Interactive Effects of Major Insect Pest of Watermelon on its Yield in Wukari, Nigeria

Watermelon is known to be infested by multiple insect pests both simultaneously and in sequence. Interactions by pests have been shown to have positive or negative, additive or non additive, compensatory or over compensatory effects on yields. Hardly has this sort of relationship been defined for watermelon visà-vis insect herbivores. A 2-year, 2-season (4 trials) field experiments were laid in the Research Farm of Federal University Wukari, to investigate the interactive effects of key insect pests of watermelon on fruit yield of Watermelon in 2016 and 2017 using natural infestations. The relationship between the dominant insect pests and fruit yield were determined by correlation (r) and linear regression (simple and multiple) analyses. Multimodel inference was used to define the predictor that impacted on fruit yield the most. Results indicated that, each pest had highly negative and significant (p < 0.05) impact on yield (range of r = 0.78 to -0.92), and that the coefficient of determination (R 2 ) values (which were indicative of the effect of pests or their complexes on yield) did not rise on addition of interaction terms. This reveals a non additive negative impact of insect interactions on the fruit yield of watermelon. This may be due to among others; competition by the pest, phenology, plant defenses or changes in nutritional content of the plant. The need to therefore employ discriminate analysis to ascertain the contribution of each pest to yield loss when multiple pest infest a crop is thus highlighted.


Introduction:
Crop plants [Watermelon, Citrullus lanatus Thunb. (Cucurbitaceae), inclusive] are often infested by complexes of pests with their attendant impact on yield 1 . Infestations are known to suppress the totality of crop agronomic performance 2,3 . The pest complex may occur jointly or in succession differing among localities and seasons. However, many scientists and publications dwell on one or two insect pests vis-à-vis yield while ignoring the complex and their cumulative effect on yield thereby disregarding a valid principle which holds that no single pest can be responsible for all the yield loss in an agro-ecosystem. That concurrent infestation by multiple insect pests does not automatically have additive negative effects on yield have been shown by evolutionary researches on wild plants 1 . Nonetheless, whether this principle applies to all cultivated plants remains a subject for continuous studies. Aside additive negative effects 4,5 , research findings have also shown compensatory or over compensatory effects on yield 1 . Studies aimed at investigating the interactive effects of insect pest complexes on fruit yield of watermelon had hardly been conducted.
Watermelon is an important vegetable fruit crop produced throughout the tropics and the Mediterranean region of the world. It accounts for 6.8% of the world area devoted to vegetable production 6 . Despite its huge nutritional, health and economic values, insect pests remain a critical constraint to its production 7,8 . Aside sap sucking and fruit feeding insects; leaf feeding beetle species are widespread and very critical to its production as they are found infesting the crop throughout it growth stages and their intensities varying among seasons 2,10 . While yield losses of up to 100% due to specific insect pest infestation have been reported on watermelon 7 , hardly any study had presented a statistical model which could be used as a basis for predicting yield vis-à-vis key insect pest pressure. Additionally, since crop pests are inherently part of every agro-ecosystem, quantifying their impact on crop performance is now an important field of study 11,12 . We therefore present here an investigation on the interactive impacts of major leaf feeding beetles (cumulatively), sap-sucking and fruit feeding insects on watermelon fruit yield.

Materials and Methods: Study Site, Field Layout and Management
Field experiments were conducted in the Research farm of Federal University Wukari, Nigeria in 2016 and 2017 early-and late-cropping seasons. Wukari has an altitude of 187m above sea level, an average annual temperature of 26.8 o C, and an average annual rainfall of 1205mm. The study area experiences a warm tropical climate characterized by wet and dry seasons. The wet season starts in April and ends in October with peaks in June and September 13 . A 2 year by 2 seasons field research conducted by Okrikata et al. 9 showed that leaf feeding beetles predominated by Aulacophora africana (Weise), Asbecesta nigripennis (Weise), A. transversa (Allard), Epilachna chrysomelina (Fab.), Monolepta nigeriae (Bryant); sap-sucking insects, mainly; Aphis gossypii (Glove), Bemisia tabaci (Genn.), and fruit feeding insects [Bactrocera cucurbitae (Coq.), Heliothis armigera (Hub.)] are the major insect pest of watermelon in the study site. In this study however, the leaf feeding beetle species were treated as a single taxon since the injury they cause were indistinguishable. Data on natural enemy populations were also disregarded as the focus was on pest populations. Forty 5 m long x 8 m wide plots were demarcated on a 0.21 hectare of field in four replications and treatments applied as reported by Okrikata and Ogunwolu 14 in which aside application of insecticide (0.5 % Cypermethrin 30g/L + Dimethoate 250g/L EC [Cyper-diforce ® ]) at the recommended rate at various growth stages and their combinations; insecticide untreated plots were regarded as the control. Additionally, to suppress the impact of pathogen and weed pests; a broad spectrum preventive fungicide; Mancozeb 80% WP. (Zeb-care ® ) which has a contact mode of action was applied at the rate of 2 kgha -1 at the vegetative, flowering and fruiting stages and, when necessary, weeding was done manually following the method described by Okrikata and Ogunwolu 14 .

Assessment of Insect Population
Sampling of insect species commenced at the 2 nd week after planting (WAP) and thereafter at weekly intervals until maturity of fruit. Leaf feeding beetles and Heliothis armigera larvae were sampled using a motorized shoulder-mounted suction machine having a 10 cm diameter inlet cone (Burkard Scientific Ltd., Uxbridge, UK.). Sampling was done by sweeping the machine through the 5 m length of the middle row of each plot at an approximate walking speed of 1m/second. Whiteflies (Bemisia tabaci Genn.) were sampled using a 15 x 15cm yellow sticky board waved across the 5 m length of the middle row of each plot on shaking the plants as described by Okrikata and Ogunwolu 14 . Estimates of population density of aphids (Aphis gossypii Glove) were made by assessing the colony size on 12 randomly selected leaves/plot using a scale from 0 -9 as described by Okrikata nad Ogunwolu 14 . Similarly, fruits infested by fruit fly were isolated and counted in each plot. Infested fruits were split open and the number of fruit fly larvae therein counted and expressed as number of fruit fly larvae/fruit as also described by Okrikata and Ogunwolu 14 .
Samples of dominant insects collected were identified at the Insect Museum of Ahmadu Bello University, Zaria, Nigeria. However, immature stages were reared to adult in the laboratory before identification.

Assessment of Marketable Fruit Yield
Fruits in a plot were harvested twice at 10 days interval, counted, weighed, and sorted into marketable and unmarketable categories. The latter comprised of fruits that were cracked, discoloured, infected with blossom end rot, misshapen and insect damaged. The proportion of the marketable fruits was then computed.

Data Analysis
The relationship between dominant insect pests and marketable fruit yield were determined by Pearson's product-moment correlation, and linear regression (simple and multiple) analyses. Multimodel inference was used to determine the independent variable that impacted on yield the

Relationships between Marketable Fruit Yield and Individual or Combined Insect Pests of Watermelon
Results presented in Table 1 indicates that all the major insect pests across years and seasons; either individually or combined showed highly negative and significant (range: 0.79 to 0.92; p < 0.05) impact on marketable fruit yield of watermelon. However, correlation coefficients were generally higher and more significant on fruits predisposed to individual than for combination of insect pests indicating that losses caused by combinations of pests were not additive.

Relationship between Marketable Fruit Yield and Major Watermelon Pests without or with Interactions
The coefficient of determination (R 2 ) values presented in Tables 2a and b indicates the interactive and non interactive effects of pests on fruit yield of watermelon. Across years and seasons; the results revealed that addition of interactive terms did not increase the R 2 values. The generally comparatively lower R 2 values observed in the interactions, showed that the effect of pest complexes did not explain all the losses in fruit yield. All the regression analyses consistently and significantly (p < 0.05) followed the linear model, and their R 2 values ranged from 62.4 -85.2% for interactive insects; and from 73.

Discussion:
It is known that injuries caused by crop pests lead to damage, and that damage leads to yield losses 12,15 . However, the extent of these relationships can be better expressed by linear regression analyses. Assessing the impacts of pests on crop performance has been shown to be very important in modelling 15,16 . Kirmse and Chaboo 17 reported that though chrysomelid beetles are season-long pests of cucurbits, they are most attractive during the 1 st 2 to 3 weeks post emergence. The ability of leaf-eating beetles to weaken seedlings and/or bring about loss of plant stands resulting to yield loss has been demonstrated by Konstantinov 18 . Our multimodel inference analyses showed that of the major pest of watermelon, the leaf-eating beetles, had the highest impact on fruit yield. It has been shown that leaf injury has serious implication on the quantity and quality of fruits produced by watermelon as the leaves play a major role in manufacturing sugar and gathering water in the fruit 19,20 .
Throughout the 2 years study, presence of H. armigera (a key fruit boring insect) in the earlyseason crop was sporadic. The early-season crop growth period was characterized by higher intensity and frequency of rain which might not augur well with H. armigera colonization, and population rise. Alternating dry and wet spells have been shown to favour its outbreak 21 . Another predominant, fruit boring insect was B. cucurbitae. The damage caused by fruit borers is very obvious. Yield losses of 30 -100% have been attributed to fruit feeding insects alone 22 .
The sap sucking insects predominated by A. gossypii and B. tabaci have direct and indirect effect on yield. Though, assessing their impact on yield was difficult as it was largely physiological, they are known to, aside sucking sap; leading to stunted growth and curly leaves, vector pathogens, and excrete honey dews which attracts fruit flies 23 .
It is therefore obvious that each of the 3 pest groups (leaf feeding, sap sucking and fruit boring) affects yield negatively, and that the individual species attack watermelon crop either simultaneously or sequentially 9 . But are their interactions additive? As evidenced by results of correlation and regression analyses, the current finding showed non additive negative interactions by the component pests. That the correlation coefficients were slightly higher and significant in plants not predisposed to multiple pests, and that R 2 values largely did not improve under interactive pest infestations, so indicate. However, of interest is that the multimodel inference analyses identified the leaf feeding beetles as the most determinants of the negative yield.
While a finding similar to the current was reported by Jörg 24 when investigating the interactions of foliar pathogens on Wheat growth indices; the current finding is at variance with that of Vesna et al. 1 who reported that, two stem and seed weevils known to individually suppress yield on Winter oil seed rape 25 ; jointly increases (positive impact) yield by > 2 folds. Hence, their findings showed non additive positive impact on yield. Arguably, as reported by Stephens et al. 5 ; the interactive effects of multiple insect attacks have been hinged on competition by insect pest components for resources, seasonal variations, plant defensive response and nutritional quality change.

Conclusion:
The current finding shows that the impact of the component insect pests on fruit yield reduction in watermelon was not additive. This may be attributed to competition among pest species, influence of natural enemy species, seasonal variations, plant defenses and changes in nutritional quality of the plant. Hence, no single insect pest is completely responsible for the totality of yield loss even though the leaf beetles had more impact. The need to recruit discriminate analyses when multiple pest infest a crop should be a focus.