Cashew leaf and nut blight disease outbreaks under unimodal rainfall pattern in Tanzania

Cashew (Anacardium occidentale L.) which is an important cash crop in Tanzania, is susceptible to various abiotic and biotic stresses. Cashew leaf and nut blight disease caused by Cryptosporiopsis spp. is a devastating disease of cashew crop resulting in a significant reduction of nut production. We conducted a study to assess the association between weather variables, particularly rainfall and temperature, on cashew leaf and nut blight disease (CLNBD) outbreaks and establish a minimum wetness duration period for initiation and development of the disease. Poisson regression analysis indicated a significant association between rainfall and temperature with CLNBD incidence (df = 3, Chi Pr < 0.001). Rainfall was positively associated with CLNBD and a unit increase of 1 mm in rainfall caused a 2 percent increase in the percentage of disease incidence (e =  + 0.01995, p < 0.011). However, temperature was negatively associated with disease incidence and a unit increase of 1 °C in temperature caused a 21.56% decline in disease incidence (e = –0.2428, p < 0 .001). Blight incidence was significantly higher (p < 0.001) during the rainy season (range: 11.7 to 36.6%) compared to the dry season (1.4 to 11.5%). Cashew trees in Mtwara District recorded higher blight disease incidence as compared to the rest of the studied areas. This study also established that symptoms for blight disease on cashew leaves appear 8 h post exposure to continuous wet periods. Thus, it appears that for the blight pathogen to initiate and complete its infection processes, a continuous wet period of at least 8 h is required. Findings from this study, particularly the association of rainfall and temperature on disease epidemics, can be used as tools for forewarning incidences of CLNBD. This would help timely initiation of appropriate management strategies.


Introduction
Plant diseases continue to be one of the major constraints to agricultural production (Palmgren et al. 2015). These diseases can be caused by biotic factors such as fungi, bacteria, viruses, and mycoplasmas or abiotic factors including light, heat, cold, drought, excessive precipitation, pH, nutrition, chemical, and mechanical injury (Manners 1993). Direct yearly agricultural production losses associated with biotic stress have been estimated to be between 20 and 40% (James 1998;Oerke 2006;Oerke et al. 2012;Savary et al. 2012). Plant disease epidemics are associated with weather patterns and climatic parameters, and each disease can be affected differently by weather parameters (Burt 2002).
In Tanzania, cashew is the topmost cash crop, contributing significantly to the foreign exchange earnings (BOT 2016) and is a source of income for more than 500,000 households (NARI 2018). However, like other crops, many different diseases can affect cashew at different stages of its growth (Freire et al. 2002;Sijaona 2013;Wonni et al. 2017). Diseases such as powdery mildew, blight, anthracnose, dieback (Nene and Sijaona 2017;Sijaona 2013), black mould, angular leaf spot and gummosis (Freire et al. 2002) have been considered as the major challenges in cashew nut production. Similarly, recent cashew attacks by Fusarium oxysporum represent yet another alarming challenge for cashew producers (Tibuhwa and Shomari 2016).
Cashew Leaf and Nut Blight Disease (CLNBD) continues to be a serious destructive disease of cashews often causing substantial yield losses. Investigations of the disease, including pathogen identification in collaboration between the Naliendele Agricultural Research Institute (NARI) and Commonwealth Agricultural Bureaux International (CABI), identified the pathogen at the genus level as a Cryptosporiopsis sp. (Sijaona et al. 2006). The disease was originally called "cashew leaf and nut blight disease" because of the pathogen's ability to infect and show symptoms both on cashew leaves and nuts. However, further observation and monitoring have indicated that the disease can also attack flowers and cashew apples. The disease was first observed in 2000 in a cashew field at Itoculo, Monapo District, Nampula in Mozambique, where it was confused with anthracnose. In 2002, similar symptoms were observed in a cashew field of the Newala District, Mtwara, Tanzania (Sijaona et al. 2006), and only five years later, the disease was recorded in all cashew growing regions in the country (Nene and Sijaona 2017). The disease infests young tissues of newly flushed leaves and developing nuts. Tender leaves show dark tan lesions bordered by dark reddish-brown margins on the leaf laminar, while older leaf lesions appear papery, silver or grey colour and the centre often falls out resulting in a so-called "shot-hole" appearance. The developing nuts and young apples become blackened, and most of them drop down prematurely whereas infected older nuts and apples become darkish and slightly sunken (Sijaona et al. 2006). The disease has been reported to be fast-spreading and can cause yield losses between 40 and 100%, if not controlled (Nene and Sijaona 2017;Sijaona 2013).
Studies have indicated that CLNBD has an incubation period of seven days under field observation (Sijaona 2013) and three days under laboratory conditions (Menge et al. 2014). Laboratory investigations on the biology of Cryptosporiopsis sp. identified those environmental factors that favor the growth and development of the pathogen, and have determined that water-based media is an ideal substrate for its growth (Menge et al. 2014).
A study by Menge et al. (2013) investigated for the environmental conditions favouring the growth of Cryptosporiopsis pathogen which include a temperature range of 25 to 30 °C and pH level of 6.0 and 7.0 Experiences of researchers from the Tanzania Agricultural Research Institute (TARI) at the Naliendele Centre and general observations have shown that growth and development of Cryptosporiopsis sp. is affected both by rainfall and temperature and that period of plant wetness is necessary for infection. However, to date, there have been no attempts of investigating the association between rainfall and temperature or wetness duration, which might be necessary for infection and subsequent development of the pathogen. Understanding the conditions favourable to disease development and disease incidence or severity can serve as the basis for not only estimating the negative impact of the pathogen on cashew yields but also in designing timely initiation of appropriate control strategies against the disease. Thus, this study aimed to assess the association between rainfall and temperature on CLNBD epidemics and the minimum wetness duration required for the initiating of the infection by the blight pathogen and further development of the disease.

Description of the study area
We conducted a study from June 2016 through May 2017 to assess the association between weather variables particularly rainfall and temperature on cashew leaf and nut blight disease outbreaks. This study was carried out in four locations in the south-eastern part of Tanzania that produces more than 80% of all cashew nut produced in the country.  (Rejani et al. 2013). The studied area covers the Coastal with altitude from below 20 m above sea level, the hinterlands and the plateaux to above 700 m.a.s.l.
The studied areas receive a unimodal type of rainfall that usually starts in November / December through April/May (Nene et al. 2016), though in recent years there have been several changes of off-season rains which fall in between July and September (Nene et al. unpublished data). The mean annual rainfall ranges from 820 to 1245 mm. The mean temperature is 26 °C in the coastal areas and 24 °C in the inland areas (Dondeyne et al. 2003).

Assessment of CLNBD incidences
Six surveys were conducted (three surveys during the dry season and three during the rainy season) for each of the four locations described above. During each blightmonitoring survey, four unsprayed cashew fields were selected randomly in each location, and ten cashew trees per field were assessed. This means forty trees were assessed in every survey for each location. Assessed cashew fields were those with improved cashew seeds distributed by TARI Naliendele (Masawe and Kapinga, 2017). The cultivars are produced from polyclonal seed gardens, however all these cultivars are found in each of the assessed cashew fields, and have similar disease susceptibility. The sampling unit was a onemetre square quadrat whereby shoots on four sides north, south, east and west and those showing the presence of blight for each tree were recorded. This means forty trees were assessed in every survey for each location. We counted all the shoots enclosed within the quadrat and the shoots that were showing symptoms of the disease. The percentage of blight disease incidence was determined by the number of shoots showing symptoms of the disease divided by the total number of shoots enclosed within the quadrat × 100. Data for rainfall and temperature were collected from the nearby weather station whereby each assessed cashew field was estimated at ten (10) kilometres from the weather station. Masasi and Nachingwea are both found in the hinterland areas with similar weather conditions, therefore, weather information from Nachingwea weather stations represents both areas. Summary of the sampling procedures is shown in Table 1 below.

Determination of leaf wetness duration (LWD)
In this experiment, leaf wetness duration (LWD) was considered as, the period when free water caused by either dew or rain remains deposited on the plant tissues. Therefore, we conducted a study to ascertain the minimum wetness duration required for the initiating of the infection by the blight pathogen and further development of the disease.
A total of thirty holes were prepared on a seedbed at TARI Naliendele centre nursery. Two cashew seeds were sown per hole spaced at 1 m apart. After three months, the stands were thinned to one plant and left for inoculation work (Fig. 1).
About 100 tender cashew leaves infected with blight fungus were collected from Naliendele cashew fields and brought to the laboratory. The leaves were thoroughly washed with distilled water, placed into Petri dishes containing filter paper, and wetted with moisture vapour. The Petri dishes were finally covered and incubated at 30 °C to enhance sporulation. After six days, the leaves were removed from the Petri dishes and the conidial suspension was made by adding 600 ml of distilled water.
The plants were later inoculated by spraying with the prepared conidial suspension. Inoculum concentrations were not determined in order to mimic field situations where disease incidence varies. The plants remained enclosed in wet plastic transparent cage or chamber for 24 h to increase humidity (Fig. 2). Six different wetness duration periods (0 h, 4 h, 8 h, 12 h, 16 h, and 20 h) were used and replicated four times. Preliminary observations indicated no any symptom of blight disease on cashew leaves in four (4) hours post application of inocula. Therefore an interval of four (4) hours duration was chosen and found to be appropriate in this experiment. Appearing of dark brown spots of blight disease were assessed on 10 upper most cashew leaves in five (5) seedlings in each wet duration periods. At the end of each respective wetness period, the plants were gently dried using a table fan. The prevailing temperatures during the experiment fluctuated between 27 to 31 °C and 20 to 24 °C for days and nights, respectively. After 24 h, the plastic bags were removed, and the plants were observed daily for blight symptoms expression for ten (10) days. The percentage of blight disease incidence was determined by counting the number of leaves showing symptoms of the disease divided by the total number of leaves per plant × 100.

Data analysis
We performed Poisson regression analysis for data on blight disease incidence in the fields to determine the association between rainfall and temperature on blight epidemics. Data for blight incidence between locations, season (dry and rainy) within locations and between locations and wetness durations were not normally distributed even after transformation. Therefore, a non-parametric one-way ANOVA (Kruskal-Wallis) test was performed to find out variations between locations. Mann-Whitney U test was used to compare the differences in blight disease incidence between dry and rainy season. Data were analysed using GenStat software 15 th edition.

Association between rainfall and temperature on blight disease epidemics
The Poisson regression results (Table 2) showed significant associations between rainfall and temperature with CLNBD incidence (df = 3, Chi Pr < 0.001). Rainfall was positively associated with CLNBD (Table 2). A unit increase in rainfall caused a 2 percent increase in the percentage of disease incidences (e = + 0.01995, p < 0.011, Exp [β] = 1.02). The temperature was associated with a low disease incidence. A unit increase in temperature caused a 21.56 percent decline in the disease incidences (e = -0.2428, p < 0 0.001, Exp [β] = 0.7844). Figure 3 presents the CLNBD incidences per cashew fields. Generally results indicated insignificances of CLNBD incidences within fields. Results in Fig. 4 indicate that, CLNBD differs significantly (df = 3, Chi Pr < 0.001) with locations when percentage disease incidences for both rainfall and dry seasons are pooled together. The highest disease incidence was 24% in Mtwara whereas cashew fields in Masasi recorded the least blight disease infection (6.6%). Nachingwea and Tandahimba recorded blight disease incidence of 17.6 and 19.7 respectively. Figure 5 shows the CLNBD incidence level between rainy and dry season (without considering locations). Significantly (Mann-Whitney U test = 1167, n = 96, p < 0.001) more incidence of CLNBD was recorded during the rainy season (December to May = 26.5%) as compared to the dry season (June to November = 7.5%). Considering locations (Fig. 6), blight infection was higher in each cashew field of Mtwara, Tandahimba, Masasi, and Nachingwea during rainy season as compared to the dry season. Masasi site recorded lower blight incidence during both rainy and dry seasons of 11.7 and 1.4 respectively compared to other locations. Higher mean percentage of blight disease was recorded in Mtwara District during rainy (36.6%) and 11. 5% during the dry season compared to other locations.

Comparison of blight disease incidence in different locations and season
Results in Table 3 indicate maximum and minimum CLNBD incidences during the rainy and dry season in studied locations. During rainy season, minimum blight incidence was 0% in Masasi while maximum recorded disease incidence was 97.5% in Tandahimba. We recorded blight incidence of between 0.0 and 85.4% during the dry season.
Rainfall during this study ranged from 15.5 mm recorded in December at Nachingwea site (Fig. 7) to 298.6 mm recorded in February at Mtwara site (Fig. 8). Results indicated that the mean temperature ranged from 24.7 °C to 27 °C during the dry period and from 24. 5 to 28.3 °C during the rainy season, whereby the lowest temperature (24.5 °C) was recorded in May in Nachingwea and highest of 28.3 °C was recorded in December in Mtwara. Results in Fig. 9 indicated that, amount of rainfall in March was higher in Tandahimba (274.7 mm) compared to other locations. Temperature in Tandahimba ranged from 24.7 (recorded in May) to 27.0 °C which was recorded in February.
Results obtained six days after the inoculation showed that blight symptoms were vividly clear on cashew leaves, which were wet for more than 8 h. Furthermore, disease incidence on those leaves increased with an increase of wetness period as shown in Fig. 10. The observation indicates that young and tender leaves were more affected than did mature ones.

Discussion
This study revealed the association between weather parameters, particularly rainfall or wetness duration and temperature on CLNBD incidence. The study also compared the disease incidence in terms of weather period (rainy and dry) and locations. The study recorded CLNBD incidence under both natural (through field surveys) and artificial (inoculated) conditions. Results indicate that CLNBD severity was significantly associated with rainfall, temperature, and wetness duration. To the best of our knowledge, this is the first report of analyzing the association of rainfall or wetness duration and temperature on CLNBD epidemics. A higher infestation of CLNBD was positively associated with rainfall, and accordingly a higher disease incidence was recorded during the rainy season compared to the dry season. However, an increase in temperature and the interaction between temperature and rainfall caused a decrease in disease incidence. A similar result was observed by Shafaullah et al. (2011) who reported an increase of blast disease of rice as rainfall increases and a decrease of the disease as temperature increases. The studied areas receive a uni-modal type of rainfall, which is from December through May. Periods of June through November is considered a dry period with almost no rains. However, in recent years because of climate, change there has been repeated several events of rains during this dry period; and these are termed as 'rains of the off-rainy season'. Cashew farmers consider this rain as 'bad rain' as it favours blight disease development. Virulent pathogens require conducive environments to infect susceptible hosts (Pannu 2013;Velásquez et al. 2018). Environmental parameters on pathogen and plants can have favourable, neutral or negative effects on disease development (Velásquez et al. 2018). Conditions such as temperature and humidity influence virulence factors of the pathogen such as virulence proteins, toxins production, reproduction and survival of the pathogen (Velásquez et al. 2018). A work by Sandhu et al.  (2017) indicated that temperature and relative humidity are highly correlated with stripe rust incidence caused by Puccinia striiformis f. sp. tritici. The month with the highest rainfall, temperature, and relative humidity recorded higher disease incidence of leaf spot of Jack fruit compared to other months (Shiblu et al. 2015). Like other fungi such as yeasts (Park et al. 2008), the infection mechanism of the Cryptosporiopsis pathogen is favoured by water. Therefore, the pathogen can be considered as hydrophilic fungi with a strong affinity to water (Babič et al. 2017;Park et al. 2008). The symptoms during observations in the fields showed that the disease infests young tissues of new flush leaves and developing nuts. The attacked leaves show angular lesions, which enlarge and coalesce, causing blighting and defoliation. Leaves showed lesions, which are dark tan and bordered by dark reddish-brown margins on the leaf laminar. This situation was noticed under severe infection particularly at the early stages of disease development. Older leaf lesions appear papery, silver or grey colour and the centre often fall out giving a shot-hole appearance.
Infection by the Cryptosporiopsis pathogen tends to spreads to the developing nuts and young apples particularly during the reproductive phase of the cashew (Nene et al. 2016). Subsequently, infected young nuts blacken and most of them drop down prematurely, resulting in significant yield losses. Infected older nuts and apples become darkish and slightly sunken (Sijaona et al. 2006).
The present study revealed that cashew leaf and nut blight disease incidence levels varied across locations. Factors such as different agricultural practices, the build-up of inoculum, weather conditions, and climate change cause variation in the disease epidemics. Lower blight disease incidence was recorded in Masasi cashew fields. Field observations indicated that cashew canopies in Masasi cashew fields had fewer new flush leaves and tender parts as compared to other areas.
Screen house results experiment show that cashew blight disease grows rapidly when exposed to continuous wet periods of more than 12 h. However, when subjected to less than 12 h of wetness, blight spots were also visible, but of reduced sizes. Disease incidence increased with increased duration of wet periods. Thus, it appears that for the blight pathogen to initiate and complete its infection processes, including penetration of host tissues and colonization, a film of water lasting more than 12 h is required. The findings show that increased wetness duration accelerates infection and trigger the occurrence of blight epidemics. Some pathogens release their spores in the films of water from rain or dew on the surface of the leaf (Hardwick 2002), sot it is also possible that longer wetness periods may increase sporulation of the CLBND pathogen.

Conclusions and recommendations
We recorded CLNBD in all of the studied areas but in varying incidence levels. This study revealed that the Cryptosporiopsis pathogen is favoured by wet conditions, and the presence of wetness condition on tender new flush leaves is associated with rapid growth and development of CLNBD. The TARI at Naliendele Centre has recommended different management strategies, which include the use of fungicides against cashew leaf and nut, blight disease. The majority of the recommended chemical brands contain active ingredients such as tebuconazole, trifloxystrobin or chlorothalonil (NARI 2018). The application of chemicals is done only during the reproductive phase, which is between June and November to protect flushing leaves, flowers, and developing nuts. The findings of this study can be used as disease predictor for forewarning incidence of CLNBD. Based on incubation period for the disease to show symptoms; three days under laboratory study (Menge et al. 2014) or seven days in the field (Sijaona 2013), it is therefore recommended that appropriate time for initiating application of chemicals can be within three days after the fall of rains.