Optimum Time for Harvesting Cassava Tubers to Reduce Losses Due to Cassava Brown Streak Disease in Northeastern DRC

The present study aimed to determine the appropriate time to harvest cassava tuberous root which minimize the losses due to cassava brown streak disease (CBSD) in the region of Yangambi, DRC. To achieve the aim of the study, 38 cassava cultivars were evaluated in Yangambi INERA’s Research Center for CBSD in the roots at harvest time between 9 and 13 months after planting (MAP). All the 38 cultivars tested showed CBSD root necrosis symptoms. Foliar symptoms occurred on 37.6% of the evaluated cultivars while CBSD root necrosis varied significantly among cultivars (7.0% to 82.5%) depending on susceptibility and the age of plant. This indicates the differential response of the cultivars to CBSD infection. Whitefly population density decreased with age of cassava, it was of 3 whiteflies per plant (9 MAP) to 1 (10 MAP). We noticed that in older plants, whitefly population decreased from 1 at 11 MAP to none at 13 MAP. Although, some cultivars did not show CBSD symptoms up to 12 MAP, they were not necessarily less attractive to whitefly. Negative relationship (r = -0.08 and r = -0.25) has been found between whitefly number and foliar symptom severity or between whitefly and tuber necrosis severity. Beyond 12 MAP, CBSD necrosis (severity score 4) was present in the tubers of 3 cultivars (EUR/2011/0148, Yafelamonene and Ybi/2011/258). Our study shows that in order to mitigate the losses due to CBSD necrosis, the optimum harvesting time for cassava tubers in Yangambi is 9 MAP.


Introduction
Cassava (Manihot esculenta Crantz) is a staple crop in Sub-Saharan Africa (FAOSTAT, 2009;Monde et al., 2013). In the Democratic Republic of Congo (DRC), cassava is economically important as it is a basic food for 70% of population, who consume both the leaves and the tubers (Molongo et al., 2015;Monde et al., 2013;Akinpelu et al., 2012). Cassava tubers harvest can take place between 9 and 12 months after planting (Monde et al., 2012), but in spite of this time range and its adaptability to pedoclimatic conditions, cassava yields in DRC remain low at < 5 t/ha (Vanlauwe et al., 2011). Cassava brown streak disease (CBSD), cassava mosaic disease (CMD) and bacterial diseases are the principal causes of low yields in DRC (Bakelana et al., 2018;Ingbabona, 2015;Legg et al., 2007).
CBSD is caused by two viruses belonging to the Ipomovirus genus, Potyviridae family. These viruses have been identified as cassava brown streak virus (CBSV) and Ugandan cassava brown streak virus (UCBSV) (Winter et al., 2010;Alicai et al., 2007;Monger et al., 2001). This viral pandemic was first described by Storey (1936) in Tanzania. In DRC, CBSD was first reported by Mahungu et al. (2003). Recently, both CBSV and UCBSV have been identified in DRC (Casinga et al., 2018;Mulimbi et al., 2012). The CBSV spread is attributed mainly to the use of infected cassava cuttings, although whitefly may also be a source of infection (Njoroge et al., 2017;

CBSD Incidence
CBSD foliar symptoms were observed on both all of the local and the improved genotypes. Differences in the average disease incidence did not vary (3.2%) among the 38 varieties (Table 1, 2). However, incidence of root necrosis varied significantly (p < 0.01) among varieties ranging from 7.0% to 82.5% (Table 1) and from 7% to 66% from 10 to 12 MAP respectively (Table 2). We also found that six (15.8%) of the improved cultivars had low CBSD root incidence (7.0%) nonetheless they displayed high necrosis severity:   Note. MAP = Month after planting. ab, a, b letter indicating the high significance difference at P < 0.001.

CBSD Severity
The mean foliar CBSD severity scores varied significantly (p < 0.01) among varieties ranging from 2.0 to 3.0. The lowest foliar CBSD severity score (2.0) was observed in two local varieties (Kotumbakotumba and Yafelamonene) and the six improved landraces here above cited to display high necrosis severity:  Very high positive correlation (from r = 0.74 to r = 0.71 at p < 0.01) was obtained between root incidence and root severity and between foliar incidence and foliar severity. Additionally, there was a moderate correlation (r = 0.54, p < 0.05) between CBSD foliar incidence and root incidence. No significant correlation (r = 0.07, p < 0.05) was found between CBSD foliar incidence, and there was negative correlation (r = -0.08, p < 0.05) between foliar severity and whitefly number (Table 4). Note. ***: Correlation significant at p < 0.01 level; **: Correlation significant at p < 0.05 level; *: negative correlation; ns: non-significant correlation.

Favorable Moment for Harvesting Cassava Tubers
Our data in Tables 3 and 1 show that gradually all the cassava cultivars developed tuber necrosis symptoms from 9 to 13 MAP. Our results in table 3 show that at 9 MAP, 35 of the 38 cultivars produce the best yield (CBSD severity score 1). Exceptions to this are the remaining 3 cultivars exhibiting disease, namely Ybi/2011/258 (severity score 3), Vuvu (TMS/94/0330) and EC/2014/0115 (both with severity score 2).
At 12 MAP, 22 of the 38 cultivars could be harvested (Table 3) At 13 MAP, 38 of the 38 cassava cultivars assessed had exhibited severe necrosis in their tubers (score 3 to 5), thus making them non-harvestable and unsuitable for human consumption. Consequently, the optimum harvest time for cassava tubers under Yangambi conditions is deemed to be 9 MAP (Table 3). Table 1 shows that the cultivars which attracted a high density of whiteflies were MVZB/0015 (3 whiteflies/plant) followed by PM/Ybi/032, EC/2014/0115, Obama2 and Vuvu (TMS/94/0330) (2 whiteflies/plant respectively). During our investigations, we noted that the average whitefly population decreased from 1 (at 11 MAP) to zero (at 13 MAP) whiteflies per plant ( Table 2). The above results suggest that the cultivar type is the one factor influencing the whitefly abundance.

Discussion
The principal objective of this study was to determine the most favorable time after planting to harvest cassava tubers in order to reduce CBSD losses. To achieve this, we assessed CBSD incidence and severity, using five harvesting times (9,10,11,12,13  The important finding from this study showed that from 9 to 12 MAP, tubers CBSD incidence varied significantly (7.0% to 66% respectively), and foliar CBSD incidence did not vary and remained at 3.2%. Furthermore necrosis severity on tubers also varied, from score 2 to score 3. This implies that the harvesting time is influenced by the disease spread and is enhanced symptom severity in the roots. This situation might be explained by the factors that affect tuber CBSD necrosis which might have been affected by plant age, the cultivar (genotype) and environmental variations like rainfall, altitude, temperature, and soil conditions (Kaweesi et al., 2014;Mbanzibwa, 2011;Hillocks & Jennings, 2003;Nichols, 1950;Storey, 1936). At the other hand, relatively low temperatures that occur at high altitude areas (> 3500 feet) during winter can result in severe CBSD symptoms manifestation and eventually death of plants-as observed by Nichols (1950). Based on our study, we think that the high relative temperature (high temperature 33.5 °C) that occurs in Yangambi, the low altitude (400 m) zone could account for the reduction in foliar chlorosis. The significant difference in time of harvesting found in Uganda by Kanju et al. (2019), and by Kaweesi et al. (2014) is similar to the finding from our work. Previous studies had indicated that there can be variation in cultivar response to CBSV infection (Hillocks & Jennings, 2003;Calvert & Thresh, 2002).
The mean foliar CBSD severity scores varied significantly (p < 0.01) among both local and improved varieties ranging from 2.0 to 3.0. The root necrosis severity scores were significantly (p < 0.01) different among varieties ranging from 2.0 to 4.6. Furthermore, we found that six (15.8%) of the improved varieties had low CBSD root incidence (7.0%) when they displayed higher necrosis severity: EC/2014/004, EC/2014/015, EC/2014/0031, EC/2014/05, PM/Ybi/01 (score 3.0 respectively) and EUR/2011/0148 (score 4.0). This variation may be due to the low titer of the virus and on the other hand, to the susceptibility of the varieties to be infected. These results are in part similar to those observed by Abaca et al. (2012) in Uganda for severity of foliar and root necrosis which significantly differed among varieties. However, our results confirm the assumption that the symptoms generated by CBSD vary considerably according to the sensitivity of the cultivar (Kanju et al., 2019;Kaweesi et al., 2014;Abaca et al., 2012). With respect to the optimum time for harvesting cassava tubers and the cropping system practiced (association of many cassava varieties on the same land portion), the most important finding from our results show that cassava should be harvested at 9 months after planting due to the low level of tuber necrosis (severity score 1)-except for the improved cultivars, Ybi/2011/258, Vuvu (TMS/94/0330) and EC/2014/0115.
Beyond 9 MAP, the number of harvestable cultivars decreased noticeably, from 32 at 10 MAP to 22 at 12 MAP (Table 3). Since the vegetative cycle for cassava is 12 MAP, this timeframe is the limit at which farmers can mitigate their losses to CBSD in the Yangambi region.
Beyond 12 MAP, all 38 cultivars displayed a high level of CBSD symptoms in their tubers (average severity score 4). At this stage, tubers are non-edible and non-marketable, leading to serious yield losses for farmers. Given the severe CBSD presence in Yangambi region, we recommend harvesting cassava tubers at 9 MAP.
Furthermore, our results show that the two following local and improved cultivars respectively, exhibited high CBSD foliar chlorosis (severity score 3.0) and less CBSD necrosis in tubers (severity score 2.0): Itchatchabindo, Ybi/2011/0184. It is known that tuber necrosis can be delayed in certain cultivars, such as Nanchinyaya, which was identified in Tanzania (Abaca et al., 2012), and AR40-6, and Namikonga also known as Kaleso (Kulembeka et al., 2012), in Uganda by Kaweesi et al. (2014) to display more foliar symptoms but no root symptoms. Tuber necrosis seems to appear after 5 or 6 MAP for the more susceptible cultivars, but in the case of Nanchinyaya, necrosis was delayed until 12 to 18 months (Bakelana et al., 2018). Thereafter, it is likely that also these cultivars from our work possess characters that will interest CBSD resistance breeding and developing varieties for specific adaptation to the infected region (Kanju et al., 2019;Kaweesi et al., 2014).
Our results showed that the number of whiteflies registered per plant varied considerably between the cultivars tested. Besides, the analysis of our results also shows that cultivars not displaying CBSD symptoms up to 12 MAP were not necessarily the less attractive to whitefly. Moreover, our data showed negative correlation (r = -0.08 and r = -0.25) between whitefly numbers found on the five top leaves of the cultivars studied and their foliar and root severity (Table 4). This finding may be explain by Van-Halder and Van-Helden (1986), who suggested that plant age may be a key determining factor in whitefly attraction. They will be useful to cassava breeders when considering CBSD-tolerant varieties and clones for wider distribution (Alicai et al., 2016).
In conclusion, the CBSD necrosis assessment outlined in this manuscript shows that the optimum harvest time for cassava tubers under Yangambi conditions is 9 MAP. This is the first field trial evaluation in DRC that has addressed the appropriate harvesting time for cassava tubers. Establishing the optimum time, by cultivar, will be useful for cassava growers in their efforts to minimize crop losses attributable to CBSD. In the longer term, this information can contribute towards developing a management strategy against CBSD in Africa.