+Bioline International Official Site (site up-dated regularly)

The sweetpotato butterfly, Acraea acerata Hew. (Lepidoptera: Nymphalidae) is known to extensively defoliate the sweetpotato crop, especially during the dry season, and this leads to significant foliage yield reductions. However, storage root yield losses due to the defoliation have not been adequately quantified. Artificial defoliation studies were undertaken at Namulonge in central Uganda, to estimate the effect of frequency and timing of defoliation on the agronomic performance of sweetpotato (Ipomoea batatus). Results showed that defoliation had no effect on shoot survival. Single quality defoliations had little effect on storage root yield but repeated defoliations significantly (P<0.05) reduced root yield. The time at which defoliation was carried out did not have a significant effect on the yield. It is possible that the plants had sufficient time to recover when defoliated once and therefore were able to channel enough assimilates towards the roots. However, repeated defoliation may not have given the plants enough time after each regeneration, given the limited growth period of six months. Any recommendation for the control of A. acerata should therefore take into account these responses given the fact that defoliation by the pest is usually less severe than was the case in this study.

Key Words: Acraea acerata, Ipomoea batatus, shoot survival, sweetpotato butterfly

Les papillons de la potate douce (Acraea acerata Hew.) (Lepidoptera: Nymphalidae) défolient considérablement la culture de la patate douce et particulièrement durant la saison sèche et ceci cause une réduction importante du rendement foliaire. Cependant les pertes des rendements des tubercules dû au défoliage n'ont pas été déterminés de façon adéquate. Des études de défoliation artificielles ont été enterprises à Namulonge, au centre de l'Uganda, pour estimer des effets de fréquences et des temps de la défoliation sur la performance de la patate douce (Ipomoea batatus). Les résultats ont montré que la défoliation n'a pas d'effet sur la survie des pousses. Une simple défoliation a un petit effet sur le rendement en tubercules mais des défoliations répétées ont réduit significativement (P<0.05) le rendement des tubercules. Le temps auquel la défoliation a été effectuée n'a pas eu d'effet significatif sur le rendement des tubercules. Il était possible que les plantes avaient eu suffisament de temps pour recuperrer une fois défoliées et ainsi étaient capables de transmetre assez de substances assimilées vers les tubercules. Cependant les défoliations répétées semblaient ne pas avoir donné assez de temps aux plantes après chaque régéneration, vue la période de croissance limitée à 6 mois. Une recommendation de contrôle d'A. acerata devrait donc prendre en considération ces réponses vu le fait que la défoliation par cet insecte est normalement moins sévère par rapport au cas de cette étude.

MotsClés: Acraea acerata, Ipomoea batatus, la survie de pousse, les papillons de la patate douce

The sweetpotato butterfly (Acraea acerata Hew.) is an important pest of sweetpotato in Uganda and other parts of East Africa (Bashaasha et al., 1993; Smit et al., 1996). The pest is found throughout the country, but is thought to be most serious in highland areas (1800 - 25000 meters above sea level). The pest is known to extensively defoliate the sweetpotato crop through feeding by caterpillars, especially during the dry season, and this is thought to lead to great reduction in both tuber and foliage yields (Hill, 1983; Skoglund and Smit, 1994; Smit et al., 1996). However, storage root yield losses due to the defoliation have never been adequately quantified (Lenné, 1991).

The premise that leaf removal by animals reduces plant productivity is based on the assumption that the photosynthetic area removed by herbivoures directly reduces plant growth. Unfortunately, the relationship between leaf removal and plant growth is not nearly as simple and straightforward as the aforementioned statement might simply (Hare, 1980). For instance, the ratio of potato yield to leaf area reduction is usually less than unity (Turnipseed, 1973), and defoliation of annual crops apparently must exceed a threshold, usually between 5 and 30%, before primary production is impaired (Mattson and Addy, 1975). It is also known that the sensitivity of plant species changes with some function of plant development, but very few studies have tried to relate the plant changes to the abundance patterns of the natural defoliators in order to determine their effect on plant growth and reproduction (Hare, 1980).

Making these determinations directly present many technical problems in terms of obtaining enough herbivoures and manipulating them to cause the desired levels of defoliation at the desired times. Thus, in most investigations, mechanical defoliation has been substituted for herbivorous defoliation. However, herbivores can be less debilitating than mechanical defoliators (Hare, 1980). Most herbivores, for example, selectively feed on plant tissues, some of which may be less "valuable" than others, whereas mechanical defoliation tends to be more nonselective (Jameson, 1963).

Nevertheless, to be able to establish a reasonable understanding of the impact of insect defoliation, artificial defoliation at specific levels are necessary. In the present study, therefore, the effect of frequency and timing of complete defoliation on the agronomic performance of sweetpotato was evaluated. This was done to mimic the damage by the sweetpotato butterfly.

Field experiments were conducted at Namulonge Agricultural and Animal Production Research Institute located 27 km north of Kampala at latitude 0°31'E and an altitude of 1,200 - 1,300 m above sea level. The soils of the area are mostly deep tropical red clay loamy types, characteristic of the lower slopes of the Buganda catena (Yost and Eswaran, 1990). They are heavy but well drained, with pH 5.0 - 6.0. Organic matter levels are 2 - 3% in the surface horizons. The climate of the region is wet tropical, with moderate temperatures whose maximum occasionally exceed 28.0°C while the minimum may fall below 15.0°C. Mean annual rainfall is about 1270 mm. The rainfall is bimodal, with a long wet season from late March to July and a shorter one from late September to January (Anon., 1990).

The experiments were established during the second (short) rainy seasons of 1994, and first (long) rains of 1995. In each season, sweetpotato was planted on medium sized mounds of about 0.9 m diameter in three blocks. Three vine cuttings were planted per mound, giving a plant population of approximately 33,000 plants ha^-1. Three weedings using a hand hoe were carried out, and no fertiliser was applied. Monitoring of sweetpotato butterfly infestation was regularly done and foliar sprays of ambush applied as soon as larvae were encountered.

The experiment was a split-split plot design with varieties as the main plot factor, defoliation frequency as the sub-plot factor and defoliation time as the sub-sub-plot factor. The sub-sub-plots measured 5 x 6 m and consisted of 30 mounds. The treatments were replicated three times.

Three sweetpotato varieties, Tororo 3, No. 39 and New Kawogo, were used in the study. Planting dates were 22 October 1994 and 6 April 1995 for the first and second seasons, respectively. Four defoliation frequencies were applied to each variety; zero (control), one, two and three repeated defoliations. The single defoliated sub-sub-plots were randomly defoliated at 1, 2 or 3 months after planting (MAP) while the double defoliated sub-sub-plots were defoliated at 1 and 2; 1 and 3 or 2 and 3 MAP. Defoliation was artificially done by removing all the leaves on a plant at their bases, leaving petioles.

Data collection was undertaken 6 months after planting, specific dates being 26 - 27 April and 28 - 29 September, 1995, for the first and second seasons, respectively. At harvest, the number of surviving shoots were counted from the middle 12 mounds for each treatment combination. After shoot removal, the mounds were carefully dug up and the total number of storage roots and their fresh weight recorded.

For purposes of comparison between control and defoliation combinations, data were analysed as a two factor experiment whereby variety had three levels and defoliation combination had eight levels (1- 8). The combinations were 1 = control, 2 = single defoliation at one MAP, 3 = single defoliation at two MAP, 4 = single defoliation at three MAP, 5 = two defoliation at one at 1 and 2 MAP, 6 = two defoliations at 1 and 3 MAP, 7 = two defoliations at 2 and 3 MAP, and 8 = three defoliations at 1, 2 and 3 MAP. Means were separated using Tukey's test.

Shoot survival. The number of shoots per mound did not vary significantly (P≤0.05) among varieties and with defoliation. Their interaction effect was also not significant. Three shoots per mound were registered at harvest for almost all the sweetpotato mounds sampled.

Number and fresh weight of storage roots. The mean number and fresh weight of storage roots per hectare did not vary significantly among the varieties grown. In both seasons, only the effect of frequency and timing of complete defoliation significantly (P≤0.05) influenced the number and fresh weight of storage roots of the three sweetpotato varieties grown.

Mean number of harvestable roots was lower in the second rains of 1994 than in the first rains of 1995 (Table 1). In each season, however, the mean number of storage roots per hectare decreased significantly (P≤0.05) with increase in defoliation frequency.

In the short rainy season (2nd rains) of 1994, single complete defoliation of sweetpotato 1 - 3 months after planting (MAP) did not significantly affect the number of storage roots harvested compared to the control. However, repeated defoliations suppressed the number of storage roots to levels significantly (P≤0.05) lower than those of the single defoliations and control. Frequency and timing of the repeat defoliation had no significant effect on storage root count (Table 1).

In the long rains (1st season) of 1995, all levels of complete defoliation significantly (P≤0.05) suppressed tuberisation of the sweetpotato compared to the control (Table 1), and statistically similar effects were observed for both single and double defoliations. Complete defoliation of the sweetpotato three times caused the greatest suppression of tuber formation, although its effect did not differ significantly from the effect of double defoliation.

Compared to the undefoliated sweetpotato (control), the decline in sweetpotato tuberisation following complete defoliation ranged from 6% for single defoliation at 3 MAP to 41% for triple defoliation at 1 - 3 MAP during the second season of 1994 (Table 1). During the first season of 1995, the corresponding decline in harvested sweetpotato storage roots were 26% for single defoliation at 2 MAP and 60% for triple defoliations.

Mean fresh weight of storage roots also varied in relation to frequency and timing of complete defoliation of sweetpotato (Table 2). Storage root weight was suppressed by defoliation, the effect being greater at higher defoliation frequencies and during the first season of 1995. In the second season of 1995, single defoliations did not significantly (P≥0.05) suppress sweetpotato storage root weight compared to the control. Double and triple defoliations all significantly (P≤0.05) suppressed storage root weight at harvest compared to the control, although the effect of double defoliation at 1 and 2 MAP did not differ significantly (P≥0.05) from the effects of single defoliation at 2 MAP (Table 2).

In the first season of 1995, all defoliation frequencies significantly (P≤0.05) suppressed sweetpotato storage root weight compared to the control (Table 2). However, storage root weights among sweetpotato subjected to single or multiple defoliations at different times did not differ significantly from each other (Table 2). Except for double complete defoliation at 1 and 2, and 2 and 3 MAP, multiple defoliations suppressed storage root yield significantly (P≤0.05) more than single defoliations.

The reductions in sweetpotato storage root fresh weight per plant over the control ranged from 17% for single defoliation at 1 MAP to 55% for triple defoliations during the second season of 1994 (Table 2). The corresponding decline in yield during the first season of 1995 were 33% for sweetpotato defoliated at 2 MAP to 69% for triple defoliations at 1, 2 and 3 MAP.

In this study, defoliation had no significant effect on sweetpotato shoot survival. Single defoliation did not significantly affect storage root yield in the first season experiment but repeated defoliation significantly (P≤0.05) reduced root yield in both seasons. The growth stage at which defoliation was carried out did not significantly affect root yield.

As evident from the data collected, there was significant effect on root yield when sweetpotato plants suffered a single defoliation during the growing season. These results suggest that, in situations of single defoliation, the sweetpotato varieties had sufficient time to channel their assimilates and/or mobilised reserves towards the regeneration of shoots after defoliation, and therefore recovered in time to channel enough assimilates towards the roots (Welter, 1993). As observed by Chalfant et al. (1990), many foliar feeders cause limited yield reductions in sweetpotato because of the compensatory ability of the plants. The crop therefore tolerates high levels of defoliation.

The study also showed that storage root yield significantly (P≤0.05) decreased with increase in defoliation frequency. It is possible that, following each defoliation, the plants directed their assimilates and/or mobilised reserves towards the regeneration of leaves (Evans, 1984; Welter, 1993). However, repeated defoliation may not give the plants enough time to channel enough assimilates to the roots after each regeneration, given the limited growth period of six months. Hare (1980) noted that while young plants certainly have sufficient reserves and potential to compensate for some insect damage, there is obviously some level of repeated damage for which plants may not compensate.

The present study also showed that the time at which defoliation was carried out did not have a significant effect on the formation of storage roots. This was also attributed to the compensatory ability of the crop or an excessive assimilate capacity. It is suggested that the time interval (one month) between defoliations, in the case of repeated defoliation, was sufficiently large to allow the plants recover.

Page et al. (1980) made attempts to study the effect of defoliation and consequent crop loss in cassava caused by the grasshopper, Zonocerus variegatus (L.) in Southern Nigeria. They observed that, although damage to cassava caused by Z. variegatus may occur every year in the dry season, it only periodically produced losses in root yield depending on it's timing in relation to natural leaf regeneration at the end of the season. Although large populations of Zonocerus are more likely to produce such a loss, a small population may do so only under the right conditions (Page et al., 1980).

During the study period, more rainfall was received at the time of defoliation in the first season than in the second season. The higher rainfall could have encouraged faster and greater recovery of defoliated sweetpotato plants and a more uniform response of sweetpotato to defoliation. This may partly explain why the effects of timing and frequency of defoliation were greater during the second season compared to the first season (Tables 1 and 2). Subsequent experiments could be planted at the end of the rainy seasons (early December or late May) to study the effect of defoliation during the dry season.

In conclusion, this study showed that single defoliation had little effect on sweetpotato shoot survival and yield, while repeated defoliation reduced root yield. The sweetpotato growth stage at which defoliation occurred did not also significantly affect root yield. Any recommendation for A. acerata should therefore take into account these responses given in particularly the fact that defoliation by pest is usually less severe than was the case in this study.

Agro-ecological variation was probably another critical factor that determined sweetpotato response to defoliation. Subsequent experiments should be carried out in higher attitude agro-ecologies where the sweetpotato butterfly is considered a serious problem.

This work was conducted at Namulonge Agricultural Research Institute when the first author was a Graduate Student at Makerere University, and the research was financed by the International Potato Centre (CIP). The support of these institutions is greatly acknowledged.

TABLE 1. Effect of frequency and timing of complete defoliation at 1 - 3 months after planting on the mean number of sweetpotato storage roots per ha and their percentage reduction compared to the control at Namulonge, 1994/95

Defoliation frequency	Defoliation time (MAP)	2nd rains, 1994/95		1st rains, 1995
		No. per ha ('000')	% reduction	No. per ha ('000)	% reduction
0	0	42.3a		57.9a
1	1	37.7ab	11	39.6b	32
1	2	37.2ab	12	42.8b	26
1	3	39.6a	6	41.1b	29
2	1,2	28.0bc	34	35.3bc	39
2	1,3	27.5bc	35	36.9bc	36
2	2,3	26.6bc	37	31.4bc	46
3	1,2,3	24.9c	41	23.2c	60

* Means of three sweetpotato varieties; Tororo 3, No. 39 and New Kawogo
MAP: months after planting
a,b,c: Means followed by the same letters in the same column are not significantly different at P<0.05 (Turkey's test)

TABLE 2. Effect of frequency and timing of complete defoliation at 1 - 3 months after planting on the mean fresh weight (t ha^-1) of sweetpotato storage roots at Namulonge, 1994/95*

Defoliation frequency	Defoliation time (MAP)	2nd rains, 1994/95		1st rains, 1995
		Tons ha^-1	% reduction	Tons ha^-1	% reduction
0	0	12.50a		11.77a
1	1	10.32a	17	7.26b	38
1	2	9.42ab	25	7.83b	33
1	3	9.61a	23	7.20bc	39
2	1,2	6.37bc	49	5.77bcd	51
2	1,3	6.07c	51	5.34bcd	55
2	2,3	5.60c	55	5.00cd	58
3	1,2,3	5.63c	55	3.67d	69

* Means of three sweetpotato varieties, Tororo 3, No. 39 and New Kawogo
MAP: months after planting
First and second rains = Fist and second seasons
a,b,c,d: Means followed by the same letters in the same column are not significantly different at P<0.05 (Tukey's test)