Bananas Research Program, USAID Bureau of Humanitarian Assistance (BHA) award number 720BHA22IO00136, the OneCGIAR Initiative on Plant Health, and the CGIAR Trust Fund ( www.cgiar.org/funders/ ); we thank all donors and organizations which globally support the work of CGIAR through their contributions to the CGIAR Trust Fund. We also appreciate support from USDA Animal and Plant Health Inspection Service (APHIS) Cooperative Agreements AP21PPQS&T00C195 and AP22PPQS&T00C133. The opinions expressed in this article are those of the authors and do not necessarily reflect the view of USAID BHA or USDA APHIS. Funder Information Declared CGIAR Seed Equal Research Initiative USAID BHA , 720BHA22IO00136 OneCGIAR Initiative on Plant Health USDA Animal and Plant Health Inspection Service (APHIS) , AP21PPQS&T00C195 , AP22PPQS&T00C133 Footnotes Updated author affiliations and contact information; Introduction updated for clarity; title modified References 1. ↵ Adhikari , U. , Nejadhashemi , A.P. , Woznicki , S.A ., 2015 . Climate change and eastern Africa: a review of impact on major crops . Food and Energy Security 4 , 110 – 132 . OpenUrl 2. ↵ Alcalá Briseño , R.I. , Batuman , O. , Brawner , J. , Cuellar , W.J. , Delaquis , E. , Etherton , B.A. , French-Monar , R.D. , Kreuze , J.F. , Navarrete , I. , Ogero , K. , et al. 2023 . Translating virome analyses to support biosecurity, on-farm management, and crop breeding . Frontiers in Plant Science 14 , 1056603 . OpenUrl PubMed 3. ↵ Aleuy , O.A. , Kutz , S ., 2020 . Adaptations, life-history traits and ecological mechanisms of parasites to survive extremes and environmental unpredictability in the face of climate change . International Journal for Parasitology: Parasites and Wildlife 12 , 308 – 317 . OpenUrl 4. ↵ Almekinders , C.J. , Walsh , S. , Jacobsen , K.S. , Andrade-Piedra , J.L. , McEwan , M.A. , de Haan , S. , Kumar , L. , Staver , C. , 2019 . Why interventions in the seed systems of roots, tubers and bananas crops do not reach their full potential . Food Secur . 11 , 23 – 42 . OpenUrl 5. ↵ Andersen Onofre , K.F. , Buddenhagen , C.E. , Rachkara , P. , Gibson , R. , Kalule , S. , Phillips , D. , Garrett , K.A. , 2019 . Modeling epidemics in seed systems and landscapes to guide management strategies: The case of sweet potato in northern Uganda . Phytopathology 109 , 1519 – 1532 . OpenUrl CrossRef PubMed 6. ↵ Andersen Onofre , K.F. , Forbes , G.A. , Andrade-Piedra , J.L. , Buddenhagen , C.E. , Fulton , J.C. , Gatto , M. , Khidesheli , Z. , Mdivani , R. , Xing , Y. , Garrett , K.A. , 2021 . An integrated seed health strategy and phytosanitary risk assessment: Potato in the Republic of Georgia . Agr Syst 191 , 103144 . OpenUrl 7. ↵ Andrade-Piedra , J. , Almekinders , C.J. , McEwan , M. , Kilwinger , F. , Mayanja , S. , Mulugo , L. , Delaquis , E. , Garrett , K. , Omondi , B.A. , Rajendran , S ., 2020 . User guide to the toolbox for working with root, tuber and banana seed systems. Lima, Peru. CGIAR Research Program on Roots , Tubers and Bananas. RTB User Guide . 8. ↵ Andrade-Piedra , J. , Dontsop , P. , Dunia , D. , Kulakow , P. , Harahagazwe , D. , Mouafo-Tchinda , R. , Omondi , A. , Ogero , K. , Rajendran , S. , McEwan , M. , 2023 . Tools4SeedSystems: Working towards resilience through root, tuber and banana crops in humanitarian settings . Capacity needs assessment for root, tuber and banana seed interventions in humanitarian settings: Cameroon and DRC. Technical Report . 65 . 9. ↵ Awmack , C. , Harrington , R. , Leather , S ., 1997 . Host plant effects on the performance of the aphid Aulacorthum solani (Kalt.)(Homoptera: Aphididae) at ambient and elevated CO2 . Global Change Biology 3 , 545 – 549 . OpenUrl CrossRef 10. ↵ Bale , J.S. , Masters , G.J. , Hodkinson , I.D. , Awmack , C. , Bezemer , T.M. , Brown , V.K. , Butterfield , J. , Buse , A. , Coulson , J.C. , Farrar , J ., 2002 . Herbivory in global climate change research: direct effects of rising temperature on insect herbivores . Global Change Biology 8 , 1 – 16 . OpenUrl CrossRef Web of Science 11. ↵ Bebber , D.P. , Ramotowski , M.A.T. , Gurr , S.J ., 2013 . Crop pests and pathogens move polewards in a warming world . Nature Climate Change 3 , 985 – 988 . OpenUrl 12. ↵ Blomme , G. , Ocimati , W. , Amato , S. , Felde , A.z. , Kamira , M. , Bumba , M. , Bahati , L. , Amini , D. , Ntamwira , J. , 2020 . Banana pest risk assessment along banana trade axes running from low to high altitude sites, in the Eastern DR Congo and in Burundi . African Journal of Agricultural Research 16 ( 9 ), 1253 – 1269 . OpenUrl CrossRef 13. ↵ Buddenhagen , C.E. , Xing , Y. , Andrade-Piedra , J.L. , Forbes , G.A. , Kromann , P. , Navarrete , I. , Thomas-Sharma , S. , Choudhury , R.A. , Andersen Onofre , K.F. , Schulte-Geldermann , E. , Etherton , B.A. , Plex Sula , A.I. , Garrett , K.A ., 2022 . Where to invest project efforts for greater benefit: A framework for management performance mapping with examples for potato seed health . Phytopathology 112 , 1431 – 1443 . OpenUrl CrossRef PubMed 14. ↵ Carisse , O. , Fall , M.L ., 2021 . Decision trees to forecast risks of strawberry powdery mildew caused by Podosphaera aphanis . Agriculture 11 , 29 . OpenUrl 15. ↵ Chevalier , M. , Broennimann , O. , Guisan , A ., 2024 . Climate change may reveal currently unavailable parts of species’ ecological niches . Nature Ecology & Evolution 8 , 1298 – 1310 . OpenUrl PubMed 16. ↵ Torquebiau , E Cilas , C. , Goebel , F.-R. , Babin , R. , Avelino , J ., 2016 . Tropical crop pests and diseases in a climate change setting—a few examples . In: Torquebiau , E . (Ed.), Climate change and agriculture worldwide , Springer , Dordrecht , pp. 73 – 82 . 17. ↵ Colwell , R.K. , Brehm , G. , Cardelús , C.L. , Gilman , A.C. , Longino , J.T ., 2008 . Global warming, elevational range shifts, and lowland biotic attrition in the wet tropics . Science 322 , 258 – 261 . OpenUrl Abstract / FREE Full Text 18. ↵ Datta , A. , Kühn , I. , Ahmad , M. , Michalski , S. , Auge , H ., 2017 . Processes affecting altitudinal distribution of invasive Ageratina adenophora in western Himalaya: The role of local adaptation and the importance of different life-cycle stages . PLOS One 12 , e0187708 . OpenUrl CrossRef PubMed 19. ↵ Ebregt , E. , Struik , P. , Odongo , B. , Abidin , P ., 2005 . Pest damage in sweet potato, groundnut and maize in north-eastern Uganda with special reference to damage by millipedes (Diplopoda) . NJAS-Wageningen Journal of Life Sciences 53 , 49 – 69 . OpenUrl CrossRef 20. ↵ Etherton , B. , Plex Sula , A. , Mouafo-Tchinda , R. , Kakuhenzire , R. , Kassaye , H. , Asfaw , F. , Kosmakos , V. , McCoy , R. , Xing , Y. , Yao , J ., 2025 . Translating Ethiopian potato seed networks: Identifying strategic intervention points for managing bacterial wilt and other diseases . Agr Syst 222 , 104167 . OpenUrl CrossRef 21. ↵ Etherton , B.A. , Choudhury , R. , Alcalá-Briseño , R. , Xing , Y. , Sulá , A.P. , Carrillo , D. , Wasielewski , J. , Stelinski , L. , Grogan , K. , Ballen , F. , 2023 . Are avocados toast? A framework to analyze decision-making for emerging epidemics, applied to laurel wilt . Agr Syst 206 , 103615 . OpenUrl CrossRef 22. ↵ Etherton , B.A. , Choudhury , R.A. , Alcalá Briseño , R.I. , Mouafo-Tchinda , R.A. , Plex Sulá , A.I. , Choudhury , M. , Adhikari , A. , Lei , S.L. , Kraisitudomsook , N. , Buritica , J.R ., 2024 . Disaster plant pathology: Smart solutions for threats to global plant health from natural and human-driven disasters . Phytopathology 114 , 855 – 868 . OpenUrl CrossRef PubMed 23. ↵ Fall , M.L. , Van der Heyden , H. , Carisse , O. , 2016 . A quantitative dynamic simulation of Bremia lactucae airborne conidia concentration above a lettuce canopy . PLoS One 11 , e0144573 . OpenUrl CrossRef PubMed 24. ↵ Fenu , G. , Malloci , F.M ., 2021 . Forecasting plant and crop disease: an explorative study on current algorithms . Big Data and Cognitive Computing 5 , 2 . OpenUrl 25. ↵ Ferdu , A. , Solbreck , C ., 2010 . Oviposition preference and larval performance of the sweet potato butterfly Acraea acerata on Ipomoea species in Ethiopia . Agricultural and Forest Entomology 12 , 161 – 168 . OpenUrl CrossRef 26. ↵ Fick , S.E. , Hijmans , R.J ., 2017 . WorldClim 2: new 1-km spatial resolution climate surfaces for global land areas . International Journal of Climatology 37 , 4302 – 4315 . OpenUrl CrossRef 27. ↵ Gallego-Tévar , B. , Gil-Martínez , M. , Perea , A. , Pérez-Ramos , I.M. , Gómez-Aparicio , L. , 2024 . Interactive Effects of Climate Change and Pathogens on Plant Performance: A Global Meta-Analysis . Global Change Biology 30 , e17535 . OpenUrl CrossRef PubMed 28. ↵ Garrett , K. , Bebber , D. , Etherton , B. , Gold , K. , Sulá , A.P. , Selvaraj , M.G ., 2022a . Climate change effects on pathogen emergence: artificial intelligence to translate big data for mitigation . Annual Review of Phytopathology 60 , 357 – 378 OpenUrl CrossRef PubMed 29. ↵ Garrett , K.A ., 2021 . Impact network analysis and the INA R package: Decision support for regional management interventions . Methods in Ecology and Evolution 12 , 1634 – 1647 . OpenUrl 30. ↵ Ziska , L Garrett , K.A. , Thomas-Sharma , S. , Forbes , G.A. , Hernandez Nopsa , J.F. , Plex Sulá , A.I ., 2022b . Climate change and plant pathogen invasions . In: Ziska , L . (Ed.), Invasive Species and Global Climate Change . CABI , pp. 22 – 49 . 31. ↵ Gotor , E. , Di Cori , V. , Pagnani , T. , Kikulwe , E. , Kozicka , M. , Caracciolo , F. , 2022 . Public and private investments for banana Xanthomonas Wilt control in Uganda: the economic feasibility for smallholder farmers. African Journal of Science, Technology , Innovation and Development 14 , 135 – 146 . OpenUrl 32. ↵ Hausfather , Z ., 2019 . CMIP6: the next generation of climate models explained . Carbon Brief . 33. ↵ Hijmans , R. , Barbosa , M. , Ghosh , A. , Mandel , A ., 2023 . geodata: Download geographic data . R package version 0.5. 34. ↵ Hijmans , R.J ., 2024 . terra: Spatial Data Analysis . R package version 1.8-0. 35. ↵ Hijmans , R.J. , Bivand , R. , Forner , K. , Ooms , J. , Pebesma , E. , Sumner , M.D. , 2022 . Package ‘terra’ . Maintainer: Vienna, Austria. 36. ↵ Hodkinson , I.D ., 2005 . Terrestrial insects along elevation gradients: species and community responses to altitude . Biological reviews 80 , 489 – 513 . OpenUrl CrossRef PubMed 37. ↵ Hu , J.S. , Wang , M. , Sether , D. , Xie , W. , Leonhardt , K.W ., 1996 . Use of polymerase chain reaction (PCR) to study transmission of banana bunchy top virus by the banana aphid ( Pentalonia nigronervosa ) . Annals of Applied Biology 128 , 55 – 64 . OpenUrl CrossRef 38. ↵ International Food Policy Research Institute , M., 2020 . Spatially-disaggregated crop production statistics data in Africa south of the Sahara for 2017 . Harvard Dataverse. 39. ↵ Jongejans , E. , Skarpaas , O. , Ferrari , M.J. , Long , E.S. , Dauer , J.T. , Schwarz , C.M. , Rauschert , E.S.J. , Jabbour , R. , Mortensen , D.A. , Isard , S.A. , Lieb , D.A. , Sezen , Z. , Hulting , A.G. , Shea , K ., 2015 . A unifying gravity framework for dispersal . Theoretical Ecology 8 , 207 – 223 . OpenUrl 40. ↵ Juroszek , P. , Bartsch , L. , Fontaine , J.F. , Racca , P. , Kleinhenz , B ., 2022 . Summary of the worldwide available crop disease risk simulation studies that were driven by climate change scenarios and published during the past 20 years . Plant Pathology 71 , 1815 – 1838 . OpenUrl 41. ↵ Kaundal , R. , Kapoor , A.S. , Raghava , G.P ., 2006 . Machine learning techniques in disease forecasting: a case study on rice blast prediction . BMC bioinformatics 7 , 1 – 16 . OpenUrl CrossRef PubMed Web of Science 42. ↵ Keshav , K. , Plex-Sula , A. , Garrett , K.A ., 2024 . geohabnet: Geographical Risk Analysis Based on Habitat Connectivity . R Package, Version 2.0.0. CRAN. 43. ↵ Kreuze , J. , Adewopo , J. , Selvaraj , M. , Mwanzia , L. , Kumar , P.L. , Cuellar , W.J. , Legg , J.P. , Hughes , D.P. , Blomme , G ., 2022 . Innovative digital technologies to monitor and control pest and disease threats in root, tuber, and banana (RT&B) cropping systems: Progress and prospects. Root , Tuber and Banana Food System Innovations: Value Creation for Inclusive Outcomes , pp. 261 – 288 . 44. ↵ Kroschel , J. , Okonyaet , J. , Juarez , H. , Forbes , G. , Kreuze , J. , Beed , F. , Blomme , G. , Legg , J. , 2014 . Management of critical pests and diseases in RTB crops under changing climates, through risk assessment, surveillance and modeling . RTB Workshop Reports; CGIAR Research Program on Roots, Tubers and Bananas (RTB): Lima, Peru, p. 20 . 45. ↵ Kruskal , J.B ., 1964 . Nonmetric multidimensional scaling: a numerical method . Psychometrika 29 , 115 – 129 . OpenUrl CrossRef Web of Science 46. ↵ Kuhn , M ., 2008 . Building predictive models in R using the caret package . Journal of Statistical Software 28 , 1 – 26 . OpenUrl CrossRef PubMed 47. ↵ Legg , J. , Bouwmeester , H. , Bigirimana , S. , Tata-Hangy , W. , Obiero , H. , Gashaka , G. , Mkamilo , G. , Ndyetabula , I. , Jeremiah , S. , Alicai , T. , 2009 . Cassava Disease Surveillance Surveys 2009 . Mapping Report, Great Lakes Cassava Initiative. p. 46 . 48. ↵ Legg , J.P. , Owor , B. , Sseruwagi , P. , Ndunguru , J ., 2006 . Cassava mosaic virus disease in East and Central Africa: epidemiology and management of a regional pandemic . Advances in Virus Research 67 , 355 – 418 . OpenUrl CrossRef PubMed 49. ↵ Loxdale , H.D ., 2018 . Aspects, including pitfalls, of temporal sampling of flying insects, with special reference to aphids . Insects 9 , 153 . OpenUrl PubMed 50. ↵ Makiola , A. , Holdaway , R.J. , Wood , J.R. , Orwin , K.H. , Glare , T.R. , Dickie , I.A ., 2021 . Environmental and plant community drivers of plant pathogen composition and richness . The New Phytologist 233 , 496 – 504 . OpenUrl PubMed 51. ↵ Malumphy , C. , Eyre , D. , Anderson , H ., 2017 . Tobacco, sweet potato or silver leaf whitefly: Bemisia tabaci . Department for Environment Food and Rural Affairs , United Kingdom. Hal , p. 6 . 52. ↵ Margosian , M.L. , Garrett , K.A. , Hutchinson , J.S. , With , K.A ., 2009 . Connectivity of the American agricultural landscape: assessing the national risk of crop pest and disease spread . BioScience 59 , 141 – 151 . OpenUrl CrossRef 53. ↵ McCulloch , G.A. , Waters , J.M ., 2023 . Rapid adaptation in a fast-changing world: Emerging insights from insect genomics . Global Change Biology 29 , 943 – 954 . OpenUrl CrossRef PubMed 54. ↵ McEwan , M. , 2016 . Sweetpotato seed systems in sub-saharan Africa: A literature review to contribute to the preparation of conceptual frameworks to guide practical interventions for root, tuber and banana seed systems . RTB Working Paper. CGIAR Research Program on Roots, Tubers and Bananas (RTB), Apartado 1558, Lima 12, Peru, p. 45 . 55. ↵ Mishra , K. , Subbanna , A. , Rajashekara , H. , Paschapur , A.U. , Jeevan , B. , Singh , A.K. , Maharana , C ., 2024 . Perceptions on Disease and Pest Status of Major Cultivated Crops in Indian Himalayas Under Changing Climate . Adapting to Climate Change in Agriculture-Theories and Practices: Approaches for Adapting to Climate Change in Agriculture in India , pp. 121 – 141 . 56. ↵ Mouafo-Tchinda , R.A. , Beaulieu , C. , Fall , M.L. , Carisse , O ., 2021 . Effect of temperature on aggressiveness of Plasmopara viticola f. sp. aestivalis and P. viticola f. sp. riparia from eastern Canada . Canadian Journal of Plant Pathology 43 , 73 – 87 . OpenUrl CrossRef 57. ↵ Mouafo-Tchinda , R.A. , Etherton , B.A. , Plex Sulá , A.I. , Andrade-Piedra , J. , Ogero , K. , Omondi , B.A. , McEwan , M.A. , Tene Tayo , P.M. , Harahagazwe , D. , Cherinet , M ., 2024 . Pathogen and pest risks to vegetatively propagated crops in humanitarian contexts: Towards a national plant health risk analysis for Cameroon and Ethiopia . bioRxiv , 2024.2002.2012.580019. 58. ↵ Mouafo-Tchinda , R.A. , Fall , M.L. , Beaulieu , C. , Carisse , O ., 2022 . Competition Between Plasmopara viticola Clade riparia and Clade aestivalis : A Race to Lead Grape Downy Mildew Epidemics . Plant Disease 106 , 2866 – 2875 . OpenUrl PubMed 59. ↵ Nakato , G.V. , Okonya , J.S. , Kantungeko , D. , Ocimati , W. , Mahuku , G. , Legg , J.P. , Blomme , G ., 2023 . Influence of altitude as a proxy for temperature on key Musa pests and diseases in watershed areas of Burundi and Rwanda . Heliyon 9 . 60. ↵ Ngatat , S. , Hanna , R. , Doumtsop Fotio , A.R. , Lienou , J.A. , Nanga Nanga , S. , Fotso Kuate , A. , Osundahunsi , B. , Fiaboe , K.K. , Ndemba , B. , Dossa , G.S ., 2024 . Distribution and diversity of emergent Banana bunchy top virus infecting banana and plantain in Cameroon , Central Africa. Journal of Phytopathology 172 , e13279 . OpenUrl CrossRef 61. ↵ Nyang’au , D. , Atandi , J. , Cortada , L. , Nchore , S. , Mwangi , M. , Coyne , D ., 2021 . Diversity of nematodes on banana ( Musa spp.) in Kenya linked to altitude and with a focus on the pathogenicity of Pratylenchus goodeyi . Nematology 24 , 137 – 147 . OpenUrl CrossRef 62. ↵ Ochieng , J. , Knerr , B. , Owuor , G. , Ouma , E ., 2015 . Agricultural commercialization and household food security: The case of smallholders in Great Lakes Region of Central Africa. International Association of Agricultural Economists, Milan , Italy 212588 , p. 30 . OpenUrl 63. ↵ Ochieng , J. , Ouma , E. , Birachi , E ., 2014 . Gender participation and decision making in crop management in Great Lakes Region of Central Africa . Gender, Technology and Development 18 , 341 – 362 . OpenUrl CrossRef 64. ↵ Ocimati , W. , Ogwal , G. , Tazuba , A.F. , Kubiriba , J. , Tugume , J. , Erima , R. , Okurut , W. , Mahuku , G. , Kutunga , D. , Blomme , G ., 2024 . Mapping the vulnerability of banana production landscapes in Uganda to banana bunchy top disease . Frontiers in Agronomy 6 , 1401478 . OpenUrl 65. ↵ Ogot , B.A. , Niane , D.T ., 1984 . The Great Lakes Region . General history of Africa IV: Africa from the twelfth to the sixteenth century , 498 – 524 . 66. ↵ Okonya , J.S. , Gamarra , H. , Nduwayezu , A. , Bararyenya , A. , Kroschel , J. , Kreuze , J ., 2021 . Serological survey and metagenomic discovery of potato viruses in Rwanda and Burundi reveals absence of PVY in Burundi and first report of TRV in potatoes in sub-Saharan Africa . Virus Research 302 , 198487 . OpenUrl CrossRef PubMed 67. ↵ Okonya , J.S. , Kroschel , J ., 2013 . Incidence, abundance and damage by the sweet potato butterfly ( Acraea acerata Hew.) and the African sweet potato weevils ( Cylas spp.) across an altitude gradient in Kabale district , Uganda. Int J AgriScience 3 , 814 – 824 . OpenUrl 68. ↵ Olivoto , T. , Lúcio , A.D.C ., 2020 . metan: An R package for multi-environment trial analysis . Methods in Ecology and Evolution 11 , 783 – 789 . OpenUrl 69. ↵ Ong’ayo , A.O ., 2018 . Displacement and cross-border mobility in the Great Lakes region-re-thinking underlying factors and implications for regional management of migration . Africa Insight 48 , 62 – 85 . OpenUrl 70. ↵ Pautasso , M. , Döring , T.F. , Garbelotto , M. , Pellis , L. , Jeger , M.J ., 2012 . Impacts of climate change on plant diseases—opinions and trends . European Journal of Plant Pathology 133 , 295 – 313 . OpenUrl 71. ↵ Pinnegar , J.K. , Hamon , K.G. , Kreiss , C.M. , Tabeau , A. , Rybicki , S. , Papathanasopoulou , E. , Engelhard , G.H. , Eddy , T.D. , Peck , M.A ., 2021 . Future socio-political scenarios for aquatic resources in Europe: A common framework based on shared-socioeconomic-pathways (SSPs) . Frontiers in Marine Science 7 , 568219 . OpenUrl 72. ↵ Poppinga , S. , Weisskopf , C. , Westermeier , A.S. , Masselter , T. , Speck , T ., 2016 . Fastest predators in the plant kingdom: functional morphology and biomechanics of suction traps found in the largest genus of carnivorous plants . AoB Plants 8 , plv140 . OpenUrl 73. ↵ Poveda , K. , Martínez , E. , Kersch-Becker , M.F. , Bonilla , M.A. , Tscharntke , T. , 2012 . Landscape simplification and altitude affect biodiversity, herbivory and Andean potato yield . Journal of Applied Ecology 49 , 513 – 522 . OpenUrl CrossRef 74. ↵ Pryke , A. , Mostaghim , S. , Nazemi , A ., 2007 . Heatmap visualization of population based multi objective algorithms . Evolutionary Multi-Criterion Optimization: 4th International Conference, EMO 2007, Matsushima, Japan, March 5-8, 2007. Proceedings 4. Springer, pp. 361 – 375 . 75. ↵ R Core Team , R., 2024 . R: A language and environment for statistical computing . R Foundation for Statistical Computing . 76. ↵ Ramcharan , A. , McCloskey , P. , Baranowski , K. , Mbilinyi , N. , Mrisho , L. , Ndalahwa , M. , Legg , J. , Hughes , D.P ., 2019 . A mobile-based deep learning model for cassava disease diagnosis . Frontiers in plant science 10 , 272 . OpenUrl PubMed 77. ↵ RTB , C., 2013 . Workshop on tools for managing seed degeneration in RTB crops (5-8 February 2013, Arusha, Tanzania) . CRP-RTB, p. 23 . 78. ↵ Rybicki , E.P ., 2015 . A Top Ten list for economically important plant viruses . Archives of virology 160 , 17 – 20 . OpenUrl CrossRef PubMed 79. ↵ Savary , S. , Willocquet , L. , Elazegui , F.A. , Teng , P.S. , Van Du , P. , Zhu , D. , Tang , Q. , Huang , S. , Lin , X. , Singh , H. , 2000 . Rice pest constraints in tropical Asia: characterization of injury profiles in relation to production situations . Plant Disease 84 , 341 – 356 . OpenUrl PubMed 80. ↵ Science Group Project on Accelerated Varietal Improvement and Seed Systems in Africa , 2024 . CGIAR project on Accelerated Varietal Improvement and Seed Systems in Africa: Annual Technical Report 2023 . 81. ↵ Selvaraj , M.G. , Vergara , A. , Ruiz , H. , Safari , N. , Elayabalan , S. , Ocimati , W. , Blomme , G ., 2019 . AI-powered banana diseases and pest detection . Plant Methods 15 , 1 – 11 . OpenUrl CrossRef PubMed 82. ↵ Simpson , G ., 2020 . ggvegan; ggplot2-based plots for vegan package . version 0.1. 0. In: https://github.com/gavinsimpson/ggvegan ; consulted : May . 83. ↵ Singh , B.K. , Delgado-Baquerizo , M. , Egidi , E. , Guirado , E. , Leach , J.E. , Liu , H. , Trivedi , P ., 2023 . Climate change impacts on plant pathogens, food security and paths forward . Nature Reviews Microbiology 21 , 640 – 656 . OpenUrl CrossRef PubMed 84. ↵ Sparks , A.H ., 2018 . nasapower: a NASA POWER global meteorology, surface solar energy and climatology data client for R . Journal of Open Source Software 3 , 1035 . OpenUrl 85. ↵ Sparks , A.H. , Forbes , G.A. , Hijmans , R.J. , Garrett , K.A ., 2014 . Climate change may have limited effect on global risk of potato late blight . Global Change Biology 20 , 3621 – 3631 . OpenUrl CrossRef PubMed 86. ↵ Sseruwagi , P. , Sserubombwe , W. , Legg , J.P. , Ndunguru , J. , Thresh , J ., 2004 . Methods of surveying the incidence and severity of cassava mosaic disease and whitefly vector populations on cassava in Africa: a review . Virus Research 100 , 129 – 142 . OpenUrl CrossRef PubMed 87. ↵ Taberlet , P. , Coissac , E. , Hajibabaei , M. , Rieseberg , L.H ., 2012 . Environmental DNA . Molecular Ecology 21 . 88. ↵ Thies , C. , Tscharntke , T ., 1999 . Landscape structure and biological control in agroecosystems . Science 285 , 893 – 895 . OpenUrl Abstract / FREE Full Text 89. ↵ Wells , K. , Clark , N.J ., 2019 . Host specificity in variable environments . Trends in Parasitology 35 , 452 – 465 . OpenUrl CrossRef PubMed 90. ↵ Were , H. , Kabira , J. , Kinyua , Z. , Olubayo , F. , Karinga , J. , Aura , J. , Lees , A. , Cowan , G. , Torrance , L ., 2013 . Occurrence and distribution of potato pests and diseases in Kenya . Potato Research 56 , 325 – 342 . OpenUrl CrossRef 91. ↵ Wickham , H ., 2016 . ggplot2: Elegant Graphics for Data Analysis . Springer-Verlag , New York . ISBN 978-3-319-24277-4. 92. ↵ Wieczorek , K. , Fulcher , T.K. , Chłond , D. , 2019 . The composition of the aphid fauna (Insecta, Hemiptera) of the Royal Botanic Gardens, Kew . Scientific Reports 9 , 10000 . OpenUrl PubMed 93. WorldData.info, 2024 . Climate comparison between Rwanda and Burundi based on a long-term analysis of the last 20 years . 94. ↵ Xiao , Q. , Li , W. , Chen , P. , Wang , B ., 2018 . Prediction of crop pests and diseases in cotton by long short term memory network . Intelligent Computing Theories and Application: 14th International Conference, ICIC 2018, Wuhan, China, August 15-18, 2018, Proceedings, Part II 14. Springer, pp. 11 – 16 . 95. ↵ Xing , Y. , Hernandez Nopsa , J.F. , Andersen , K.F. , Andrade-Piedra , J.L. , Beed , F.D. , Blomme , G. , Carvajal-Yepes , M. , Coyne , D.L. , Cuellar , W.J. , Forbes , G.A. , Kreuze , J.F. , Kroschel , J. , Kumar , P.L. , Legg , J.P. , Parker , M. , Schulte-Geldermann , E. , Sharma , K. , Garrett , K.A ., 2020 . Global cropland connectivity: A risk factor for invasion and saturation by emerging pathogens and pests . BioScience 70 , 744 – 758 . OpenUrl CrossRef PubMed 96. ↵ Yamamura , K. , Kiritani , K ., 1998 . A simple method to estimate the potential increase in the number of generations under global warming in temperate zones . Applied Entomology and Zoology 33 , 289 – 298 . OpenUrl CrossRef View the discussion thread. Back to top Previous Next Posted October 25, 2025. Download PDF Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Pathogen and pest communities in agroecosystems across climate gradients: Anticipating future challenges in the highland tropics Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Pathogen and pest communities in agroecosystems across climate gradients: Anticipating future challenges in the highland tropics Romaric A. Mouafo-Tchinda , Aaron I. Plex Sulá , Berea A. Etherton , Joshua S. Okonya , Gloria Valentine Nakato , Yanru Xing , Jacobo Robledo , Ashish Adhikari , Guy Blomme , Déo Kantungeko , Anastase Nduwayezu , Jan F. Kreuze , Jürgen Kroschel , James P. Legg , Karen A. Garrett bioRxiv 2025.01.08.631994; doi: https://doi.org/10.1101/2025.01.08.631994 Share This Article: Copy Citation Tools Pathogen and pest communities in agroecosystems across climate gradients: Anticipating future challenges in the highland tropics Romaric A. Mouafo-Tchinda , Aaron I. Plex Sulá , Berea A. Etherton , Joshua S. Okonya , Gloria Valentine Nakato , Yanru Xing , Jacobo Robledo , Ashish Adhikari , Guy Blomme , Déo Kantungeko , Anastase Nduwayezu , Jan F. Kreuze , Jürgen Kroschel , James P. Legg , Karen A. Garrett bioRxiv 2025.01.08.631994; doi: https://doi.org/10.1101/2025.01.08.631994 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Ecology Subject Areas All Articles Animal Behavior and Cognition (7635) Biochemistry (17691) Bioengineering (13892) Bioinformatics (41937) Biophysics (21452) Cancer Biology (18588) Cell Biology (25504) Clinical Trials (138) Developmental Biology (13378) Ecology (19899) Epidemiology (2067) Evolutionary Biology (24320) Genetics (15609) Genomics (22506) Immunology (17736) Microbiology (40394) Molecular Biology (17181) Neuroscience (88605) Paleontology (666) Pathology (2832) Pharmacology and Toxicology (4824) Physiology (7641) Plant Biology (15156) Scientific Communication and Education (2045) Synthetic Biology (4294) Systems Biology (9825) Zoology (2271)