A team of 29 subject matter experts (SMEs) conducted the assessment. The team was convened from regulatory agencies (L’Autorité Burundaise de Régulation des Produits Vétérinaires, des Pesticides et des Aliments (ABREVPA), Direction Générale de la Mobilisation pour l’Auto-développement et de la Vulgarisation Agricole (DGAEVMA), Direction de la Protection des Végétaux du Burundi (DPV), and Office National de Contrôle et de Certification des Semences (ONCCS); national (Institut des Sciences Agronomiques du Burundi (ISABU) and international agricultural research institutions (CABI); academia (Université du Burundi, Université Lumière de Bujumbura, and Université de Ngozi); and extension institutions (Direction de Vulgarisation, Formation, Recherche et Développement (DVFRD). The SMEs had experience in bacteriology, entomology, mycology, nematology, and virology.

The assessment was carried out across two workshops. The first workshop, held from 17 to 26 April 2023, focused on training the team in horizon scanning, utilising the CABI Horizon Scanning Tool, and conducting rapid risk assessments in accordance with the guidelines outlined in Supplementary File S1 . The premium version of the Horizon Scanning Tool was used to create a preliminary list of plant pests that had not yet been reported as present in Burundi. The parameters set in the Horizon Scanning Tool involved determining the area at risk (Burundi) and selecting regions from which non-native pest species were likely to be introduced (source areas). The source areas included all geographical regions in Africa, Asia, Europe, North America, Oceania, and South America. All parameters except for the type of organism were left open. The considered pest organisms consisted of arthropods, bacteria (including phytoplasma), chromista, fungi, molluscs, protists, plant parasitic nematodes, and viruses (including viroids). Following the scan, the list was downloaded in Excel (.xlsx) format for further analysis. The current list of regulated pests for Burundi was also examined to determine if there were pests not included in the horizon scan output that could be considered for rapid risk assessment. The assessment conducted in the first workshop was validated in the second, held from 25 to 29 November 2024.

The risk-scoring guidelines employed were based on the framework outlined by Roy et al. (2019) but have been modified by Kenis et al. (2022) and Mulema et al. (2022). Roy et al. (2019) evaluated the likelihood of introduction (entry), establishment, spread, and potential negative impacts on biodiversity and ecosystem services; however, this study examined the possibility of introduction (entry), establishment (including aspects of spread), and potential socio-economic and environmental (biodiversity) impacts. A five-point scoring system outlined in Supplementary File S1 was utilised for the four parameters (entry, establishment, socio-economic impact, and biodiversity impact).

A score of one indicated that the organism was absent from Africa and unlikely to be found in an imported commodity; two indicated absence from Africa (the pest is not known to occur in Africa) but likely to be infrequently imported on a commodity; three was ascribed to three scenarios: either present in Africa (excluding neighbouring countries to Burundi) and spreading slowly; or absent from Africa but demonstrated to spread very quickly across several continents, or often associated with a commodity that is commonly imported, or frequently intercepted in Burundi; four, indicated presence in Africa (excluding neighbouring countries) with rapid spread, or presence in a neighbouring country with slow spread; and five, indicated presence in a neighbouring country (the Democratic Republic of the Congo (DR Congo), Rwanda, and Tanzania) with rapid spreads. A pest could have been present in Africa but absent from Burundi, suggesting that it was reported elsewhere on the African continent but not yet known to occur in Burundi. The global pest distribution data were obtained from the CABI Compendium, the European and Mediterranean Plant Protection Organisation (EPPO) Global Database, and the Global Biodiversity Information Facility (GBIF).

Hulme et al. (2008) detailed the three mechanism and their respective pathways through which an alien species may enter new geographical or political areas. These mechanisms include the importation of a commodity (via three pathways: contaminant, escape, and release), the arrival of a transport vector (through one pathway: stowaway), and natural spread from a neighbouring region (via two pathways: corridor and unaided). The contaminant pathway involves organisms unintentionally introduced with a specific commodity, including obligate pathogens, parasites, and commensals of traded plants and animals. The escape pathway comprises organisms that are intentionally introduced but unintentionally escape, including feral crops and livestock, pets, garden plants, and live baits. The release pathway primarily involves organisms introduced deliberately for release, including biocontrol agents, game animals, and landscape plants. Only the contaminant pathway was considered because comprehensive risk assessments are typically conducted for organisms deliberately introduced into countries or regions.

The stowaway pathway, also considered in this assessment, involves organisms that are unintentionally introduced because they are attached to or within a transport vector, such as molluscs. The corridor pathway involves organisms that are unintentionally introduced as a result of human infrastructure development, linking previously unconnected regions. This pathway was not considered because it does not apply to the majority of pests. Lastly, the unaided pathway comprises organisms that are introduced through natural dispersal across political borders. Examples include flying insects and pollen, spores, and soil infested with pathogenic microorganisms that wind currents may carry. Therefore, the contaminant (CO), stowaway (ST), and unaided (UN) were considered for this study because they are more relevant to the introduction and spread of pests.

A score of one indicated that Burundi was climatically unsuitable or that host plants were absent; two, that only a few areas in Burundi were climatically suitable, or host plants were rare; three, that large regions of Burundi were climatically suitable, whilst host plants were rare; or that only a few areas in Burundi were climatically suitable, but host plants were at least moderately abundant; four, that large regions of Burundi were climatically suitable, but host plants were moderately abundant; and five, that large areas in Burundi were climatically suitable, whilst host plants were very abundant. To ensure consistency in assessing the capacity for the establishment, the SMEs weighted the likely hosts to be grown in Burundi ( Supplementary File S2 ).

A score of one indicated that the alien species did not attack plants that were cultivated or utilised; a score of two signified that the alien species damaged plants that were only occasionally cultivated or utilised; a score of three denoted that the alien species damaged plants that were regularly cultivated or utilised, but without threatening the cultivation, utilisation, or trade of this crop; a score of four meant that the alien species had the potential to threaten, at least locally, the cultivation of a plant that was regularly cultivated or utilised, or to periodically attack a crop that was key to the economy of Burundi without posing a threat to the latter; and a score of five indicated that the alien species had the potential to threaten, at least locally, a crop that was essential to the economy of Burundi.

A score of one indicated that the alien species would not affect any native species; two meant the alien species would impact individuals of a native species without affecting its population level; three signified that the alien species had the potential to reduce the population levels of a native species; four implied that the alien species could eradicate a native species locally or affect populations of a protected or keystone species; and five denoted that the alien species possessed the potential to completely eliminate a native species or to locally eradicate a keystone species. Native species in this context include animals, plants, and other organisms, including pathogens.

All assessments were conducted in plenary by all SMEs. The scores for each species attribute were thoroughly discussed, and any discrepancies were resolved until a consensus was reached before the final score was confirmed. As detailed in Blackburn et al. (2014), a low, medium, and high confidence rating was provided for each of the four assessed attributes, the likely pathway of arrival, and the overall risk score. The vectors and vectored pathogenic organisms of prioritised pests were also evaluated.

Consequently, the overall risk score was calculated using the following formula, resulting in scores ranging from 250 to two.

An overall score of 54 was deemed the threshold for recommending a management action. This score was chosen because it is only achievable for species with a score of three for all assessable attributes or those exceeding three in at least three parameters. A score of three or more suggests a higher likelihood of entry, establishment, and significant impact (socio-economic or biodiversity). Scores below three are regarded as low risk as they indicate a reduced likelihood of entry, establishment, and impact; a score of three is classified as moderate, while scores above three (four and five) signify an increased likelihood of entry, establishment, and impact (contrasting with the lower scores of one and two). A no-action approach was recommended for species that recorded an overall risk score below 54. In contrast, a management action was proposed for all pests that recorded an overall risk score above 54. The suggested actions included detection surveillance for pests reported in a neighbouring country to establish pest status in Burundi, regulation supported by Pest Risk Analysis (PRA), contingency planning for preparedness, raising awareness, management by industry (relevant value chain system), and research.

Detection surveillance was also considered if the assessment team confirmed a previous interception or if there was a pest frequently affecting a host commodity imported from a country where the pest is endemic. Pest Risk Analysis is essential in providing guidelines for managing pathways to limit the introduction of quarantine pests. However, action was deemed necessary for certain species that recorded an overall risk score below 54. These are species currently regarded by trading partners as high-risk A1 or A2 quarantine pests, such as subspecies of Xylella fastidiosa. While a no-action recommendation would be made for some species that recorded an overall risk score above 54, the final decision on regulation or deregulation was left to the National Plant Protection Organisation (NPPO). This was for species whose primary pathway of introduction was deemed implausible. This category included pests transmitted by vectors rather than through seed (Seed-contaminated or seed-transmitted), yet the vectors were unlikely to be introduced alongside a viable pathogenic organism. Others in this category included soil-borne pests that are unlikely to accompany the commodity. A prime example would be soil-borne fungi that are not seed-transmitted. Furthermore, a no-action recommendation was also suggested when insufficient information was available for the assessment of a particular species or vectors of assessed species, as well as for pests vectored by assessed species already reported in Burundi.

A three-stage approach was utilised to develop the final list of non-native plant pest species that could impact crop and forest productivity if introduced to Burundi. Firstly, a global horizon scan identified 8,747 species that have not yet been reported as existing in Burundi. These species included 5,647 arthropods, 311 bacteria (including 54 phytoplasma), 171 chromista (oomycetes), 1,719 fungi, 27 molluscs, 206 nematodes, nine protists, 12 viroids, and 645 viruses. The initial output from the Horizon Scanning Tool was subsequently filtered to include only species with complete names, resulting in 8,541 species, which comprised 5,517 arthropods, 296 bacteria (including 54 phytoplasmas), 165 chromista, 1,678 fungi, 27 molluscs, 192 nematodes, nine protists, 28 viroids, and 629 viruses ( Supplementary File S3 ). Of the 8,541 species, 641 had been reported as invasive. These 641 species consisted of 397 arthropods, 41 bacteria (including 14 phytoplasma), 15 chromista (oomycetes), 94 fungi, 11 molluscs, 38 nematodes, two viroids, and 43 viruses. This data is detailed in Supplementary File S2 (Sheet: All Horizon Scanning Data). The list of 8,541 species was further refined by selecting those that impacted major value chains cultivated in Burundi, resulting in 1,801 species.

The current list of regulated pests for Burundi was developed in 2006, and since then, the status of some pests has changed. The list comprises 123 species, 93 of which were part of the horizon scanning output, except for 30 (27 of which had already been reported in Burundi). Of the 123 species, 110 are specific, while the remaining 13 are represented as general species, including Anastrepha spp., Ceratitis spp., Chilo spp., Cuscuta spp., Cylas spp., Dacus spp., Ditylenchus spp., Erwinia spp., Heterodora spp., Lyriomyza spp., Meloidogyne spp., Sclerospora spp., and Striga spp. Two of the remaining 110 species, Corynebacterium pisi and Elsinoe piperis, could not be determined, while for one species, Bostrynchopsis villosula, the available information was insufficient for assessment, leaving only 107 species. The 107 species included 23 not yet reported in Africa and 84 that have already been reported in Africa.

The 84 species documented in Africa included 24 species already recorded in Burundi. The remaining 60 species were reported in a neighbouring country. A total of 78 species were part of the horizon scanning output. The 78 species included 19 not reported in Africa, 59 reported in Africa, of which 27 were reported in a neighbouring country. Additionally, a number of species with the genera Anastrepha, Ceratitis, Chilo, Cylas, Dacus, Ditylenchus, Erwinia, Heterodora, Lyriomyza, Meloidogyne, and Sclerospora were also part of the output. However, eight species, including Allantophomopsiella pseudotsugae, Coffee blister spot agent, Helicobasidium brebissonii, Nilapavarta lugens, Popilia japonica, Robbsia andropogonis, Ustilaginoidella oedipigera, and Wheat striate mosaic virus were not and have been considered for rapid risk assessment. All data have been consolidated in Supplementary File S3 .

A total of 1,801 species from the horizon scanning output were assessed to identify those capable of transmitting or being transmitted by other species. The species that could act as vectors or be vectored but were not included in the horizon scanning output were also evaluated to determine their associated risks. The species capable of vectoring the assessed species numbered 105, while those that could be vectored numbered eight. The 105 species included 99 arthropods, one fungus (Olpidium bornovanus), four plants, all Cuscuta species (C. campestris, C. ceanothi, C. europaea, and C. reflexa), and one nematode (Paratrichodorus anemones). The arthropods belonged to the orders Acarida (one), Coleoptera (five), Hemiptera (88), Hymenoptera (four), and Thysanoptera (one). Most of the Hemiptera species were from the families Cicadellidae (54) and Aphididae (19). The eight species that could be vectored included two fungi (Microbotryum succisae and Neocosmospora euwallaceae), one protist (Gregarina sitophili), and five viruses (millet red leaf persistent luteovirus, raspberry leaf spot virus, watermelon mosaic potyvirus-II, wheat striate mosaic virus, and wheat yellow mosaic virus).

The Soft Rot Pectobacteriaceae, formerly known as the Soft Rot Enterobacteriaceae, includes two genera; Dickeya and Pectobacterium, both of which are part of the bacterial family Pectobacteriaceae. These two species cause highly destructive diseases in many value chains, particularly in the Solanaceae and Poaceae families. The horizon scanning analysis identified eight species of Dickeya and nine species of Pectobacterium. Six of the Dickeya species (D. chrysanthemi, D. dadantii, D. dianthicola, D. fangzhongdai, D. paradisiaca, and D. solani) affect potatoes, while one of each affects maize (D. zeae) and rice (D. oryzae). Although D. zeae was initially considered a single species, studies by Hugouvieux-Cotte-Pattat and Van Gijsegem (2021) demonstrated diversity among D. zeae strains, leading to two distinct clades, one of which was later elevated to D. oryzae. Seven of the nine identified species of Pectobacterium are known to affect potatoes. They include P. aroidearum, P. atrosepticum, P. betavasculorum, P. brasiliense, P. carotovorum, P. parmentieri and P. polaris. The previous species P. carotovorum subsp. carotovorum has been elevated to a number of species, including P. carotovorum, P. parvum, and P. peruviense (Waleron et al., 2018; Portier et al., 2019; Pasanen et al., 2020). Pectobacterium punjabense, recently isolated from potatoes, is a new species closely related to P. parmentieri (Sarfraz et al., 2018). Therefore, the inclusion of the species D. oryzae, P. parvum, P. peruviense and P. punjabense was guided by their close relationship to the species (D. zeae, P. carotovorum, and P. parmentieri) identified through horizon scanning.

The total number of species suggested for rapid risk assessment was 1,924. This total included 1,803 species generated through horizon scanning. Eight species from the current list of regulated pests for Burundi, which were not part of the horizon scanning output, were added, bringing the total to 1,811. Additionally, four species from the genera Dickeya (D. oryzae) and Pectobacterium (P. parvum, P. peruviense, and P. punjabense) were included, as they were not part of the original horizon scanning output but were deemed essential due to their phytosanitary consequences, raising the number to 1,815. Furthermore, 104 vectors that transmit the assessed species and five pathogenic organisms transmitted by the assessed species were added, completing the total number of prioritised species for assessment. The 1,924 species included 416, which accounted for approximately 22%. More than half (n=992, 52%) of these species had already been reported in Africa, with about one-third of the 967 (n=356, 36%) documented in a neighbouring country. The 1,924 species comprised 550 arthropods, 153 bacteria, 79 chromista, 518 fungi, 19 molluscs, 141 nematodes, nine protists, 16 viroids, and 435 viruses. Of the 416 invasive species, there are 250 arthropods, 25 bacteria (including six phytoplasmas), seven chromista, 61 fungi, nine molluscs, 34 nematodes, one viroid, and 29 viruses. All data have been consolidated in Supplementary File S2 (Sheet: Assessment of prioritised pests).

The 550 arthropods prioritised for rapid risk assessment included 250 species (45%) recorded as invasive. Of these, 237 (43%) had been reported in Africa, and of that proportion, 96 (41%, n=237) were found in a neighbouring country ( Table 1 ). Among the 250 species listed as invasive, 101 were reported outside Africa, while 150 were reported within Africa, including 70 in a neighbouring country. The highest overall risk score was 175, attributed to 23 species, whereas the lowest was two, noted for 28 species. Six of the species, namely Dieuches humilis, Nysius euphorbiae, Phyllocoptes arcani, Unkanodes albifascia, Unkanodes sapporonus, and Uroleucon escalantii, could not be assessed due to lack of enough sufficient information. Most arthropods were likely to be introduced solely as contaminants (84%, n=461) ( Table 2 ). The remaining 18% could be introduced through various pathways, with 5% (n=30) entering as contaminants or stowaways and 10% (n=53) as contaminants, stowaways, or unaided ( Table 2 ). The remaining six species not assessed accounted for 1%. The 450 assessed species also included 313 (57%) known vectors. All data have been consolidated in Supplementary File S4 .

The 153 bacteria species prioritised for assessment included 25 (16%) already documented as invasive. Of these, 61% (n=93) were reported in Africa, with 33% (n=31) reported in a neighbouring country ( Table 1 ). The highest overall risk score was 140, recorded for 18 species, while the lowest was two, recorded for four species: Candidatus Phytoplasma rubi, Pseudomonas syringae pv. atropurpurea, Xanthomonas axonopodis pv. betlicola, and Xanthomonas translucens pv. arrhenatheri. One species (sesame phyllody phytoplasma) was not assessed due to a lack of sufficient information. The majority of assessed bacteria (78%, n=120) were likely introduced solely as contaminants, while the fewest (7%, n=11) could be introduced only as stowaways. The remaining 14% (n=21) could enter through either pathway as contaminants or stowaways ( Table 2 ). The pathway of introduction could not be determined for one species (Sesame phyllody phytoplasma), as indicated above. Thirty-two (21%) of the assessed species were vectored, of which all were phytoplasmas, except for Candidatus Liberibacter americanus, Candidatus Liberibacter asiaticus, Candidatus Liberibacter solanacearum, Dickeya zeae, Pantoea ananatis, Pantoea stewartii subsp. stewartii, Rathayibacter rathayi, Rathayibacter tritici, Spiroplasma citri, Spiroplasma kunkelii, Xylella fastidiosa subsp. fastidiosa (Xff), Xylella fastidiosa subsp. multiplex (Xfm), and Xylella fastidiosa subsp. pauca (Xfp). All data have been consolidated in Supplementary File S5 .

The 79 assessed Chromista (oomycetes) included seven species noted as invasive. Most (67%, n=53) of the species had been reported in Africa, with 18 of these (34%, n=53) documented in a neighbouring country ( Table 1 ). The highest overall risk score was 175, recorded for Peronosclerospora maydis and Phytophthora nicotianae, while the lowest was six, recorded for Peronospora radii, Phytophthora aleatoria, Phytophthora pinifolia, and Pythium sulcatum. Half (49%, n=39) of the species were likely to be introduced solely as stowaways, while 41% (n=32) were likely to be introduced as contaminants. The remaining 10% (n=8) could potentially be introduced through multiple pathways, specifically as contaminants or stowaways ( Table 2 ). All data have been consolidated in Supplementary File S6 .

The 518 fungi prioritised for assessment included 61 (12%) documented as invasive. A proportion of 61% (n=318) had been reported in Africa, of which 43% (n=138) were recorded in a neighbouring country ( Table 1 ). The majority of assessed fungi (75%, n=390) were likely introduced exclusively as contaminants, while the fewest (1%, n=7) could be introduced solely as stowaways. The remaining 24% (n=121) could be introduced through multiple pathways ( Table 2 ). This group includes 28 (5%) that could be introduced as both contaminants and stowaways and 93 (18%) that could be both contaminants and unaided. The highest overall risk score was 175, recorded for Lasiodiplodia theobromae, Phakopsora pachyrhizi, Puccinia polysora, Uromyces appendiculatus, and Ustilago maydis, while the lowest was two, recorded for seven species including Erysiphe biocellata, Exobasidium vaccinii, Myrothecium verrucaria, Passalora sequoiae, Phomopsis vaccinii, Sirococcus clavigignenti-juglandacearum, and Tilletia lolii. Nine of the assessed fungal species, which include Albonectria rigidiuscula, Amylostereum areolatum, Fusarium oxysporum f.sp. cubense Tropical Race 4 (Foc TR4), Haematonectria haematococca, Leptographium procerum, Microbotryum succisae, Microbotryum violaceum, Neocosmospora euwallaceae, and Raffaelea lauricola, were likely to be disseminated by vectors. All data have been consolidated in Supplementary File S7 .

The 19 molluscs prioritised for assessment included nine species documented as invasive, representing 47% ( Table 1 ). Nine (47%) of the assessed molluscs had been reported in Africa, but only one had been documented in a neighbouring country (Tanzania) ( Table 1 ). Species recorded in Africa included Achatina fulica, Arion hortensis Ferussac, Bradybaena similaris, Cornu aspersum, Deroceras laeve, Deroceras reticulatum, Limax maximus, Pomacea canaliculata, and Rumina decollata with P. canaliculata also reported in Kenya. The highest overall risk score was 120, recorded for Achatina fulica, while the lowest was six, recorded for Meghimatium pictum. All the assessed molluscs could be introduced exclusively as stowaways ( Table 2 ). All data have been consolidated in Supplementary File S8 .

The 141 prioritised plant-parasitic nematodes included 34 (24%) species identified as invasive. Of these, 57% (n=81) were reported in Africa, with 27% (n=21) documented in a neighbouring country ( Table 1 ). The highest overall risk score was 140, recorded for 13 species, while the lowest was 2, noted for two species: Longidorus macrosoma and Trophotylenchulus piperis. A proportion of 31% (n=44) was likely to be introduced exclusively as contaminants, while 35% (n=49) was likely introduced as stowaways. The remaining 34% (n=48) could be introduced either as contaminants or stowaways. Four nematodes, Bursaphelenchus fungivorus, Bursaphelenchus mucronatus, Bursaphelenchus xylophilus, and Rhadinaphelenchus cocophilus, could be vectored by arthropods, while 17 were known vectors for bacteria and viruses. All data have been consolidated in Supplementary File S9 .

The four evaluated Cuscuta species (C. campestris, C. ceanothi, C. europaea, and C. reflexa) were not part of the horizon scanning output but were included for assessment due to their role as vectors for the species being evaluated. With the exception of C. ceanothi, the remaining three Cuscuta species were reported in Africa, and only C. campestris was found in a neighbouring country, Tanzania. All were likely to be introduced as contaminants, except for C. campestris, which could be introduced as either a contaminant or stowaway. The highest overall risk score recorded was 100 recorded for C. campestris, while the lowest was 35, attributed to Cuscuta ceanothi. All data have been consolidated in Supplementary File S10 .

The nine protists prioritised for assessment included three species, Plasmodiophora brassicae, Polymyxa betae, and Polymyxa graminis, all reported in Africa, with none observed in a neighbouring country ( Table 1 ). The highest overall risk score was 105, recorded for Polymyxa graminis, while the lowest was 12, recorded for two species, Physarum cinereum and Phytomonas ananas. Eight of the assessed protists could be introduced exclusively as stowaways, whereas only Ligniera vasculorum could be introduced as either a contaminant or stowaway ( Table 2 ). Two protists, Polymyxa betae and Polymyxa graminis, are known vectors of viruses. Polymyxa betae is known to vector beet necrotic yellow vein virus, while Polymyxa graminis vectors soilborne wheat mosaic virus, wheat spindle streak mosaic virus, wheat streak mosaic virus, and wheat yellow mosaic virus. Additionally, three protists, Gregarina sitophili, Phytomonas leptovasorum and Phytomonas staheli can spread by arthropod vectors. All data have been consolidated in Supplementary File S11 .

The 16 viroids prioritised for assessment included only one viroid, the potato spindle tuber viroid, identified as invasive. Twelve viroids, accounting for 75% of the total, were reported in Africa, with none recorded in a neighbouring country. The majority (75%, n=12) of the viroids were likely introduced as contaminants, while the remaining 19% (n=3) were introduced either as contaminants or stowaways. For one viroid, the pathway could not be determined based on the available information. The highest overall risk score was 140, recorded for potato spindle tuber viroid and tomato apical stunt viroid, while the lowest was three, recorded for one species (coconut cadang-cadang viroid). All data have been consolidated in Supplementary File S12 .

The 435 virus species prioritised for assessment included 29 identified as invasive. Forty-two percent of the total assessed viruses (n=181) were reported in Africa, along with 49 reported in a neighbouring country ( Table 1 ). Most of the viruses were likely introduced exclusively as stowaways (49%, n=214), while the least were contaminants (16%, n=68). The remaining 34% (n=147) could have been introduced as either contaminants, stowaways, or both. Due to insufficient information, the potential pathways for six viruses- including cowpea mosaic virus, millet red leaf persistent luteovirus, okra yellow vein mosaic virus, pepper chlorotic spot virus, pigeon pea yellow mosaic virus, and wheat yellow mosaic virus- could not be determined. The highest overall risk score recorded was 175, noted for 11 viruses, while the lowest was two, noted for 14 viruses. A proportion of 84% (n=366) of the viruses are transmitted by known vectors, while 21 are expected to be vectorised; however, the specific vectors are yet to be identified. All data have been consolidated in Supplementary File S13 .

The total number of known vector species assessed was 354. These included 328 arthropods, three fungi, 12 nematodes, four plants, and three protists. Of the 350 species, 26 were not assessed because they had already been reported as present in Burundi, while an additional six, including Dieuches humilis, Nysius euphorbiae, Phyllocoptes arcani, Unkanodes albifascia, Unkanodes sapporonus, and Uroleucon escalantii, were not assessed due to insufficient information to conduct the assessment leaving 317. The 317 species comprised 296 arthropods (33%), three fungi, 12 nematodes, four plants, and two protists. The fungi included Olpidium bornovanus, Olpidium brassicae, and Olpidium radicale; the plants included Cuscuta. campestris, Cuscuta ceanothi, Cuscuta europaea, and Cuscuta reflexa; and the protists included Polymyxa betae and Polymyxa graminis. The nematodes belonged to the genera Longidorus, Paratrichodorus, Trichodorus, and Xiphinema. Most of the arthropods were Hemiptera, specifically from the family Cicadellidae, followed by the Aphididae. The highest overall risk score recorded was 175 for eight species, which included Apis mellifera, Bemisia afer, Bemisia tabaci, Chilo patellus, Cicadulina mbila, Frankliniella schultzei, Peregrinus maidis, and Thrips tabaci. The lowest overall risk score recorded was two for 28 species. All data have been consolidated in Supplementary File S14 (Sheet: Vector species assessment).

The total number of vectored species was 215, comprising 28 bacteria (including phytoplasma), nine fungi, five nematodes, one protist, and 172 viruses. Two species, Ralstonia solanacearum and tobacco rattle virus, were not assessed as they were already present in Burundi. Additionally, three species, millet red leaf persistent luteovirus, sesame phyllody phytoplasma, and wheat yellow mosaic virus- were excluded from the assessment due to insufficient information. One species, Angiostrongylus cantonensis, was also not assessed because it affects humans rather than crops. The remaining 210 species included 109 (52%) reported as present in Africa, of which 24 were noted in a neighbouring country. The majority of the vectors for the assessed species were arthropods, accounting for 90% (n=195). However, other vectors included nematodes from the genera Anguina (A. agrostis and A. tritici), Longidorus (L. attenuatus and L. elongatus), Paratrichodorus (P. anemones, P. minor, P. pachydermus, P. porosus, and P. teres), Trichodorus (T. primitivus, T. similis, and T. viruliferus), and Xiphinema (X. americanum, X. diversicaudatum, and X. rivesi); fungi from the genus Euwallacea (E. fornicatus and E. perbrevis); molluscs from the genus Pomacea (P. canaliculata); and protists in the genus Polymyxa (P. betae and P. graminis). The highest overall risk score was 175, reported for six species (alfalfa mosaic virus, bean yellow mosaic virus, maize mosaic virus, potato virus A, potato virus M, and potato virus Y), while the lowest was two, recorded for eight species, all viruses, including black raspberry necrosis virus, blueberry leaf mottle virus, blueberry shoestring virus, chickpea chlorotic dwarf Syria virus, parsnip mosaic virus, raspberry leaf mottle virus, raspberry leaf spot virus, and ryegrass mosaic virus. All data have been consolidated in Supplementary File S14 (Sheet: Vectored species assessment).