African Crop Science Journal, Vol. 9. No. 2, pp. 411-419
THE EFFECT OF MAIZE STOVER USED AS MULCH ON TERMITE DAMAGE TO MAIZE AND ACTIVITY OF PREDATORY ANTS
M. B. SEKAMATTE, M.W. OGENGA-LATIGO1 and A. RUSSELL-SMITH2
(Received 30 August, 2000; accepted 9 November, 2000)
Code Number: CS01023
Cultural control of termites in agriculture and forestry is attracting renewed interest following the ban on use of persistent organochlorine (cyclodiene) insecticides (Logan et al., 1990). Non-chemical control involves methods which attempt, without the use of commercial pesticides, to (i) prevent termite access to plants, (ii) reduce termite numbers in the vicinity of plants, or (iii) reduce susceptibility/increase resistance of the plants themselves (Logan et al., 1990). Published research on any of these methods is scarce, although numerous cultural procedures have been suggested, including measures to enhance plant vigour and to manipulate termite numbers and behaviour.
The benefits of application of mulch to the soil surface are widely appreciated in tropical agriculture (Schroth et al., 1992) but the possibility of protecting crops against termite damage by use of mulches has remained controversial. Logan et al. (1990) reviewed the conflicting evidence for the use of crop residues and other organic mulches to control termites. Two conflicting theories have been advanced; (i) that removal of residues and other debris from the field will reduce potential food supplies and hence lead to a reduction in termite numbers and subsequent attack; and (ii) that leaving the residues in the field or adding further organic matter will provide alternative food to which the termites will be attracted, thereby reducing levels of attack on crops.
Gold et al. (1991) tested mulches of plants with insect repellent properties (Neem and Ipomea fistulosa) and observed significant protection to groundnuts (Arachis hypogaea (L.) ). This study and other earlier experiments did not relate the quantity of mulch used with the levels of termite attack on groundnuts. In general, however, published studies reporting the significance of noninsecticidal mulches such as maize stover on termite attack on crops are lacking. The significance of crop residues on termite activity in field crops is also poorly understood.
Dawkins (1949) gathered empirical evidence in Uganda that repeated mulching of tree nurseries with grass, over a period of one year, greatly reduced termite foraging. Although not strictly applied as mulch, tree trunks, branches and woody remnants are used in the Philippines and Sierra Leone to reduce termite attack in rice fields (Litsinger et al., 1978; Raymundo, 1986). Leaves of the savanna tree Vitex doniana buried alongside cassava cuttings in Uganda, attracted termites away from cassava, reduced termite damage and led to increased sprouting (Epila and Ruyooka, 1988). The use of wooden logs in the Philippines aside, none of the published works directly deals with non-insecticidal mulches and levels of termite damage to crops.
On the basis of the published literature and the authors observations in recent field surveys in Uganda (Sekamatte et al.-unpublished data), the present study was instituted with two objectives. First to determine the impact of different quantities of mulch on termite damage to maize and second, to assess the impact of mulching on relative incidence of predatory ant species and their role in regulation of termite activity in the maize cropping system.
MATERIALS AND METHODS
Study site. Field experiments were conducted at Namulonge Agricultural and Animal Production Research Institute (NAARI), near Kampala at latitude 32° 27E and longitude 0° 32N, and a mean elevation of about 1200 m above sea level. There are two planting seasons; the first, and normally more reliable one, occurs from March to June and the second from September to November, with mean annual level of 1300 mm. Maximum temperatures are in the range of 25 - 30 °C and minimum temperatures range between 13 - 17 °C.
Effect of mulching on termite damage to maize. Five levels of maize stover mulch were used as treatments. These were: 0 kg (control), 10 kg, 20 kg, 30 kg and 40 kg per plot. The stover amounts were used to provide a surface coverage of 0%, 10%, 40%, 60% and 80%, respectively, in five 8 m x 8 m plots. Maize (var. Longe-1) was planted in the plots at a spacing of 75 cm between plant rows and 30 cm between plants within rows. Two maize seeds were planted per hill but thinned to one plant per hill 14 days after crop emergence. The experiments were planted 28 April 1998 and 17 October 1999. Two weeks after seedling emergence the number of standing plants per plot was determined on the middle five rows of each plot. The maize stover treatments were applied 3 weeks after seedling emergence to cover the middle five rows in the respective plots. The treatments were laid out in a completely randomised block design replicated three and four times in the 1998 and 1999 cropping seasons respectively.
Assessment of termite damage and ant activity. To termite damage on maize in different plots, both destructive and non-destructive sampling methods were used.
Destructive sampling was carried out by randomly uprooting ten maize plants using a hand hoe from each plot. This type of sampling was done once at seedling, silking, green cob, and dry plant stages. The first destructive sampling was done at 4 weeks after crop emergence. Caution was taken not to destroy the root system while uprooting the plants. Presence or absence of termites attacking roots in the different treatments was recorded on each sampling occasion.
In the case of non-destructive sampling 40 plants, randomly selected from within the middle 10 rows of each plot were carefully examined and scored for termite damage once every two weeks till dry cob stage. Termite damage to roots, stems, leaves and cob was scored using a scale of 1 -5, where 1 = no damage, 2 = 1-25%, 3 = 26 -50%, 4 = 51 - 75% and 5 = 76 - 100%. A damage rating of 5 in most cases caused plants to lodge.
During non-destructive sampling for termites, maize plants with evidently established colonies of Lepisiota or Myrmicaria or more than three ants nesting or foraging within 15 cm diameter around a sampled maize plant stem base were recorded in-situ. Ant activity was then expressed as the percentage of maize plants with ant nests out of 40 sampled plants per plot.
Statistical analysis. Treatment effects, percentage of attacked plants and damage scores on roots and stem base were determined by ANOVA procedures. Data were subjected to square root transformation and conformed to assumptions of ANOVA as dictated by tests of normality (PROC. GENSTAT 4.1) and tests of homogeneity of variance. Data for compound damage indices were analysed for individual sample dates. Other data were summarised as seasonal means, which were tested for normality and constant variance as above and subjected to ANOVA using procedures for completely randomised block design using the MSTATC software. To further test for the effects of mulching, orthogonal comparisons were made and regression analysis (GENSTAT) used to test for linear relationships of mulch quantity on the response variables. The means were separated by Fishers Least Significant Difference test at P = 0.05.
Above ground termite damage on maize. The extent of termite damage to maize was significantly (P < 0.05) influenced by the amount of mulch in the plots (Table 1). In the first cropping season, the compound damage index was significantly lower in plots with 30 kg and 40 kg mulch coverage than in the other treatments. The percentages of plants with cut stems in plots with 30 kg and 40 kg mulch cover were also significantly lower than those in plots with 10 kg and 20 kg mulch levels, while the unmulched plots had significantly higher number of plants with cut stems.
In the second cropping season, a similar pattern was observed in termite damage to plants under different levels of mulch. However, the effect of mulching was more pronounced than in the first cropping season possibly because of a higher termite infestation in the experimental field. All levels of mulch led to significantly lower termite damage on maize compared to unmulched plots. The reduction in compound index during this season in mulched plots (compared to unmulched plots) ranged between 6.5 and 100%. Reduction in percentage of plants with cut stems ranged between 21.6% and 100% . For both parameters, however, only the 30 kg and 40 kg mulch per plot caused significantly (P < 0.01) lower termite attack on maize plants than the lower levels of mulch coverage.
In both cropping seasons, no significant (P = 0.05) differences in scores of root damage were observed between unmulched plots and plots with 10 kg and 20 kg amount of mulch. Root damage scores in plots with 10 kg and 20 kg mulch levels were, however, significantly higher than scores in plots with 30 kg and 40 kg mulch cover. Root damage in the 40 kg mulch level plots in the two cropping seasons was 77% and 53% lower than in the unmulched plots, respectively. There were no significant differences between plant lodging among treatments in the first cropping season. In the second cropping season, however, percentages of lodged maize plants were significantly (P=0.05) lower in plots with 30 kg, and 40 kg mulch cover than in the other treatments. A significant negative relationship between compound damage index and amount of mulch was observed in the two cropping seasons. The relationship between quantity of mulch and compound damage in the first cropping season was given by 36.85 0.1804x (r = 0.69) and by 15.95 0.0265x (r = 0.95) in the second cropping season.
Below-ground termite infestation of maize. The main termite species infesting maize were Microtermes for most of the season with minor incidences of species from other genera in the post silking stages of the maize crop. Mean rates of infestation on roots of maize under different mulch treatments were relatively low and ranged from 16.5 to 23.5% and 8.3 to 17.8% in 1998 and 1999 cropping seasons, respectively (Fig 1). The differences were, however, not statistically significant.
Effect of mulching maize on incidence of predatory ants. Ant species encountered in the experimental plots at Namulonge were mainly Myrmicaria sp., and Lepisiota sp. Mean incidence of predatory ants at different crop growth stages in mulched and unmulched maize plots over the two cropping seasons are presented in Figure 2. In the first cropping season, percentage of plants with nesting ants on the first two sampling occasions at 28 and 42 days after crop emergence (DAE) was low but higher in mulched than in unmulched plots (Fig. 2). The effects of mulching during this season began to show after 42 DAE. Eight weeks after planting (56 DAE), mulched plots had comparatively higher numbers of maize plants with ant nests compared to unmulched plots (Fig. 2). On subsequent sampling dates at 70, 84, 98 and 102 DAE, wide differences were observed among the levels of mulch. On most of sampling dates, the percentage of plants with nesting ants in plots with 10 kg per plot mulch cover was closer to that in unmulched plots. On the other hand, a higher percentage of plants with ant nests were consistently recorded in plots with higher quantities of mulch (20, 30 and 40 kg per plot) than in plots with 10 kg or without mulch cover although the differences between the 30 and 40 kg treatments was not that great (Fig. 2).
No clear distinctions were observed between the 20 kg and 30 kg levels of mulch cover in percentage plants with nesting ants over the four sample dates. A sharp decline in ant nesting was observed between 84 DAE at which peak ant nesting was recorded and 98 DAE but the trend among treatments maintained the same pattern. (Fig. 2)
Despite the missing data at 42 and 102 DAE, during the second cropping season, a similar trend in ant nesting near maize plants was observed. Overall, however, the nesting of ants was lower compared to the previous season (Fig. 2). Peak incidence of ant nests in maize plots during the second cropping season was recorded at 98 DAE and plots with higher quantities of mulch had significantly (P < 0.05) higher numbers of plants with ant nests.
Analysis of variance of the percentage of plants with ant nests at peak incidence indicated highly significant (P<0.01) differences among the treatments (Table 2). Ant nests were significantly more abundant in plots with 20 kg, 30 kg and 40 kg of mulch per plot than in plots with 10 kg or without mulch (Table 2). In both seasons, however, the peak number of plants with ant nests was not significantly different between plots with 30 kg and 40 kg of mulch. A highly significant positive correlation was obtained between the amount of mulch and percentage of plants with ant nests. The relationship was defined by ant count (y) = 0.0174x 0.05 (r2 = 0.8723) and y = 0.0283x 0.133 (r2 = 0.911) in the 1998 and 1999 cropping seasons, respectively. A highly significant negative relationship was also obtained between percentage ant nesting and damage parameters of termites on maize. The relationship between percentage of damaged plants and ant count was expressed as y = 16.87 0.361x (r = 0.88) and y = 23.33 0.426x (r = 0.92) in the two cropping seasons, respectively.
Maize grain yield. Maize grain yields in the two cropping seasons are presented in Figure 3. In the first cropping season, maize grain yield in unmulched plots was not significantly lower than yield in plots mulched with 10 and 20 kg per plot. Plots with 30 kg and 40 kg per plot of mulch yielded significantly (P = 0.05) better than the other treatments but did not differ significantly between themselves (Fig. 3). An almost similar pattern was obtained in yields during the second cropping season but no significant differences were obtained (Fig. 3). The yield gain (above yields from unmulched plots) caused by the 30 kg and 40 kg mulch cover during the first cropping season were 34.4% and 35.7%, respectively.
In the present study, provision of maize stover as mulches above 4.7 t. ha-1 significantly reduced Macrotermes and Pseudacanthotermes (Macrotermitinae) attack on maize. Coaton (1950) also observed that mulching crops with various items such as hay, manure, wood shavings or threshed maize cobs dramatically reduced attack by termites. Wardell (1987) makes a similar observations in forestry and plantations systems and recommended leaving as much debris as possible on the site after clearing, to ring weed rather than clean-weed young stands and to use dried out stoloniferous and rhizomatous material from planting holes as mulch.
Although Microtermes species were more abundant in mulched than unmulched plots, the differences were not significant. By contrast, Macrotermes and Pseudacanthotermes species were significantly less abundant in mulched than in unmulched plots. Two factors are presumed to be responsible for the observed responses: (1) the higher concentration of food resource for the termites, and (2) enhanced predation by ants within the microhabitat created by mulches. In Table 1 the level of termite attack on maize was shown to be lower in plots with higher quantities of mulch. The significant negative relationship between proportions of damaged plants and amount of mulch indicated that the two were inversely related. This supports the long established view that leaving crop residues in the field or adding further organic matter will provide alternative food to which the termites will be attracted thereby reducing levels of attack (Logan et al., 1990). Further support for this view is provided by the results in Table 1 which indicate no or slight differences in termite attack to plants between plots without mulch and plots with low levels of mulch.
The reduced termite damage on the other hand, could have been a result of enhanced predation of termites by ants beneath mulch. The results in Figure 2 indicated a steady progression of predatory ant activity across the season with significantly higher ant abundance in mulched plots compared to unmulched plots at peak incidence. The significant positive relationship obtained between quantity of mulch and mean number of plants with ant nests provided clear support for increased predation while that between quantity of mulch and termite attack indicated the positive effect of predation on termites. While it is also possible that both improved food sources for termites and enhanced predation termites accounted for our observations in this experiment, it is difficult to distinguish the effects of the two mechanisms.
Our results on termite predation by ants in mulched plots provide evidence on the importance of natural control by the local predators Lepisiota sp. and Myrmicaria species. The results further suggest that the benefits of the enemies hypothesis (Root, 1973) widely acknowledged in mixed cropping (Altieri and Letourneau, 1984; Andow, 1991) could be applied to termites in maize monoculture through mulching.
The higher abundance of ant predators in mulched plots observed in this study could be attributed to an improved micro-climate for the ants under mulch, presumably associated with increased humidity and shade near the soil. On the other hand, it could be the higher numbers of termite prey under mulch, which attracted ants in mulched plots. It is highly possible that interplay of the two mechanisms occurred although on basis of the present data, a distinction between them cannot be clearly demonstrated.
These results suggest that it is the relationship between quantity of mulch available in the field and the duration of mulch material in relation to maturity period of the crop that might be important. On basis of the present results, a theoretical distinction emerges on the effect of non-insecticidal mulches between perennial crops and annual crops. Reasonable proportions of the 30 kg and 40 kg of maize stover used as treatments in both cropping seasons of this study, were still available in the respective plots at crop maturity. This suggested that the maize plants in these plots were constantly mulched all through the stages susceptible to termite attack.
Reports of maize residue effects as reservoirs of lepidopteran stem borer larvae have been published (Macharia, 1989; Saxena et al., 1989). On the other hand, a range of benefits of applying organic material as mulch to the soil has also been reported (Webster and Wilson, 1966)). They include the reduction of soil erosion hazards, better infiltration of rain water and less evaporation, lower soil temperatures, supply of organic matter and nutrients, higher biological activity, better root growth and suppression of weeds (Schroth et al., 1992). Adetola et al. (1995) also reported higher densities of detrivore and phytophage micro arthropods in maize plots mulched with stover compared to those in bare fallow and unmulched control plots. Undoubtedly, a combination of a range of factors could have been responsible for the significantly greater yields of maize in plots with high amounts of mulch in the first cropping season. In recent studies involving ant baits, Sekamatte et al. (unpubl.), recorded significant increase in maize yields due to improvement in termite predation by Lepisiota and Myrmicaria species. The results suggested the important role of predators in reducing maize yield loss by termites.
Under the smallholder conditions in Uganda, maize stover is available and, considering the effort needed to mulch the relatively small hectarage compared to the benefits (mulching) could be viewed as a justifiable practice for subsistence farmers. However, until more critical experiments are conducted in which the effect of mulching is isolated from effects due to predation and also to altered soil environments such as moisture and nutrients, the importance of mulch as alternative food for termites remains anecdotal.
We would like to thank our colleagues Jimmy Akono, James Kayongo and Solomon Kaboyo for their assistance with the field work. This publication is an output from a research project funded by the United Kingdom Department of International Development (DFID) for the benefit of developing countries. Project R6653, Crop Protection Research Programme. The views expressed are not necessarily those of DFID.
The following images related to this document are available:
Line drawing images[cs01023b.gif] [cs01023a.gif] [cs01023c.gif]