Effects of sweet potato feathery mottle virus, sweet potato chlorotic stunt virus and their co-infection on sweet potato yield in Western Burkina Faso

Abstract To determine the effects of sweet potato feathery mottle virus (SPFMV), Sweet potato chlorotic stunt virus (SPCSV) and their co-infection on sweet potato yield, twelve sweet potato varieties were assessed in a hotspot area in Western Burkina Faso. The experiment was carried out in a randomized complete-block design with the twelve varieties in three replications. Data were collected on plant growth parameters, plant virus symptoms and yield parameters. Additional testing for selected sweet potato viruses was done using a nitrocellulose membrane enzyme-linked immunosorbent assay (NCM-ELISA) and RT-PCR. SPFMV and SPCSV were the viruses detected in this study. Varieties Djakani and Ligri were virus-free and had the highest average yields out of twelve sweet potato varieties assessed. Field monitoring indicated that 58% of plants were found to be virus-infected. The results suggest that severe symptoms were associated with sweet potato virus disease (SPVD) and yield reduction. However, the interaction of SPCSV with other viruses, which may result in synergistic negative effects on sweet potato yield and quality, needs further research.


Introduction
Sweet potato (Ipomoea batatas) is cultivated and consumed in many tropical and sub-tropical regions, including several countries in Africa (Rey et al. 2012). In Sub-Saharan Africa, sweet potato is the third most important root and tuber crop after cassava (Manihot esculenta) and yam (Dioscorea spp.) (FAOSTAT 2018). Sweet potato is an important food crop and the orange-fleshed sweet potato (OFSP) has the potential to address malnutrition and vitamin A deficiency among children under five and lactating women populations. OFSP varieties which are rich in beta-carotene, a precursor of vitamin A, and other micronutrients are promoted by several Non-Governmental Organizations, as a candidate to prevent malnutrition in children under five (Kimura et al. 2007) cassava and maize were developed. In orange and salmon-fleshed sweetpotatoes, (all-E. However, sweet potato productivity is limited by viral diseases. Wherever sweet potato is grown, viruses are also present ). Because sweet potato is vegetatively propagated (by taking cuttings directly from a previous crop or from sprouted tubers), it is prone to the accumulation of viruses and other pathogens (Souto et al. 2003; Cuellar et al. 2015) and 46 were found to be positive. All were symptomless in sweet potato and generated leaf curling and/or chlorosis in Ipomoea setosa. The five most divergent isolates, based on complete genome sequences, were used to study interactions with Sweet potato chlorotic stunt virus (SPCSV. More than 30 (Somé et al. 2015); five of these have been released in 2014, five are being processed for release and many others are at different stages of selection. In Western Burkina Faso, a significant proportion of the cultivated land is used for sweet potato and it is the main food crop. Because Burkina Faso is now actively developing OFSP as a strategy for food security (Somé et al. 2015) and to address malnutrition, studies were conducted to identify major threats to sweet potato production, including viruses. The relationship between virus diseases and sweet potato yield losses has not been yet reported in a Burkina Faso context. This study was undertaken (i) to assess in a hotspot area in Western Burkina Faso the effects of virus infection on yield losses on twelve sweet potato varieties, (ii) to determine which virus or group of viruses has a greater effect on the twelve OFSP varieties production, and, then (iii) to identify promising varieties resistant to SPVD.

Materials and methods
The study was conducted from July to November 2017 at the INERA station of Farako-Bâ (N 11°5' 36.402'' W 4°20› 4.581''), located in Western Burkina Faso. The climate of this locality is Sudano-Guinean type with an alternation of two seasons: a rainy season from June to October and a dry season, from November to May. Western Burkina Faso is the sweet potato production hub and virus hotspot environment. The average rainfall ranges from 900 to 1200 mm. The rainfalls recorded during the experiment times were 120.7 mm, 152.7mm, 118.5 mm and 18.4 mm in July, August, September and October respectively. Minimum/ maximum temperatures during crop duration ranged from 26C/32C to 30C/37C. Relative Humidity ranged from 52% to 81% for the same period. The soil is of tropical ferruginous type with low organic matter (<2%) and, predominantly sandy to loamy texture with a low cation exchange capacity (CEC). The mineral reserves are also low, particularly in potassium (Bado 2002).

Field experiment
Planting materials were multiplied in insect-proof net tunnels and in adjacent field plots at the primary multiplication stage at Kamboinsé near Ouagadougou (Burkina Faso). So, cuttings from each of twelve sweet potato varieties plants were healthy. The experimental design was a randomized complete block design with twelve varieties in three replications. Each variety was planted on one ridge of 4 meters long. The ridges were 1m apart and on each ridge the planting spacing was 30cm. The field was plowed, prepared in ridges, and vines were planted on July 2017. NPK fertilizer (14-23-14) was applied at the dosage of 200Kg. ha -1 21 days after planting. Weeding was done on the 46 th day during the vegetative stage and after according to the need.

Data collection
Data were collected on plant growth parameters (vine length, vine diameter, soil cover), plant virus symptoms and yield parameters. The virus symptoms were recorded using a scale of 1 to 9 as described by CIP in McEwan et al. (2015), and leaf samples were harvested 60 days after planting.

Yield
Storage root number, storage root weight and fresh vine biomass weight per plot were recorded at harvest at 4 months after planting to compute storage root and biomass yield. The root shape, skin colour, flesh colour, level of root flesh oxidation and growth habit for all cultivars were recorded following the CIP/IBPGR descriptors for sweet potato as described by Huaman (1991). For quantitative traits, a mean from five measurements was obtained (Huaman 1991).

Virus symptoms collection
Virus symptoms were recorded monthly from each plot using the 1 to 9 scales (McEwan et al. 2015), with scale 1 for no virus symptom and the scale 9 for severe virus symptom with stunted plants that are dying ( Table 2).

NCM-ELISA for virus diagnosis
To assess the presence of viruses, a nitrocellulose membrane ELISA (NCM-ELISA) test kit with polyclonal antibodies was used according to the manufacturer protocol. The kit was kindly supplied by International Potato Center (CIP), Sub-Saharan office, Nairobi, Kenya, and was able

Data analysis
Data were analyzed using SAS (version 9.4, SAS institute Inc., USA). An analysis of variance was computed for all the collected parameters and the mean comparisons were performed for the quantitative variable to evaluate the level of variability among the varieties and replication using the least significance difference at p<0.05 level. Thereafter, correlation analysis was also computed to estimate the relatedness among variables, especially the relationship between the virus single or co-infection on the storage root and upper ground biomass yield.
Yield performance: The overall yield performance during this experiment was very low due to poor rainfall (410mm). However, the varieties Ligri and Djakani (average 1t/ha and 0.5t/ha respectively) ( Figure 2) that were virus-free had the highest average yields compared to the other varieties evaluated in this study. While BF64-7, BF59xCIP-1-20, BF59xCIP-4 and Kb_Pourpre varieties infected by SPVD had the lowest yields ( Figure 2). The average biomass weight ranged between 0.17 and 1.12 t/ha. Varieties Ligri and BF7755-10 were the highest weight with an average of 1.12 and 0.98 t/ha respectively. Data related to the growth parameters, upper ground biomass and root yield were subjected to an analysis of Pearson correlation coefficients with significant positive correlations between symptoms, viruses, upper ground biomass and root yield ( Table 3).
The Pearson Correlation analysis (Table 3) showed that symptoms were highly (P <0.0001) and positively correlated with SPFMV, SPCSV and SPFMV+SPCSV infections. SPFMV symptom severity was significantly and negatively correlated with upper ground biomass production and storage root yield.
Our results showed that SPFMV had a higher negative impact on upper ground biomass and root yield compared to SPCSV and SPFMV+SPCSV ( Table 3). Out of Caromex, the other varieties BF59xCIP-1-26, BF64-7 and Kb_Pourpre that were in coinfection had a yield of zero tons per hectare (Figure 2b).

Discussion
The rainfall amount and distribution during the experiment was very poor with around 410.3mm of rain being recorded for the minimum rainfall required of 600mm for sweet potato. Therefore, the overall performance of the experiment showed a very low yield. Storage root yield was far lower compared to those reported by Some et al.  Kim et al. 2017). However, the variety Caromex was SPVD infected but had a yield of 0.5 ton per hectare, among the highest. Caromex variety could be considered as SPVD tolerant. The positive correlation between upper ground biomass and storage root yield to SPVD was also reported by Njeru et al. (2004) in Kenya. Indeed, they had found that single SPFMV and SPCSV infection had no significant effect on biomass contrary to SPVD . Likewise, contradictory studies in other countries on the effect of SPFMV on yield of sweet potato cultivars have been reported. Studies have even reported no effects on storage roots and upper-ground biomass yield in comparison with healthy plants (Trenado et al. 2007;Adikini et al. 2016). Some studies reported SPFMV infected plants producing a better yield than the healthy control (Gutiérrez et al. 2003), while others have reported yield reduction of up to 46% (Mukasa, 2004;Njeru et al. 2004;Domola et al. 2008). In this study, the statistical analyses showed the negative correlation between the SPVD and the root yield was not very significant (P= 0.16,). The varieties that were positive for SPVD and had severe symptoms could be considered as susceptible; however, this needs to be confirmed by further studies. Varieties Djakani and Ligri varieties had no SPVD symptoms and were found to be virus free using NCM-ELISA and RT-PCR, they might be considered as resistant varieties. Abidin et al. (2017) reported fewer virus symptoms on Ligri variety in an experiment conducted in Ghana.
SPFMV was the most prevalent virus detected among evaluated varieties in this study, which is in accordance with many studies that reported that it is the most widespread virus on sweet potato crops in the world (Ateka  2017) concluded that Ligri was infected with SPFMV, nevertheless had mild symptoms. SPCSV was present but did not seem to be very important in sweet potato fields in the western Burkina Faso. Typical symptoms of SPCSV observed were chlorotic spots, purpling and yellowing of the middle and mature leaves, similar to symptoms reported by Adikini et al. (2016).
This study has demonstrated that SPVD and generally virus-infection are serious threats to sweet potato production in Western Burkina Faso as yields are significantly affected. Varieties that were virus-free (Ligri and Djakani) had the highest average yields compared to the infected ones. Based on these results, it has been shown that sweet potato variety yield performance depended on the use of disease-free planting material or the use of SPVD resistant varieties like Ligri or Djakani.