MATERIALS AND METHODS

The study was carried out at Mwea Irrigation Scheme (0°40’S; 37°18’E) (Jaetzold et al., 2006) in Kirinyaga County, Central Province, Kenya. It is in the low altitude agro ecological zone, Lowland Midland zone 4 and 1,100-1200 meters above sea level (Jaetzold et al., 2006). The study was from 22 September, 2010-29 June, 2011 and represented two dry and two wet periods. Description of treatments (irrigation water provision and cropping system) is given in Table 1.

Rice farms with these factors were randomly selected in the whole scheme with help from the management of Mwea Rice Growers Multi-Purpose Cooperative Society. The study was arranged as a 2×3 factorial design and each treatment replicated three times. The first factor was irrigation water provision at three levels flood, sporadic and System for Rice Intensification.

This is alternate wet and dry irrigation where the crop is maintained in semi aerobic conditions during the vegetative phase, followed by shallow flooding after panicle initiation (Nyamai, 2012). The second factor was cropping system also at three levels (main crop, double crop and ratoon). Sampling was done using 1 m² quadrants, which were placed randomly in the farm.

Table 1:	Irrigation water provision and cropping system categories involved in the Mwea irrigation scheme during the study from September 2010- July 2011

Three quadrants were undertaken in each farm. Sampling was done after every 14 days starting from 22 September, 2010 to 2 July, 2011. A different part of the farm was sampled on the subsequent sampling date. The average size of sampled farms was 0.4 ha.

Data taken included, number of M. separatella larva, white heads and tunneled tillers. At harvest 20 hills in a quadrant were randomly selected and from these, yield components data was taken. All the tillers from each experimental plot were dissected and the number of tunneled tillers and M. separatella immatures (larvae and pupae) counted. The count of white hands and tunneled tillers at each sampling was subtracted from previous total and expressed as a percentage. Prior to analysis pest data was normalized using square root transformation (SQRT(X+0.5) while percent white heads and percent stem tunneling were arcsine-square root transformed (arc_p = arcsin (SQRT (p)). Data were subjected to ANOVA and means separated by Fishers Least Significance Difference (LSD) at p<0.05 where differences were significant by use of GENSTAT Version 12 statistical software.

RESULTS

The number of white heads/m² was low in areas of flood water provision and in the Systems for Rice Intensification (SRI). High M. separatella infestation occurred in sporadic water provision farms.

Highest number of white heads occurred in areas where rice was double-cropped and the lowest in the ratoon crop. Analysis of variance revealed no significant differences in the number of white heads between the on-season crop and ratoon cropping system and between flood and SRI methods of irrigation water provision. It was also shown that irrigation water provision (df 2, 17, p = 0.034) and cropping system (df 2, 17, p = 0.002) had significant effects on the number of white heads and there was no interaction between irrigation water provision and cropping system (df 2, 17, p = 0.076). Cropping system (df 2, 17, p = 0.028) had significant influence on the number of tunneled tillers m^–2 but the effects of water provision on this parameter were not significant (df 2, 17, p = 0.059) (Table 2).

The number of M. separatella larvae was highest in the sporadic irrigation water provision and where rice was double cropped. Larvae were present in the double-cropped rice irrespective of method of irrigation. There were no larvae in on- season and ratoon crop where rice cultivation was under flood water and SRI water provision. The ANOVA showed that irrigation water provision (df 17, 2, p = 0.04) and cropping system (df 17, 2, p = 0.015) significantly influenced the number of larvae/m². There was significant interaction between irrigation water provision and cropping system (df 17, 2, p = 0.037) (Table 3).

Fig. 1:	Mean number of whiteheads in different water provision and cropping systems at Mwea irrigation in October-December, 2010 and March and April, 2011

Table 2:	Number of white heads and tunneled tillers in different irrigation water provision schedules and cropping systems at Mwea irrigation scheme
Means in a column with the same letter are not significantly different by Fishers LSD (p<0.05), ns: Not significant

Table 3:	Number of larvae in different irrigation water provision schedules and cropping systems at Mwea irrigation scheme
Means in a column with the same letter are not significantly different by Fishers LSD (p<0.05)

High M. separatella infestation in the scheme occurred in double cropped rice in November, 2010 (6%) and April, 2011 (12%). The October peak was in the double cropped rice while the April peak was in the crop trans-planted in early February. Similar trends occurred in the other two systems (Fig. 1).

Table 4:	Weight of rice grains and percent filled grains in different irrigation methods and cropping systems at Mwea irrigation scheme
Means in a column with the same letter are not significantly different by Fishers LSD (p<0.05), ns: Not significant

Table 5:	Number of productive panicles in different water provision and cropping systems at Mwea irrigation scheme
Means in a column with the same letter are not significantly different by Fishers LSD (p<0.05)

Table 6:	Mean number of white heads/m2 at different growth stages

The lowest weight of grains and percent filled grains were found in the sporadic irrigation water provision. The ANOVA indicated that irrigation water provision (p = 0.030) had significant effect on the weight of rice grains but the influence of cropping system (p = 0.059) was not significant. Irrigation water provision (p = 0.040) and cropping system (p = 0.050) had significant influence on percent filled grains but there was no interaction between irrigation water provision and cropping system on this parameter (p = 0.096) (Table 4). The least number of productive panicles was in all the cropping systems under sporadic water provision, with the lowest in the double cropped rice. Irrigation water provision (p<0.001) and cropping system (p<0.001) had significant effect on number of productive panicles and there was interaction between irrigation water provision and cropping system (p<0.001) (Table 5).

The number of white heads was high at panicle initiation, peaked at booting and decreased slightly towards maturity. The infestation was highest in the sporadic water provision. There were no significant difference in the number of white heads between rice crops under flood water provision and SRI (Table 6).

DISCUSSION

The influence of different planting regimes at Mwea irrigation scheme was evident: high pest infestations were in areas where rice has been double-cropped. These results are consistent with findings from other studies which indicate that if a rice-growing area comprises a mixture of farms sown at different times, there will be many oviposition periods, leading to overlapping generations (Umeh et al., 1992). Litsinger et al. (2009) reported that insect pest damage particularly from stem borers was greater in areas where farmers staggered their planting up to 3 months apart and used a set of varieties that matured in 140-210 days than where farmers planted rice within 1 month, used 120-d varieties and planted at the start of rainy season. In an experiment at Philippine Rice Research Institute (PhilRice) at Agusan del Norte, Batayan et al. (2005) found out that the incidence of whiteheads due to white stem borer, Scirpophaga innotata Walker, damage was lower in rice planted earlier on December, 15 and that rice that was planted late on February, 15 had the highest whitehead incidence These cultural practices are similar to Mwea, where farmers cultivate rice throughout the year and use medium duration varieties like Basmati 370 which has a maturity period of 122 days and (BW 196) which take 135 days to mature (Anonymous, 2003).

The high M. separatella infestation in a sporadic double crop is in agreement with work of Litsinger et al. (2009) who reported that expanded growth period favoured vegetative pest build up in flooded rice cultivation. They also found out that rain fed wetland rice system which may be equated with sporadic irrigation at Mwea was more prone to physiological stresses than flood irrigated crop and this minimizes crop compensation and accentuates losses (Litsinger et al., 2009). It was reported from Philippines that losses of 5-71% were highest in the sites with the smallest rice area in which pests were concentrated and in the poorest soils, which constrained yield compensation (Litsinger et al., 2009). Similarly, high M. separatella occurred in sporadic irrigated areas at Mwea, which were areas with unsuitable soils for rice growing and inadequate supply of water for flooded rice cultivation.

The findings in this study of two M. separatella population peaks at Mwea are consistent with reports of CABI. (1996) of a major peak in November and a minor peak in April. However in this study, the major peak was in April and the minor peak in October. This shift is possibly due to continuous cropping of rice which was not previously, the case before 1998 (Kuriah et al., 2003) which result in the presence of the pest throughout the year. The presence of low pest infestations in the ratoon, contradicts previous findings of high M. separatella infestation in the study area (CABI., 1996). Possible reasons for this may be that due to the recent importance of rice straw as fodder for livestock, there has been a reduction of straw being left standing or stacked in the fields and thus a substantial reduction of the reservoirs for the pest. The ratoon crop also matures much earlier than the double crop and thus, may escape attack.

The evidence of reduction in weight of the grains and productive panicles and increase in number of empty panicles is consistent with other findings that M. separatella infestations cause a reduction in productive tillers and yield components (Litsinger et al., 2006). The 15% white heads damage in this study is within the range observed by Sarawar on stem borer damage on aromatic rice (Sarwar, 2011). The presence of high M. separatella infestations in the double crop planted rice irrespective of whether the rice culture was conventional or SRI, underpinned the importance of double cropping, which can be equated to late planting, on stem borer infestations. This is consistent with findings from a study that tested three planting times, first planting, 14 days before farmers’ planting time, the second, simultaneous with farmers and the third, 14 days after farmer planting. They found out that at three weeks after transplanting, stem borer infestation in the third planting time was significantly higher than the other two (Suharto and Usyati, 2005).

Reports have also been made of low stem borer infestation in rice grown under the System of Rice Intensification (SRI) (Uphoff, 2007) but in this study there was no evidence that M. separatella population infestation was lower in the SRI system than in the conventional flooded method of growing paddy rice.

CONCLUSION

Irrigation methods and cropping systems greatly influence fluctuation of Maliarpha separatella infestation and in case of double cropping it will be necessary to control the stem borer. Efforts should be made to ensure that farmers synchronize planting dates and to include a rice break between two consecutive rice seasons, by growing a suitable dry land crop. Attempts should also be made to research on the influence of System for Rice Intensification on M. separatella infestations.

ACKNOWLEDGMENTS

The authors are grateful to Mwea rice farmers in the study sites for their willingness and cooperation. The funding from the Director, Kenya Agricultural Research institute (KARI) now Kenya Agricultural and Livestock Research Organization (KALRO) and the Secretary/Chief Executive Officer, National Council for Science and Technology (NCST) is acknowledged.