Influence of Agro Ecology on Rice Varietal Resistance to Sitophilus oryzae (Coleoptera: Curculionidae) and Sitotroga cerealella (Lepidoptera: Gelechiidae) in Benin

The rice weevil, Sitophilus oryzae (L.) and the Angoumois grain moth, Sitotroga cerealella (Oliv.), are serious pests of stored rice in Africa. This study aims to evaluate the influence of growing environments to rice resistance to these pests. Thus, eleven rice varieties including 6 upland NERICA, 2 Oryza glaberrima and 3 O. sativa were grown at four agro ecological zones of Benin. After harvest, samples of 1500 grains of each genotype were infested with 20 adults of S. oryzae (10 males and 10 females) and 1500 additional ones were infested with 50 eggs of S. cerealella. Results showed significant effect of agro ecological zones on pest incidence and on varietal resistance as well. O. glaberrima varieties (TOG 5681 and CG 14) were the most resistant in each location whereas the resistance of NERICA and Sativa varieties varied from tolerant to susceptible according to the growing ecology. This result highlights the impact of growing environment on rice resistance status and will provide the best advice to farmers on how to choose best genetic material according to cropping ecology.


Introduction
The rice weevil, Sitophilus oryzae (L.) and the Angoumois grain moth, Sitotroga cerealella (Oliv.), are serious pests of stored grains [1].Female moths deposit eggs on grain kernels.After hatching, larvae begin feeding on the endosperm or germ.Then the last instar larva creates a large hole to allow the adults exiting out of the grain [2].S. oryzae deposits its egg directly into the kernel where the larva hatches and completes development until the adult stage, and gets out throughout the hull [3].Most of the losses occurring during cereal storage can primarily be attributed to biotic agents such as insects, diseases, and rodents ([4] cited by [5]), but the contribution of insects is predominant.According to FAO [6], half of the storage losses are usually caused by insects.Stored grain pest damage differed in different stored grains, including rice causing serious economic loss nationally [7].Because of the high adoption of improved varieties in Benin such as NERICA and the subsequent high production between 0.5 and 1.1 t per ha depending of the country in SSA [8], grain storage has gradually become a common practice by farmers [9].Consequently, storage losses due to insect pests, which were not considered as key issue in Benin, began a serious threat for producers, processors and traders.Also, Togola et al. [9] reported that, the incidence of insect pest to the different farmer storage units depended on the storage duration and regions where samples were collected.Thus, economic losses in paddy storage are estimated at 2 FCFA (Franc of the African Financial Community) for each kilo of rice stored over 2 -3 months.Varietal resistance appears as good solution of stored rice losses [10].Several studies designed genetic traits and morphologic characters as key parameters contributing to varietal resistance ( [11]- [13] cited by [14]).Also, Cogburn and Bollich [15] noted that environmental or biochemical factors may also influence susceptibility or tolerance of rice varieties.But some recent studies such as those of Arthur et al. [3] noted that growing environment can also influence susceptibility to stored products insects.This ecological factor needs to be explored as key option of habitat management of storage insect pests.Therefore the objectives of this study were to identify environmental influences on resistance to S. cerealella and S. oryzae in NERICA varieties and their parents.

The Study Area
This study was conducted in North, Center and South region of Benin.Malanville (North) lies between 11.5˚ and 12˚ latitude from north to south and 50 km from east to west on 60 km [16].It climate is Sudano Sahelian type with one dry season (from November to April).The average rainfall is 750 mm per year and temperature differences vary between 16˚C and 25˚C.Soils are sandy clay, ferruginous [16].Rainfall during the wet season (from June to October 2012) was 728.8 mm, the mean temperatures ranged from 26˚C and 29˚C and relative humidity varies from 70% to 81% [16].Glazoue (Center) lies between 7˚30' and 8˚90' latitude north and 2˚05' and 2˚22' longitude East [17].It has a sub-equatorial climate with two rainy seasons and two dry seasons.In recent years due to climate change, we observed a late onset of rains and dry pockets inside the rainy season.The soil is tropical ferruginous with several variants namely sandy white, sandy black and hydromorphic [17].Rainfall during the wet season (from June to October 2012) was 838.6 mm, the mean temperatures ranged from 25˚C and 28˚C and relative humidity has varied from 77% to 81% [17].Lokossa (South) lies between 6˚37'60'' latitude North and 1˚43'0'' longitude East and is characterized by a subequatorial climate.The rainfall average varies between 900 and 1100 mm per year.The soil is mostly hydromorphic [18].Rainfall during the wet season (from June to October 2012) was 857.5 mm, the mean temperatures ranged from 26 and 28°C and relative humidity has varied from 82 to 85% [18].International Institute of Tropical Agriculture (IITA) is located in the town of Abomey Calavi.His geographical coordinates are 6˚25' north, 2˚19' east.His climate is of type subequatorial.The soil is sandy [19].Rainfall during the wet season (from June to October 2012) was 748.9 mm, the mean temperatures ranged from 26 and 28°C and relative humidity has varied from 82% to 85% [19].Figure 1 below shows different sites of study.

Plant Materials
Eleven rice varieties were selected representing the six upland NERICA certified and grown (NERICA1, 2, 3, 4, 14 and 15) by several African countries as well as their parents, three O. sativa (WAB 5650, WAB 56104 and WAB 18118) and two O. glaberrima (TOG 5681 and CG 14).These varieties were collected from Africa Rice Center (AfricaRice) at Cotonou, Benin.

Experimental Design and Crop Management
The first experiment was carried out with the objective to evaluate environmental influences on phenotypic variability of rice during rainy season 2012 in the four agro ecological zones of Benin mentioned above.A randomized complete block design with 3 replications in each site was used.Plot size for each genotype was 5 m 2 with 25 rows, 1 meter long and 40 cm apart from each other.Plots were fertilized with 200 kg•ha −1 as a basal dressing at 21 DAS and with 50 kg•ha −1 of urea at 45 DAS.Standard agronomic practices (such as weeding and fertilizers application) were followed during crop growth stages.Number of panicle and 1000 grains weight were recorded on four plants randomly identified and labeled, selected from the middle row in each experimental plot.At maturity, number of panicles and 1000 grains weight were taken.Before sowing, soil samples at 0-20 cm depth were collected separately at each block to determine major chemical components and soil physical characteristics.A second experiment including screening of rice varieties for resistance to S. cerealella and S. oryzae was set up.Seeds were stored at 4˚C for two weeks to kill any insects or mites present into the samples.Insects reared in the laboratory were used for this experiment.The laboratory was conditioned at a temperature of 25˚C ± 1˚C, a relative humidity of 70% ± 5% and 12/12 hours light/dark.The following methods are used for the screening.

Classification Status of These Insects
• Rice weevil

Statistical Data Analysis
The collected data were subjected to analysis of variance with two factors (ANOVA) to determine the site-varieties interactions.When the interactions are significant, another ANOVA was performed for each site to evaluate significant parameters that established the resistance of varieties to the 2 insect pests.The Student Newman Keuls test (SNK) was also performed to separate the various means of parameters.The data were analyzed with the software Genstat release 14.

Variation of Physic-Chemical Components of Soil from Study Locations
Based on Jamagne [25] triangle test, the soils textures from Lokossa and Glazoue were found to be sandy silty, those from IITA station were sandy and those from Malanville were silty.The results about the Hydrogen potential analysis showed that soils from IITA station and Malanville were acidic, and those from Lokossa and Glazoue were slightly acidic.The soils of four locations displayed a low Carbon and Nitrogen ratio (C/N), which was between 10 and 14 with a low Effective Cation Exchange Capacity (ECEC).The concentration in Phosphorus was low in soils from Lokossa, Glazoue and Malanville but high in soil from IITA.Table 1 showed the results of physico-chemical analysis of the soils from each location.

Climatic Variations in the Different Locations
The graph above (Figure 2) shows the variation in rainfall during the study period in the different study areas.Histograms of rainfall values of IITA, Lokossa and Glazoue have a bimodal rainfall, with rainfall maxima in June and October at IITA and July and September at Lokossa and Glazoue.The rainfall histogram of Malanville meanwhile has a unimodal rainfall regime with a small pocket of drought in August probably due to climate perturbations.The following two curves (Figure 3 and Figure 4) show the temperature and humidity fluctuations during the rainy season (June to October) in four study areas.In South of Benin (IITA and Lokossa), the air temperatures are constant (25˚C -27˚C) and humidity is high (75% -85%) while in the Centre (Glazoue) and the North (Malanville), temperatures are high (28˚C -29˚C) and humidity is also high (80%).

• Number of panicle per plant
The combined analysis of variance for number of panicles per plant showed highly significant (P < 0.01) differences between locations and varieties but no interaction was observed between these 2 parameters.The highest    panicles number was produced at Glazoue followed by Malanville with an average of 9.2 and 9.1 respectively; whereas, the lowest panicles number was recorded at Lokossa (6.758) and IITA (5.879).Highly significant differences (P < 0.01) existed between varieties from IITA.The O. glaberrima varieties TOG5681 and CG14 have obtained the maximum number of panicles per plant 10.9; 11.8 respectively, while the NERICA varieties and their sativa parents (WAB5650, WAB56104 and WAB18118) have recorded between 3.4 and 5.5 panicles per plant.The same clustering was observed in the other locations but the number of panicles per plant was higher than in the previous site (Table 2).For each variety, the number of panicles varied significantly according to the growing • Weight of 1000 grains The analysis of variance showed significant differences (P < 0.01) of 1000 grain weight according to growing locations, varieties and locations across varieties.The 1000 grain weight of cultivars grown at Glazoue, Malanville and Lokossa were higher than the cultivars from IITA.Varieties of O. sativa and NERICA had higher grain weight than varieties of O. glaberrima (Table 2).

Varietal Resistance to S. oryzae and S. cerealella • Resistance to S. oryzae
The analysis of variance (Table 3) for rice resistance parameters showed highly significant (P < 0.01) differences between locations, varieties and varieties across locations interactions (Table 3).The clusters analyses between locations showed that rice varieties grown in IITA were more susceptible (with 39.24 damaged grains) compared to Lokossa (5.15), Glazoue (1.61) and Malanville (1.22).Nevertheless, damage were highly significant (P < 0.01) between varieties from IITA station, Lokossa and Glazoue but not significant (P < 0.05) for the varieties from Malanville.The O. glaberrima CG14 and TOG5681 showed low numbers of damaged grains at all locations (N ≤ 12 damaged grains).Among the cultivars grown in IITA, the varieties NERICA1, NERICA14, NERICA3, NERICA4 and the O. sativa parent (WAB 5650 and WAB 56104) had high number of damaged kernels ranging from 47.3 to 87.3 damaged grains.Conversely, the varieties NERICA15 and the O. sativa parent WAB 18118 have recorded intermediate number from 19 to 20 damaged grains.However, for the remaining three localities (Lokossa, Glazoue and Malanville) the same varieties have obtained very low damaged grains ranged from 0 to 12 grains maximum (Table 4).Adult progeny of S. oryzae were not significant (P < 0.05) among the cultivars from Lokossa, Glazoue and Malanville except for the cultivars from IITA station.For the IITA station, the highest number of insects were observed on varieties NERICA 1, NERICA14, NERICA3, NERICA4 and parents O. sativa WAB 5650 and WAB 56104 ranged from 27 to 53 insects while the varieties NERICA2, NERICA15, WAB 18118, CG14 and TOG5681 have had the lowest numbers from 20 to 25 insects.The insects population did not increased on the varieties from the three remaining locations (Table 4).Signification to Test F: NS = Not significant at 5%, ** = significant at 1%, CV= Coefficient of variation; Means followed by the same letter not significant at 5%.

• Resistance to S. cerealella
The results from the analysis of variance for rice resistance parameters (Table 5) showed highly significant (P < 0.01) differences between locations, cultivars and locations × cultivars.The number of damaged grains from IITA location was greater than those from other sites.This trend remains the same as well for number of emerged S. cerealella as for number of damaged grains.The analysis of variance for resistance parameters was performed on all varieties from each environment.Significant differences (P < 0.05) existed between the varieties for all parameters for each location.On the cultivars from IITA, adult progeny was most important, except from the O. glaberrima parent CG14 (20 insects).Fewer damaged grains was recorded on TOG5681 (5 grains) and CG14 (16.33 grains) from Lokossa location followed by WAB5650 (19.33 grains) and NERICA4 (20 grains).In opposite, NERICA14, WAB56104 and WAB18118 were the most damaged varieties with 39.67, 38.67, 35.67 damaged grains respectively.The remaining varieties (NERICA15, NERICA2, NERICA1, NERICA3) were moderately tolerant to S. cerealella.As for Glazoue location, CG14 and TOG5681 still drew few S. cerealella population with 6.33 and 7 emerged adults.They were followed by NERICA15 (11.33 individuals), WAB18118 (20.33 individuals) and NERCICA3 (20.67 individuals).As for damage status, WAB56104 and WAB5650 were as much infested as, NERICA14, NERICA1 and NERICA2 ranging from 27 adults to 45.67 adults.The same trend was shown by samples from Malanville location (Table 6).All these results enabled us to identify three groups of varieties according to their level of susceptibility to S. cerealella.The O. glaberrima varieties TOG5681 and CG14 proved the most resistant regardless of the site while the O. sativa (WAB5650, WAB56104 and WAB18118) and some NERICA (NERICA14, NERICA1 and NERICA2) were identified as the most sensitive.The remaining varieties were tolerant regardless of the site.Signification to Test F: NS = Not significant at 5%, ** = significant at 1%, CV= Coefficient of variation; Means followed by the same letter not significant at 5%.

Discussion
In this study, the experiments were conducted in different locations of Benin especially in upland ecologies where climate and soil characteristics were quite different.The result showed that pH values under the different environments were between 5 and 6, suitable enough for rice cultivation [26].As for soil nutrient, it varied significantly from a location to another.The soil of IITA was poorer in organic matter and in nitrogen than the soils of Glazoue, Lokossa and Malanville.According to Ralijaona [27], soil nutrient such as nitrogen (N), phosphorus (P) and potassium (K) ensure the proper development of rice plant.Therefore, they have a significant effect on the grain yield.Also, high variability of soils textures was observed.The component was sandy at IITA station-Benin, sandy silty at Lokossa and Glazoue and silty at Malanville.The sandy silty soils have a greater capacity to retain water than sandy soils and can allow rice to grow better on this type of soil.Genotype-Envi-ronment interaction effect was evaluated in order to describe the morphological variability between varieties used.Genotype across environment interaction was not significant for the agro traits except weight of 1000 grains implying differential response of genotype under four locations for this last character.Similar report was earlier made by Asad et al. [28] and Sreedhar et al. [29] for paddy yield.This result could be explained by the combined action of soil fertility and water regime.Genotypic variability is another factor of resistance to insect pests.Indeed, highly significant differences were observed between the tested materials.The O. glaberrima parents TOG 5681 and CG 14 showed significantly higher panicles number under the different environments.While the O. sativa parents WAB5650, WAB56104, WAB18118, NERICA1, NERICA2, NERICA3, NERICA4, NERICA14 and NERICA15 exhibited low panicle number under the same environments.NERICA varieties and their parents have behaved in the same way for all the characters studied in the different locations.
Phenotypic variability observed in the same genotype under different locations indicated that environmental factors can greatly influence on rice cultivation.The results of this study are similar to those of Steel [30] and Morkinyo and Ajibade [31] who showed that morphological variations can be observed with the same genotype due to environmental conditions such as fertility of soil, water regime, light and temperature.In addition, Moukoumbi et al. [32] reported similar results with the lowland varieties.
Growing NERICA rice varieties and their parents in different environments strongly influenced their susceptibility to infestation by the rice weevil and the Angoumois grain moth.The results indicate that interaction between varieties and location was significant.Significant differences existed between locations and varieties for all parameters studied.Progeny production and damaged grains of both S. oryzae and S. cerealella were greater on cultivars from IITA than at Glazoue, Lokossa and Malanville.NERICA varieties except NERICA2 and NERICA15 from IITA were highly susceptible to S. oryzae while the same varieties from Glazoue, Malanville and Lokossa were resistant.The O.sativa parents WAB 5650, WAB56104, WAB 18118 exhibited the same trend.More S. cerealella progeny and more feeding damage were produced on cultivars from IITA except CG 14. Damage and adult progeny among varieties that were relatively susceptible varied widely among locations.Thus, NERICA varieties and parent O. sativa ranged from susceptible to tolerant in the remaining locations.Similar results were obtained by Cogburn et al. [33] and Arthur et al. [3], who screened different cultivars of rough rice for resistance to S. oryzae and S. cerealella, and in a later test grew cultivars that appeared to be resistant in different geographic locations and tested the seeds from those cultivars for resistance.Some cultivars showed consistent results in progeny production regardless of growing location, while results for other cultivars appeared to be influenced by location.Furthermore, Cogburn et al. [33] reported that environmental influences more strongly susceptible varieties than resistant ones.This seems to be strong evidence for a genetic basis for resistance.The O. glaberrima parents CG14 and TOG 5681 were found to be resistant to infestation by S. oryzae and S. cerealella in all locations confirming the relatively good resistance of O. glaberrima previously reported by Sauphanor [20].Other factor such as pest status can play a key role on varietal resistance independently to the effect of cropping location.For instance, number of damaged grains of CG14 and TOG 5681 was higher at IITA than Glazoue, Lokossa and Malanville to infestation by S. cerealella.Rizwana et al. [23] showed that no variety was completely immune to the infestation of this pest.The same conclusions were reported by Pandey et al. [34], Khatak and Shafique [35], Qayyum [36], Khatak and Shafique [37], Ratnasudhakar [38], Tirmizy et al. [39] and Khatak et al [40].Also, Haryadi and Fleurat-Lessard [41] reported that climatic conditions and cropping techniques of the rice are factors contributing to the physical or chemical characteristics of each variety and consequently can explain different level of their resistance.Authors such as Breese [42], Russell and Cogburn [22], Chanbang et al. [43], Rizwana et al. [23] reported that infestation of S. oryzae and R. dominica were not found in rice grains with an intact husk and the mode of entry was assumed to be through a crack in the hulls caused either by natural means, through breakage from harvesting, or from the drying process.Moreover, Shafique and Chaudry [44] affirmed that resistance in paddy to storage insects has been attributed to various physico-chemical characteristics of rice grains such as intact hulls and tough siliceous hull of rough rice.Positive correlations were observed between protein, moisture, ash contents and the insect progeny, grain weight loss and damage [23].Finally, Russell and Cogburn [22] concluded that more than one mechanism were operating against the insects and that insects developed more slowly on resistant varieties than on susceptible ones.

Conclusion
We can conclude that rice resistance to S. oryzae and S. cerealella is the result of a range of factors, including soil fertility, climatic conditions, water regime, physico-chemical characteristics of grains and pest status.Above all, the environmental factors such as soil fertility (nutrients content) are more important because they can influence the other factors.

Figure 1 .
Figure 1.Map of Benin showing the different sites of study.

Figure 2 .
Figure 2. Rainfall of different locations in 2012.

Figure 3 .
Figure 3. Temperatures of different locations in 2012.

Figure 4 .
Figure 4. Relative humidity of different locations in 2012.

Table 1 .
Physico-chemical characteristics of soil samples collected from different locations.

Table 2 .
Phenotypic variability of 11 rice varieties grown in four agro ecological locations.
Signification to Test F: ** = significant at 1%, CV= Coefficient of variation; Means followed by the same letter not significant at 5%. locations.NERICA varieties had high number of panicles in Malanville site while O. sativa and O. glaberrima parents got high number of panicle in Glazoue.

Table 3 .
Mean squares of resistance parameters (to S. oryzae) obtained from the analysis of variance.

Table 4 .
Adult progeny and damage caused by S. oryzae on 11 cultivars from IITA Station, Lokossa, Glazoue and Malanville.

Table 5 .
Analysis of variance of 3 resistance parameters of rice produced at 4 locations to S. cerealella.