Severe root symptoms of Cassava Frogskin Disease (CFSD) remerged in Colombia during 2020–2023 significantly impacting cassava yield in Orinoquia, one of the main cassava producing regions. Using a finite population sampling strategy (ca. 136 fields with ~30 observations per field = 4620 observations), field surveys in the main cassava-producing provinces showed high prevalence values of severe root symptoms. The upper limits of prevalence values were observed in Fuente de Oro (66.67%), Granada (63.33%), Vista Hermosa (56.67%), Puerto Lleras (46.67%), and El Castillo (23.33%). Molecular diagnostics of a subsample (n=149) identified torradovirus species 2 (CsTLV-2) and the oryzavirus CsFSaV as the most frequent pathogens (27.63% and 57.89%, respectively) in this region. The overall prevalence of CsTLV was 27.63%. Notably, among the torradoviruses, CsTLV-2 (100%) showed a clear dominance over CsTLV-1 (9.52%). In contrast, samples collected in the Department of Cauca—where the disease has been endemic and reported since the early 1970s—exhibited a distinct torradovirus profile. The total prevalence of CsTLV was 46.58%, with CsTLV-1 and CsTLV-2 detected at frequencies of 52.94% and 85.29%, respectively. In contrast, the incidence of CsFSaV was approximately threefold lower than that observed in Orinoquia. It is noteworthy that the diversity of torradoviruses was higher in Cauca and that in Orinoquia CsTLV-2 showed characteristics consistent with a recent population expansion. Altogether, the data presented here indicate that the 2020–2023 outbreak of CFSD was characterized by a high prevalence of severe roots symptoms and widespread infections by CsTLV-2 and the oryzavirus CsFSaV, in comparison with other pathogens detected in the analyzed samples.
cassava frogskin diseaseCsFSaVCsTLV-1CsTLV-2genetic diversityMinisterio de Ciencia, TecnologÍa e Innovación10.13039/100022965110390385717 CT 80740 444 2021The author(s) declared that financial support was received for this work and/or its publication. This publication is derived from the project “Sustainable and competitive cassava pro-duction system with a focus on territorial development”, within the framework of the program “Strengthening the cassava value network in Colombia through co-innovation in primary pro-duction, transformation, and access to markets with sustainability, competitiveness and circularity criteria.” code 110390385717 CT 80740 444–2021 financed by MINCIENCIAS. Financial support from CGIAR’s Sustainable Farming Program is acknowledged.section-at-acceptanceDisease ManagementIntroduction
Cassava (Manihot esculenta Crantz) is cultivated in many regions across the world and has become a crucial crop for food security, particularly in tropical and subtropical areas (Mohidin et al., 2023). In 2022, the total cassava planted area was distributed as follows: 63.1% in Africa, 29% in Asia, and 7.8% in the Americas (FAOSTAT, 2022). In Colombia, cassava cultivation occurs under diverse environmental conditions, encompassing 32 departments from 34 of this country and extending over an area exceeding 200 thousand hectares (Rosero Alpala et al., 2023). However, several biotic constraints limit cassava production across the tropics (Legg et al., 2015; Siriwan et al., 2020; Pardo et al., 2022; 2023). In Colombia, Cassava frogskin disease (CFSD), an endemic disease of cassava was first described in the 1970s from severely affected fields identified in the department of Cauca (Pineda et al., 1983). The disease has now been reported in Costa Rica, Venezuela, Brazil and Paraguay (Cardozo Téllez et al., 2016; Oliveira et al., 2014; de Souza et al., 2014; Alvarez et al., 2014). CFSD primarily affects the roots, restricting starch accumulation. A major challenge in diagnosing CFSD is the frequent absence of visible symptoms in the aerial parts of most cassava varieties, making disease severity apparent only upon root harvest (Pardo et al., 2022). Since its first report, it is estimated that CFSD has caused more than 90% of crop losses in the country’s main cassava-producing areas (Pineda et al., 1983), including the departments of Meta, Arauca, and Cauca during several outbreaks.
Since its first reports, samples with CFSD presented evidence for mixed pathogen infections, all of which recently identified at sequence level thanks to the use of molecular techniques and next-generation sequencing strategies (Pardo et al., 2022). Currently up to four viral agents and one phytoplasma of ribosomal group III have been identified in CFSD-affected plants. The first virus identified in CFSD plants was named as cassava frogskin associated virus (CsFSaV), and up to this date this oryzavirus, appears to be the most prevalent in Colombia (Calvert et al., 2008; Jimenez et al., 2024). Three other viruses were identified in plants with symptoms of the disease in Colombia and Brazil: cassava new alphaflexivirus (CsNAV), cassava polero-like virus (CsPLV) and cassava torrado-like virus (CsTLV) (Carvajal-Yepes et al., 2014; De Oliveira et al., 2020). In addition, a phytoplasma belonging to ribosomal group 16Sr III was identified in CFSD plants in Colombia, Paraguay, Costa Rica and Brazil (Alvarez et al., 2009; de Souza et al., 2014; Cardozo Téllez et al., 2016).
Pathogen accumulation in vegetative propagated crops is not unusual, as in the absence of proper seed quality certification protocols, infections build up over successive crop cycles (Thomas-Sharma et al., 2017). This characteristic confounds diagnostics and shows the limits of validating Koch’s postulates to identify a causal agent (Fox, 2020). Interestingly, a recent approach has found out that among all pathogens detected in mixed infections in CFSD, single infections by torradoviruses could explain the disease (Jimenez et al., 2024). This causal association between torradoviruses and CFSD was determined using sentinel plants, where those that presented characteristic roots symptoms (longitudinal lip-like fissures in the root skin, a reduction in the size and thickness of the root, opaque and brittle root skin, no symptoms in the above-ground parts of the plant) (Pardo et al., 2022). where single infected by torradoviruses (Jimenez et al., 2024). The availability of a molecular diagnostic protocol to analyze field collected planting material confirmed that the severe root symptoms observed in CFSD-affected plants was associated with this pathogen (Jimenez et al., 2024).
From 1930 to 2023, outbreaks of cassava diseases have been reported worldwide. According to Legg and Thresh (2000), cassava mosaic disease (CMD), caused by cassava mosaic geminiviruses (Family Geminiviridae; Genus Begomovirus), led to epidemics in Madagascar and Uganda during the 1930s and 1940s, and a rapid but more localized spread was observed along the coast of Tanzania in the 1930s and the coast of Kenya in the 1970s. The disease was first reported in East Africa in 1894. As noted by Legg (1999), during the 1990s a severe epidemic of African cassava mosaic virus (CMD) occurred in central and eastern Africa, eventually covering all of Uganda and parts of neighboring Kenya, Tanzania, Sudan, and the Democratic Republic of the Congo, causing significant production losses.
Further work by Legg et al. (2011) compared the regional epidemiology of the cassava mosaic virus and cassava brown streak virus pandemics in Africa. Several regional studies indicate a continuous pattern of annual spread toward the west and south along a sustained front. In contrast, outbreaks of cassava brown streak disease (CBSD) have been reported in Uganda and other regions of East Africa that had previously not been affected by the disease. Recent data reveal major contrasts in the regional epidemiology of these two pandemics: (i) CMD radiates from an initial center of origin, whereas CBSD appears to spread from independent foci; and (ii) the severe CMD pandemic has emerged from recombination and synergistic interactions among virus species, whereas the CBSD pandemic seems to represent a novel host–pathogen encounter.
The study by Casinga et al. (2021), which examined the expansion of the cassava brown streak disease epidemic in eastern Democratic Republic of the Congo between 2016 and 2018, assessed foliar incidence and severity of CBSD, reporting disease incidence levels of 13.2% in 2016 and 16.1% in 2018. More recently, Godding et al. (2023) developed a predictive model for an emerging cassava epidemic in sub-Saharan Africa—an epidemic, stochastic, landscape-scale model that captures the spread of the disease in Uganda and can be readily extended to generate predictions for the 32 major cassava-producing countries in sub-Saharan Africa.
Our results show that the prevalence of CFSD, defined as the proportion of plants affected at a given time, reached severe levels of more than 60% in the Orinoquia region and exceeded 30% in the Andean region. At the molecular level, CsTLV-2 and CsFSaV were the most prevalent viruses in the Orinoquia region during the outbreak. Although severe disease incidence was lower in Andean region, a greater diversity of torradoviruses—and a higher occurrence of other viruses—was detected. Together with the causal association between CsTLV-2 and severe root symptoms of CFSD recently reported by Jimenez et al. (2024), we hypothesize that the torradovirus CsTLV-2 has played a key role on the development of the 2020–2023 outbreak disease in Colombia.
Materials and methodsDigital data collection
To assess the degree of impact, a survey was designed to quantify and identify the areas with the highest presence of CFSD root symptoms in cassava fields. The survey included at least six simple questions intended to help visually recognize the disease and concluded with questions related to its presence on the farm and the areas affected (Figure 1). Data were collected in four departments of Colombia on a single date per location during 2023, which were selected according to the percentage of production of cassava per hectare, according to municipal agricultural evaluations (EVAS for Spanish acronym, https://www.agronet.gov.co/estadistica/Paginas/home.aspx?cod=1 2022) and reports obtained by virtual survey. Meta and Arauca departments represent the Eastern Plains region of Colombia, and contain major cassava cultivated areas in this region, here, nine municipalities were explored (3 from Arauca and six from Meta), summarizing 4220root digital samples from plants aged between 8–10 months after planting (Table 1). From the inter-Andean valley region, five municipalities from Cauca department were visited. Cauca is the main producer of cassava raw material for fermented starch in Colombia. At each farm visited, a visual assessment was conducted on the roots (one root per plant) of thirty randomly selected cassava plants across one transect diagonal within one hectare field. In total we scouted 141 fields. Each plant was then rated for disease severity on a scale from 1 to 5, following the classification system developed by CIAT (Pardo et al., 2022). After scoring, ten (10) plants were randomly selected for tissue sampling. Digital root RGB images exhibiting disease symptoms rated at levels 4 or 5 were collected and GPS coordinates were recorded.
Results of the survey conducted to identify regions and cassava production areas affected by CFSD. (A) Number of reports per department. (B) Area reported as at risk by producers based on 41 survey responses.
Two horizontal bar charts comparing Colombian departments. Panel A shows the number of CFSD presence and absence reports, with Meta having the highest reports of CFSD presence. Panel B displays total area affected by CFSD, with Meta and Cordoba having the largest affected areas. CFSD not observed is indicated in white, CFSD presence in gray.
Number of plants inspected in this work (we observed one storage root per plant). A subset of 149 samples were processed for molecular diagnostics. Dx = # of samples used for standard molecular diagnostics. The lower and upper limits of prevalence were calculated according to the “finite population sampling” protocol, based on the collection of 30 samples following an X-shaped transect path.
Department
Municipality
Cultivated area (ha)*
Temperature (°C)
Precipitation (mm)
Altitude (masl)
# roots inspected/# leaf tissue sampling
CFSD Prevalence (range)
Orinoquia (Arauca)
Fortul
1200
25.96
1.768,9
176
60 (16)
10 – 30
Saravena
1620
25.9
1.922,6
251
60 (13)
0 – 10
Tame
3300
26.14
1.769,5
225
60 (12)
6,6–23.3
Orinoquia (Meta)
El Castillo
1248
25.9
2.781
379
540
0 – 23.3
Fuente de Oro
1850
26.62
2.568,5
291
1470 (6)
0 – 66.6
Granada
1493
26.6
2.533,1
347
1140 (15)
0 – 63.3
Puerto Lleras
710
27
2.632,4
238
810 (8)
0 – 46.6
Vista Hermosa
1120
26.1
2.548,1
292
120 (6)
10 – 56.6
Buenos Aires
710
19.9
2.232,4
1093
60 (34)
16.6 -36.6
Andean(Cauca)
Caldono
360
14.52
2.198,8
1415
90 (9)
0 – 3.3
El Tambo
2300
23.9
2.406,5
955
60 (10)
0
Guachené
201
23.8
1.800
1043
60 (8)
0 – 6.6
Timbío
1060
15.7
2.080
1805
90 (12)
0 – 3.3
*Reported area during year 2023.
Leaf tissue collection and storage
Young leaves were collected from 149 plants, representing 8 municipalities in Orinoquia and 5 in inter-Andean valley region (Supplementary Table 1). Between three and five grams of the youngest upper leaves or apical buds were collected. The plant material was wrapped in a paper towel for protection, then placed in a small Ziploc® bag containing fresh silica gel and labeled collection bags to ensure proper preservation for further analysis. Samples were stored at room temperature for up to two weeks or at 4 °C for long-term preservation, with silica gel replaced every 2–3 months to maintain tissue desiccation. Plant tissue collection was carried out independently of root symptoms digital recording, therefore no correlation could be determined between severe root symptoms and molecular diagnostic results.
Statistical analysis
The analysis of virus and phytoplasma prevalence data was conducted using generalized linear models (GLMs). Given the binary nature of the molecular diagnostic response variable (Positive/Negative), a binomial distribution with a logit link function was employed to estimate the prevalence of each pathogen by department. To assess significant differences in pathogen prevalence among departments, multiple comparisons were made. Statistical significance was adjusted using the Holm correction method, with a significance level of α = 0.05 (RStudio 2025.09.01).
Nucleic acid extraction and molecular diagnostics
Total RNA and DNA extraction, reverse transcription, and PCR specific to each viral and phytoplasma species were performed following standardized protocols established by the International Center for Tropical Agriculture (Jimenez et al., 2021; Pardo et al., 2022). Nucleic acid extraction was conducted using the CTAB (cetyltrimethylammonium bromide) method, as described by Jimenez et al. (2021). For RNA samples, complementary DNA (cDNA) synthesis was carried out using random hexamer primers, enabling the simultaneous detection of multiple viruses through a single RT-PCR reaction. An internal control primer targeting the endogenous nad5 gene (NADH dehydrogenase subunit 5) was used as a quality control for cDNA synthesis (Menzel et al., 2002) to confirm the presence of amplifiable cDNA and reduce the likelihood of false-negative results. After confirming RNA integrity and cDNA synthesis using nad5, virus-specific detections were conducted. The thermal cycling conditions and primer sequences used for standard diagnostics are detailed in Supplementary Tables 2 and 3, respectively.
RT-PCR amplification and primer design for phylogenetic analysis
Two pairs of primers were designed to amplify genomic regions of CsTLV species using the Primer Design tool in Geneious Prime® v.2023.2.1 (Biomatters Ltd., New Zealand). Primer design was based on a multiple sequence alignment of CsTLV isolates retrieved from GenBank. For CsTLV-1, primers targeted a 1030 bp fragment of the RNA-dependent RNA polymerase (RdRp) region of RNA1: CsTLV-Yop12 RdRp_Fw (5′-GAAAAGTTCCTGTTCCTGG-3′) and CsTLV-Yop12 RdRp_Rv (5′-CCCCGACTCAAGAACACCAA-3′). For CsTLV-2, primers were designed to amplify a 1651 bp fragment of the conserved protease-polymerase (Pro-Pol) region of RNA1: CsTLV-2909-Fw (5′-GCAGCCTTCCAGTTACCTGT-3′) and CsTLV-4559-Rv (5′-GCTCGACTCAGGAAAACCAG-3′). The PCR reaction mix was prepared with 12.5 μL GoTaq® Green Master Mix (Promega Corp), 0.5 μL of each primer (10 pmol) and 9.5 μL ultrapure water (Invitrogen, Life Technologies, CA, USA) to complete a final volume of 25 μL. PCR amplifications were performed in a Mastercycler Nexus®. For the RdRp region of CsTLV-1, the thermal cycling conditions consisted of an initial denaturation at 94 °C for 5 min, followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 40 s, and extension at 72 °C for 1 min, with a final extension at 72 °C for 10 min. Amplification of the Pro–Pol region of CsTLV-2 was carried out under the following conditions: initial denaturation at 94 °C for 5 min, followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at 57 °C for 40 s, and extension at 72 °C for 1 min 40 s, with a final extension at 72 °C for 10 min.
Nanopore sequencing
Libraries were prepared using the Ligation Sequencing Kit SQK-NBD114.96 according to the manufacturer’s online protocols (Oxford Nanopore Technologies, ONT). Amplicons ranging from 1,030 to 1,651 bp were quantified using the Qubit™ 1X dsDNA High Sensitivity Assay on a Qubit® 4.0 Fluorometer (Invitrogen, Life Technologies), and 400 ng of DNA per sample was used for library preparation. DNA repair and end-preparation were performed by mixing the amplicons with 0.875 μL NEBNext FFPE DNA Repair Buffer, 0.5 μL NEBNext FFPE DNA Repair Mix, 0.875 μL Ultra II End-Prep Reaction Buffer, and 0.75 μL Ultra II End-Prep Enzyme Mix (New England Biolabs, MA, USA). The reactions were purified using AMPure XP beads (Beckman Coulter, CA, USA) and eluted in 10 μL of nuclease-free water (NFW). Native barcode was ligated by mixing 7.5 μL of End-prepped DNA prepared in the previous step, 2.5 μL of native barcode (ONT) and 7 μL of Blunt/TA ligase master mix (New England Biolabs). The reaction was stopped by adding 4 μL of EDTA. Barcoded samples were pooled with 0.4X AMPure XP Beads. For the clean-up steps, 700 μL of short fragment buffer was added, and the pellet was eluted in 35 μL of NFW. Adapters were ligated by mixing 30 μL of pooled barcoded sample, 5 μL native adapter (ONT), 10 μL of 5X NEBNext Quick Ligation Reaction Buffer and 5 μL Quick T4 DNA Ligase (New England Biolabs, MA, USA). Amplicons were cleaned up again using AMPure XP Beads, and a successive clean-up step was performed by adding 125 μL of short fragment buffer (ONT). The pellet was resuspended in 15 μL of elution buffer (ONT). Library concentration was measured using a Qubit dsDNA HS Assay Kit and a Qubit 2.0 fluorometer (Invitrogen, USA). A total volume of 75 μL was used for library preparation, consisting of 37.5 μL of sequencing buffer, 25.5 μL of library beads, and 12 μL of DNA library. The library was loaded onto R10.4.1 flow cell (FLO-MIN114) and sequenced for 24 h on a MK1D device (ONT).
Sequence processing and phylogenetic analysis
Raw reads were basecalled with Dorado duplex (v. 0.5.3) using the hac model. The resulting reads were then separated into simplex, which were demultiplexed using Dorado (kit SQK-NBD114.96). The consensus sequence of each PCR product was assembled with the program Amplicon_sorter (Vierstraete and Braeckman, 2022) using strict parameters. Phylogenetic trees were generated in MEGA X (Kumar et al., 2018) using nucleotide (nt) sequences of the RdRp region (RNA1) from torradovirus isolates available in GenBank, along with the sequences obtained in this study. The tree was inferred using the Neighbor-Joining method with 1,000 bootstrap replications, and evolutionary distances were calculated using the Kimura 2-parameter model. A genetic and population diversity analysis was conducted for the two CsTLV species, comparing populations from the Orinoquia and Cauca regions. All analyses were performed in DnaSP v5 using the default parameters (Librado and Rozas, 2009). The dataset was evaluated for the number of haplotypes, haplotype diversity (Hd), nucleotide diversity (π), the number of polymorphic sites (k), and neutrality statistics including Tajima’s D and Fu and Li’s D and F (Fu and Li, 1993). Haplotype diversity was defined as the probability that two randomly chosen haplotypes from the sample are different, whereas nucleotide diversity represents the average number of nucleotide differences per site between pairs of sequences (Beerenwinkel et al., 2012).
Results
Despite the producers’ difficulties in recording information through virtual platforms, 41 responses were obtained for the survey. More than 90% of these responses came from departments located in the Orinoquia region (Meta, Arauca, and Casanare (Figure 1A). Among them, 83% reported having observed symptoms in their cassava crops. The extent of cassava production areas at risk was also a matter of concern for the institutions, since more than 300 has were reported to be in risk (Figure 1B). This status was the preliminary evidence of severe CFSD in this region.
Status of severe CFSD in Orinoquia during the 2020–2023 outbreak
Field surveys in the Orinoquia region took into account the area and yield of cultivated cassava (Table 1; Figure 2). In this case, the departments of Meta and Arauca represent 85.1% of cultivated cassava in this region and the 8.6% in all Colombia (Agronet, 2024). As CFSD roots symptoms analysis were limited to severe symptoms of degrees 4 and 5, as described in Pardo et al. (2022), therefore, the prevalence percentages indicated in this work actually represent an underestimate of the actual CFSD prevalence during the 2020–2023 outbreak Figure 3A illustrates the distribution of 4220 fields across three CFSD prevalence intervals (0-5%, 6-15%, and >15%) in the Orinoquia and Andean (Cauca) regions, representing the prevalence of this disease in the fields. In Cauca, 75% of the fields fell within the 0-5% prevalence interval, whereas in the Orinoquia region, 37.3% and 34.5% of the fields were within the 6-15% and >15% intervals, respectively, indicating a higher overall prevalence of the disease in Orinoquia (Figure 3). Higher estimated values suggest that severe CFSD reached above 80% severe disease prevalence during the 2020–2023 outbreak, confirming field reports from farmers and other stakeholders from Meta and Arauca. Analysis of predicted prevalence across the surveyed provinces revealed notable spatial variation in disease occurrence. Fuente de Oro had the highest predicted prevalence (66.67%), whereas Saravena showed the lowest (10%). In the Andean region, Buenos Aires exhibited the highest prevalence (36.67%), while El Tambo recorded no detectable prevalence (0%) (Table 1). These results highlight the heterogeneous distribution of the disease across regions, suggesting that management practices such as cultivar selection, use of clean planting material, positive selection for planting material and others, may influence prevalence levels. (Table 1;Figure 3B). CFSD prevalence of severe root symptoms in Cauca show that the disease is still impactful in this region. (Table 1).
Map showing the sampling regions and the prevalence of root symptoms of CFSD across three departments in Colombia. Balloon size represents CFSD prevalence, and darker shading in municipalities within the surveyed departments indicates cultivated areas.
Map of Colombia displaying municipal boundaries, major regional divisions shaded in brown and yellow, and red circles indicating specific locations of interest. A north arrow appears in the upper right corner, and a scale bar is at the bottom left.
Disease prevalence in cassava fields from Cauca and Orinoquia regions of Colombia. (A) CFSD field-prevalence intervals observed in Cauca (blue) and Orinoquia (orange) regions. Numbers above the bars indicate the percentage of observations. (B) Box plots showing lower and upper prevalence limits with significant differences (p < 0.05) according to the Wilcoxon non-parametric test. Darker colors indicate higher data density.
Panel A presents a clustered bar chart comparing field prevalence percentages across two regions, Cauca (blue) and Orinoquia (orange), with Cauca showing higher fields at low prevalence and Orinoquia showing higher fields in the moderate and high prevalence categories. Panel B displays a dot plot of prevalence percentages for individual fields in each region, with mean values highlighted by red lines; Orinoquia has more fields and a higher average prevalence than Cauca.Occurrence of pathogens in the field
A total of 149 samples were used to investigate the occurrence of pathogen infections in Orinoquia during the outbreak, with Cauca included for comparison. The percentage of prevalence results from proportion of samples infected by each pathogen. Interestingly, when the data was disaggregated by specific pathogen and region the results showed different distributions in each region. In the Andean region (Cauca), the highest prevalence was for torradoviruses (46.58%), where CsTLV-1 reach 52.94% and CsTLV-2 85.29% while the oryzavirus CsFSaV reached only 20.55% and the phytoplasma close to 0% (Figure 4). However, in the Orinoquia region, torradoviruses reached (27.63%), the oryzavirus CsFSaV 57.89% and phytoplasma less than 2% of the total number of pathogens detected. It is noteworthy that among torradoviruses, CsTLV-1 percentage was 9.52% and CsTLV-2 was detected in 100% of the CsTLV infected samples. In conclusion, the reported 2020–2023 CFSD outbreak was not only characterized by the high prevalence of severe root symptoms but by a high percentage of CsFSaV and CsTLV-2 infections in Orinoquia as compared to Cauca. Other pathogens reported in cassava (CsPLV, CsVX and CsNAV) in Orinoquia showed low percentages with no significative differences among them detected by Tukey´s test (p<0.05). On the contrary in the Andean region (Cauca) CsPLV and CsVX showed higher percentages, reaching 39.73% and 41.1%, respectively.
Occurrence of pathogens in fields during the 2020–2023 outbreak in Orinoquia. Pathogens were identified from 149 cassava leaf tissue samples collected in (A) Orinoquia (n = 76) and (B) Andean Cauca (n = 73) regions of Colombia. The pathogens tested: cassava frogskin associated virus (CsFSaV), cassava polerolike virus (CsPLV), cassava virus X (CsVX), cassava new alphaflexivirus (CsNAV), 16SrIII-L phytoplasma (16SrIII-L), and cassava-torrado-like virus (species 1 and 2). Error bars indicate the standard error. Different letters above the bars indicate significant differences obtained by Tukey’s multiple comparison test (p-value < 0.05).
Bar chart compares prevalence percentages of six virus types in Orinoquia (n equals 76) and Cauca (n equals 73) regions, with error bars. CsFsaV dominates in Orinoquia, while CsPLV and CsVX are highest in Cauca. Stacked bars depict CsTLV-1 and CsTLV-2 prevalence in red and blue, respectively. Letters above bars indicate statistical significance groups.Torradovirus diversity in Colombia
Unlike CsFSaV and phytoplasma, torradoviruses have been shown associated with severe root symptoms of CFSD, in single infections (Jimenez et al., 2024) and as shown above only one species CsTLV-2 appears to have expanded in the Orinoquia region during the 2020-2023 outbreak in this region. To explore the diversity of this virus in more detail we designed new primers to amplify and sequence the RdRp region of 18 isolates collected in Orinoquia and 19 isolates collected in Cauca (Supplementary Table 2). Phylogenetic analysis revealed the formation of distinct clades corresponding to the two torradovirus species previously reported (Jimenez et al., 2024). It is noteworthy that CsTLV-2 isolates from Cauca (Figure 5) formed a monophyletic cluster, separate from those CsTLV-2 isolates reported in in Meta and Arauca. On the other hand, CsTLV-1 isolates were almost exclusively detected in samples from Cauca and were distributed at least in three well-supported clusters. Bootstrap values supported the robustness of the main branches, confirming the presence of both CsTLV species and several divergent strains in Colombia. To assess genetic and population diversity, analyses were conducted for the CsTLV virus in both species, CsTLV-1 and CsTLV-2. Furthermore, the genetic diversity of CsTLV occurrence was examined across different regions in a specific time period to identify potential geographic patterns of genetic variation (Table 2). These findings demonstrate clear difference between CsTLV populations in Cauca and Orinoquia that were differentiated by new designed primers that allowed to detect several lineages in RdRp region, however, further studies should be done to dissect spatio-temporal factors related to pathogen diversity and their functional or epidemiological implications.
Phylogenetic relationship of CsTLV isolates. Phylogenetic tree using the neighbor-joining method based on nucleotide sequences of the RdRp region. Sequences in red represent isolates from Cauca, green from Arauca, and blue from Meta. Numbers of branches indicate percentage of bootstrap support out of 1000 bootstrap replication.
Phylogenetic tree diagram displaying relationships among various labeled sequences or isolates, with CsTLV-2 and CsTLV-1 clades marked on the right. Isolate labels are colored green, blue, red, or black, representing different groups or origins. Bootstrap values at branching points indicate statistical support for evolutionary relationships. A scale bar indicating genetic distance is shown at the bottom left.
Genetic diversity indices and neutrality tests were calculated for both CsTLV species using DnaSP v5.
Species
N
Nr. of sites
S
Hd
Pi
K
Tajima’s D
Fu and Li’s D
Fu and Li’s F
CsTLV-1
15
869
186
1
0.087
75.36
0.942 (P > 0.10)
0.689 (P > 0.10)
0.877 (P > 0.10)
CsTLV-2
41
854
311
0.99
0.077
63.25
− 1.309 (P > 0.10)
0.126 (P > 0.10)
-0.473 (P > 0.10)
CsTLV-2 -Orinoquia
19
854
84
1
0.01823
15.5
-1.583 (P > 0.10)
-2.142 (P > 0.10)
-2.301 (P > 0.10)
CsTLV-2 -Cauca
12
869
47
0.97
0.01249
9.97
-1.646 (P > 0.10)
-1.733 (P > 0.05)
-1.950 (P > 0.05)
n: number of samples; S: number of polymorphic sites; k: average number of nucleotide differences; π (Pi): nucleotide diversity; Hd, haplotype diversity. NS, not statistically significant (P > 0.10).
The alignment lengths were similar between the two groups (869–854 bp). CsTLV-1 had 186 polymorphic sites (S), whereas CsTLV-2 exhibited a greater number of segregating sites (S = 311). Genetic diversity differed markedly between the two species. CsTLV-1 showed the highest nucleotide divergence (π = 0.087; k = 75.36) despite having fewer polymorphic sites, and its haplotype diversity was maximal (Hd = 1.000). Although neutrality test values were positive, Fu and Li’s D* and F* were not statistically significant (P > 0.10), suggesting that the observed variation is consistent with neutral evolution and may reflect population structure rather than recent selective or demographic events.
In contrast, CsTLV-2 displayed more variable sites but lower nucleotide diversity (π = 0.077; k = 63.25). Tajima’s D and Fu and Li statistics were negative but not statistically significant (P > 0.10), indicating an excess of low-frequency variants that may be compatible with recent population expansion or purifying selection, although these trends are not strongly supported statistically. The regional analysis revealed moderate nucleotide diversity in both areas, with CsTLV-2 from Orinoquia showing higher genetic diversity than the population from Cauca. Nevertheless, both groups exhibited very high haplotype diversity and negative neutrality values. Taken together, these patterns suggest recent population expansion in both regions, with Orinoquia potentially harboring more diverse viral lineages.
Overall, CsTLV-1 appears more genetically divergent and structured, whereas CsTLV-2 shows patterns consistent with recent expansion and regional variation in genetic diversity.
Discussion
During the period 2020–2023 the Orinoquia region reported a larger than usual prevalence of CFSD. This outbreak impacted negatively in fresh root production and availability of clean planting material, consequently, reductions in yield of up to 19.93% (2021-2022) and 31.01% (2021-2024) for the municipality of Vista Hermosa were reported right after the disease spread in this region (Agronet, 2024). We show that in Orinoquia, root symptoms of CFSD reached higher than 60% prevalence and were detected in all surveyed fields. Similar behavior was observed with Cassava Brown Streak Disease (CBSD), showing an incidence above 16% (Casinga et al., 2021). At molecular diagnostic level, samples from this region were characterized by high prevalences of CsFSaV (57.89%) and CsTLV (27.63%) as compared to other pathogens previously reported in plants with CFSD (Pardo et al., 2022). Higher prevalence of CsTLV infections and CFSD symptoms were observed in the Andean region, department of Cauca (46.6%), a region where the disease was first reported (Pineda et al., 1983). However, in this study, the field evaluation was limited only to severe grades to avoid wrong subjective qualification in plants with low severity, therefore, any correlation was established between high or low severity with pathogens presence. From CsTLV infection in this region, CsTLV-1 reached 52.94% and CsTLV-2 with 85.29% while the oryzavirus CsFSaV reached only 20.55%. These findings are in agreement with those reported by Jimenez et al. (2024), where CsFSaV and phytoplasma were found not necessary for the development of CFSD severe root symptoms.
In Orinoquia, CsFSaV and CsTLV-2 were widespread and CsNAV, CsVX and CsPLV showed the lowest prevalence in the field. While in Cauca CsTLV, CsVX and CsPLV where more prevalent than CsFSaV. While 16SrIII-L phytoplasma (16SrIII-L) was not detected in Cauca and Arauca. This may reflect the dynamic nature of mixed infections, by pathogens that may not share the same vectors nor the same conditions of accumulation when colonizing each region (Pardo et al., 2022; Jimenez et al., 2024). Further studies are needed to improve understanding of the CFSD complex across different regions of Colombia. Indeed, it has been also reported that each region hosts different cassava genotypes adapted to specific agroclimatic conditions, and according to market and consumer preferences in each region (Contreras-Valencia, 2024; Floro et al., 2018). It is noteworthy that each of these regions produces a different set of cassava genotypes. Previous studies have recognized that in the Orinoquia, the predominant variety is Brasilera; while in Cauca, the Algodona landrace dominates (Contreras-Valencia, 2024; Floro et al., 2018). The occurrence of severe CFSD in both regions each with different agroclimatic conditions, and in a wide range of distinct cassava genotypes, highlights CFSD as a major biotic constrain to cassava production across Colombia.
Phylogenetic analysis of CsTLV based on the RdRp region revealed two well-supported clades corresponding to the recognized species within the genus Torradovirus: CsTLV-1 and CsTLV-2. This finding is consistent with previous reports describing two distinct CsTLV lineages in Colombia and other regions (Leiva et al., 2022; Jimenez et al., 2024). Phylogenetic relationships further showed that CsTLV-1 and CsTLV-2 isolates were widely distributed across the Cauca region, suggesting ongoing regional diversification.
Evolutionary and population genetic analyses revealed high haplotype diversity in both species. Similar patterns have been reported for several plant RNA viruses (Nguyen et al., 2013; Ge et al., 2014; Gao et al., 2017), and cases of high haplotype diversity combined with low nucleotide diversity are often interpreted as signatures of recent population expansion (Pagán et al., 2016; Gilbert, 2002; Rodríguez-Nevado et al., 2017). CsTLV-1 displayed slightly higher overall genetic diversity and positive neutrality test values, including Tajima’s D, consistent with population stability, balancing selection, or the long-term maintenance of divergent lineages (Tajima, 1989). However, these neutrality tests were not statistically significant, suggesting that the observed patterns may also reflect population structure rather than recent selective or demographic events.
In contrast, CsTLV-2 exhibited negative values across neutrality tests (Tajima’s D and Fu and Li’s D and F), indicating an excess of rare variants. Although these values were not statistically significant, the overall pattern is compatible with recent population expansion or purifying selection (Kustin and Stern, 2021). The CsTLV-2 sublineages from Orinoquía and Cauca showed a combination of low nucleotide diversity, high haplotype diversity, and negative neutrality statistics, a pattern commonly observed in rapidly expanding viral populations and frequently associated with outbreak dynamics (Meena et al., 2019; Abadkhah et al., 2021). Notably, the Orinoquía population exhibited higher genetic diversity than the Cauca population, suggesting the presence of more diverse viral lineages in this region.
Interestingly, isolate CM5460-10, collected in Valle del Cauca in 1998, contained representatives of both torradovirus species, indicating that CsTLV-1 and CsTLV-2 have likely co-existed in this region for an extended period. Together, these findings—supported by high bootstrap values—reinforce the existence of two divergent CsTLV species and suggest that the recent CFSD outbreak was characterized by the widespread occurrence of CsTLV-2 isolates that are genetically distinct from those currently found in Cauca (Figure 5).
Conclusions
The recent outbreak of CFSD in the Orinoquia region was characterized by high prevalence of severe root symptoms and the co-infection of the oryzavirus CsFSaV and the torradovirus CsTLV-2, in most locations surveyed. The widespread and genetically homogeneous distribution of CsTLV-2 in the Orinoquia, may reflect a more recent expansion of this virus, related with the recent outbreak in this region. Comparative sequence analysis of torradoviruses shows that both recorded species are present in Colombia, where CsTLV-1 is more diverse and not as widespread as CsTLV-2. Overall, these findings establish an important baseline for future comparative studies, particularly in light of the recurrent and reemergent nature of this disease in the Americas.
Data availability statement
The original contributions presented in the study are publicly available. This data can be found here: NCBI GenBank, accession PX562882-PX562922.
Author contributions
JM: Writing – original draft, Writing – review & editing, Visualization, Formal analysis, Data curation, Methodology, Investigation. AV-B: Formal analysis, Data curation, Writing – original draft, Methodology, Investigation. MS-G: Data curation, Methodology, Writing – original draft, Investigation. LL-R: Writing – original draft, Data curation, Methodology, Investigation. AR: Methodology, Supervision, Conceptualization, Investigation, Writing – review & editing, Writing – original draft, Funding acquisition, Project administration. AML: Investigation, Writing – original draft, Software, Visualization, Formal analysis, Data curation, Methodology. JJ: Conceptualization, Data curation, Methodology, Investigation, Writing – original draft, Visualization, Formal analysis. WJC: Conceptualization, Investigation, Validation, Resources, Writing – review & editing, Supervision, Writing – original draft, Methodology.
Acknowledgments
The authors would like to express their gratitude to Juan Manuel Pardo and Rafael Rodriguez from CIAT for their technical support. We thank the Ministry of Agriculture and Rural Development (MADR) for the financial support to Agrosavia and Ministry of Science and Technology (MINCIENCIAS) for financial support of this study.
Conflict of interest
The authors declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The author WJC declared that they were an editorial board member of Frontiers, at the time of submission. This had no impact on the peer review process and the final decision.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fagro.2026.1751653/full#supplementary-material
General information of samples randomly selected for molecular diagnostic analysis.
The following list contains the list of primers using for diagnostics, along with their respective nucleotide sequences, that were utilized in the present study.
The following is a detailed description of the PCR thermal profile for the detection of the viral species under study, the internal control of cDNA synthesis, and phytoplasma by qPCR.
Geographic distribution and sampling details of cassava plants collected across three departments in Colombia. The red dots correspond to the exact sampling locations.
ReferencesAbadkhahM.HamidiH.KoolivandD. (2021).
Phylogenetic and population analysis of lettuce mosaic virus isolates based on the coat protein gene. J. Genet. Resour.7, 211–219. doi: 10.22080/jgr.2021.20849.1238AlvarezE.MejíaJ. F.LlanoG. A.LokeJ. B.CalariA.DudukB.. (2009).
Characterization of a phytoplasma associated with frogskin disease in cassava. Plant Dis.93, pp.1139–1145. doi: 10.1094/PDIS-93-11-1139, PMID: 30754574AlvarezE.PardoJ. M.DufourD.MorenoJ. L.AlvarezE. (2014).
The metabolism of carbohydrates in roots of cassava (Manihot esculenta Crantz) infected with frogskin disease. Phytopathology104, 7. doi: 10.1094/PHYTO-104-11-S3.1, PMID: 25337637BeerenwinkelN.GünthardH. F.RothV.MetznerK. J. (2012).
Challenges and opportunities in estimating viral genetic diversity from next-generation sequencing data. Front. Microbiol.3, 329. doi: 10.3389/fmicb.2012.00329, PMID: 22973268CalvertL. A.CuervoM.LozanoI.VillarealN.ArroyaveJ. (2008).
Identification of three strains of a virus associated with cassava plants affected by frogskin disease. J. Phytopathol.156, pp.647–pp.653. doi: 10.1111/j.1439-0434.2008.01412.x, PMID: 41711423Cardozo TéllezL.PardoJ. M.ZacherM.TorresA.ÁlvarezE. (2016).
First report of a 16SrIII phytoplasma associated with frogskin disease in cassava (Manihot esculenta) in Paraguay. Plant Dis.100, pp.1492–1492.doi: 10.1094/PDIS-09-15-1102-PDN, PMID: 40211709Carvajal-YepesM.OlayaC.LozanoI.CuervoM.CastanoM.CuellarW. J. (2014).
Unraveling complex viral infections in cassava (Manihot esculenta Crantz) from Colombia. Virus Res.186, 76–86. doi: 10.1016/j.virusres.2013.12.011, PMID: 24374265CasingaC. M.ShirimaR. R.MahunguN. M.Tata-HangyW.BashiziK. B.MunyerenkanaC. M.. (2021).
Expansion of the cassava brown streak disease epidemic in eastern Democratic Republic of Congo. Plant Dis.105, pp.2177–2188. doi: 10.1094/PDIS-05-20-1135-RE, PMID: 33258425Contreras-ValenciaK. (2024). “
Comprensión del segmento de la yuca de consumo fresco en Colombia,” in Manual de manejo de yuca para consumo fresco en diferentes regiones de Colombia. Ed.
Rosero AlpalaEn E.A. (Mosquera-Colombia:
La Corporación Colombiana de Investigación Agropecuaria (Colombia: AGROSAVIA), 12–19.
de OliveiraS. A. S.FerreiraC. F.DiamantinoM. S. A. S.SantosT. A.PereiraJ. D. S.de OliveiraE. J. (2020).
First report of cassava torrado-like virus, cassava polero-like virus and cassava new alphaflexivirus associated with cassava frogskin disease in Brazil. J. Plant Pathol.102, pp.247–pp.247. doi: 10.1007/s42161-019-00384-6, PMID: 41732346de SouzaA. N.da SilvaF. N.BedendoI. P.CarvalhoC. M. (2014).
A phytoplasma belonging to a 16SrIII-A subgroup and dsRNA virus associated with cassava frogskin disease in Brazil. Plant Dis.98, 771–779. doi: 10.1094/PDIS-04-13-0440-RE, PMID: 30708630FAOSTAT. doi: 10.1111/1477-9552.12247, PMID: 41711423FloroV. O.IVLabartaR. A.Becerra López-LavalleL. A.MartinezJ. M.OvalleT. M. (2018).
Household determinants of the adoption of improved cassava varieties using dna fingerprinting to identify varieties in farmer fields: a case study in Colombia. J. Agric. Econ.69, 518–536. doi: 10.1111/1477-9552.12247, PMID: 41711423FoxA. (2020).
Reconsidering causal association in plant virology. Plant Pathol.69, 956–961. doi: 10.1111/ppa.13199, PMID: 41711423FuY. X.LiW. H. (1993).
Statistical tests of neutrality of mutations. Genetics133, 693–709. doi: 10.1093/genetics/133.3.693, PMID: 8454210GaoR.XuY.CandresseT.HeZ.LiS.MaY.. (2017).
Further insight into genetic variation and haplotype diversity of Cherry virus A from China. PLoS One12, e0186273. doi: 10.1371/journal.pone.0186273, PMID: 29020049GeB.HeZ.ZhangZ.WangH.LiS. (2014).
Genetic variation in potato virus M isolates infecting pepino (Solanum muricatum) in China. Arch. Virol.159, 3197–3210. doi: 10.1007/s00705-014-2180-6, PMID: 25233939GilbertG. S. (2002).
Evolutionary ecology of plant diseases in natural ecosystems. Annu. Rev. Phytopathol.40, 13–43. doi: 10.1146/annurev.phyto.40.021202.110417, PMID: 12147753GoddingD.StuttR. O.AlicaiT.AbidraboP.Okao-OkujaG.GilliganC. A. (2023).
Developing a predictive model for an emerging epidemic on cassava in sub-Saharan Africa. Sci. Rep.13, p.12603. doi: 10.1038/s41598-023-38819-x, PMID: 37537204JimenezJ.CaicedoS.PardoJ. M.Gil-OrdóñezA.Alvarez-QuintoR.MollovD.. (2024).
Single torradovirus infections explain the mysterious cassava frogskin disease in the Americas. Sci. Rep.14, p.29648. doi: 10.1038/s41598-024-81142-2, PMID: 39609480JimenezJ.LeivaA. M.OlayaC.Acosta-TrujilloD.CuellarW. J. (2021).
An optimized nucleic acid isolation protocol for virus diagnostics in cassava (Manihot esculenta Crantz.). MethodsX8, 101496. doi: 10.1016/j.mex.2021.101496, PMID: 34754767KumarS.StecherG.LiM.KnyazC.TamuraK. (2018).
MEGA X: molecular evolutionary genetics analysis across computing platforms. Mol. Biol. Evol.2018, 5. doi: 10.1093/molbev/msy096, PMID: 29722887KustinT.SternA. (2021).
Biased mutation and selection in RNA viruses. Mol. Biol. Evol.38, 575–588. doi: 10.1093/molbev/msaa247, PMID: 32986832LeggJ. P. (1999).
Emergence, spread and strategies for controlling the pandemic of cassava mosaic virus disease in east and central Africa. Crop Prot.18, pp.627–pp.637. doi: 10.1016/S0261-2194(99)00062-9, PMID: 41674266LeggJ. P.JeremiahS. C.ObieroH. M.MaruthiM. N.NdyetabulaI.Okao-OkujaG.. (2011).
Comparing the regional epidemiology of the cassava mosaic and cassava brown streak virus pandemics in Africa. Virus Res.159, pp.161–pp.170. doi: 10.1016/j.virusres.2011.04.018, PMID: 21549776LeggJ. P.KumarP. L.MakeshkumarT.TripathiL.FergusonM.KanjuE.. (2015). “
Cassava virus diseases: biology, epidemiology, and management,” in Advances in virus research, vol. 91. (Netherlands:
Academic Press), 85–142.
LeggJ. P.ThreshJ. M. (2000).
Cassava mosaic virus disease in East Africa: a dynamic disease in a changing environment. Virus Res.71, pp.135–pp.149. doi: 10.1016/S0168-1702(00)00194-5, PMID: 11137168LeivaA. M.JimenezJ.SandovalH.PerezS.CuellarW. J. (2022).
Complete genome sequence of a novel secovirid infecting cassava in the Americas. Arch. Virol.167, pp.665–pp.668. doi: 10.1007/s00705-021-05325-2, PMID: 34977988LibradoP.RozasJ. (2009).
DnaSP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics25, 1451–1452. doi: 10.1093/bioinformatics/btp187, PMID: 19346325MeenaR. P.PrabhaK.BaranwalV. K. (2019).
Genome characterization of citrus yellow vein-clearing virus: limited heterogeneity of viral genomes in Mandarivirus-infecting different citrus species. 3 Biotech.9, 348. doi: 10.1007/s13205-019-1876-4, PMID: 31497466MenzelW.JelkmannW.MaissE. (2002).
Detection of four apple viruses by multiplex RT-PCR assays with coamplification of plant mRNA as internal control. J. Virol. Methods99, 81–92. doi: 10.1016/s0166-0934(01)00381-0, PMID: 11684306MohidinS. R. N. S. P.MoshawihS.HermansyahA.AsmuniM. I.ShafqatN.MingL. C. (2023).
Cassava (Manihot esculenta Crantz): A systematic review of its pharmacological activities, traditional uses, nutritional values, and phytochemistry. J. Evidence-Based Integr. Med.28, 2515690X231206227. doi: 10.1177/2515690X231206227, PMID: 37822215NguyenH. D.TranaH. T. N.OhshimaK. (2013).
Genetic variation of the Turnip mosaic virus population of Vietnam: A case study of founder, regional and local influences. Virus Res.171, 138–149. doi: 10.1016/j.virusres.2012.11.008, PMID: 23201192OliveiraS. A. S. D.AbreuE. F. M.AraújoT. S.OliveiraE. J.AndradeE. C.GarciaJ. M. P.. (2014).
First report of a 16SrIII-L phytoplasma associated with frogskin disease in cassava (Manihot esculenta Crantz) in Brazil. Plant Dis.98, pp.153–pp.153. doi: 10.1094/PDIS-05-13-0499-PDN, PMID: 30708600PagánI.FraileA.García-ArenalF. (2016). “
Evolution of the interactions of viruses with their plant hosts,” in Virus evolution: current research and future directions. Eds.
WeaverS. C.DenisonM.RoossinckM.VignuzziM. (
Caister Academic Press, Poole), 127–154. doi: 10.21775/9781910190234.05PardoJ. M.AlvarezE.Becerra Lopez-LavalleL. A.OlayaC.LeivaA. M.CuellarW. J. (2022).
Cassava Frogskin Disease: Current knowledge on a re-emerging disease in the Americas. Plants11, p.1841. doi: 10.3390/plants11141841, PMID: 35890475PardoJ. M.ChittarathK.VongphachanhP.HangL. T.OeurnS.ArinaitweW.. (2023).
Cassava witches’ broom disease in Southeast Asia: A review of its distribution and associated symptoms. Plants12, p.2217. doi: 10.3390/plants12112217, PMID: 37299196PinedaB.JayasingheU.LozanoJ. C. (1983).
La enfermedad ‘cuero de sapo’ en yuca (Manihot esculenta Crantz). ASIAVA4, 10–12.
Rodríguez-NevadoC.MontesN.PagánI. (2017).
Ecological factors affecting infection risk and population genetic diversity of a novel potyvirus in its native wild ecosystem. Front. Plant Sci.8, 1958. doi: 10.3389/fpls.2017.01958, PMID: 29184567Rosero AlpalaE. A.Ceballos LascanoH.Rodríguez HenaoE. (Eds.) (2023). Aportes y perspectivas del mejoramiento genético de yuca en el fortalecimiento de su red de valor en Colombia (Mosquera-Colombia:
Corporación Colombiana de Investigación Agropecuaria (Colombia: AGROSAVIA).
SiriwanW.JimenezJ.HemniamN.SaokhamK.Lopez-AlvarezD.LeivaA. M.. (2020).
Surveillance and diagnostics of the emergent Sri Lankan cassava mosaic virus (Fam. Geminiviridae) in Southeast Asia. Virus Res.285, 197959. doi: 10.1016/j.virusres.2020.197959, PMID: 32407870TajimaF. (1989).
Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics123, 585–595. doi: 10.1093/genetics/123.3.585, PMID: 2513255Thomas-SharmaS.Andrade-PiedraJ.Carvajal YepesM.Hernandez NopsaJ. F.JegerM. J.JonesR. A. C.. (2017).
A risk assessment framework for seed degeneration: Informing an integrated seed health strategy for vegetatively propagated crops. Phytopathology107, 2. doi: 10.1094/PHYTO-09-16-0340-R, PMID: 28545348VierstraeteA. R.BraeckmanB. P. (2022).
Amplicon_sorter: A tool for reference-free amplicon sorting based on sequence similarity and for building consensus sequences. Ecol. Evol.12, e8603. doi: 10.1002/ece3.v12.3, PMID: 41735180
Edited by: Pei Li, Kaili University, China
Reviewed by: Patrick Chiza Chikoti, Zambia Agriculture Research Institute (ZARI), Zambia