Phenology and behaviour of the clearwing moth, Monopetalotaxis candescens (Lepidoptera: Sesiidae), associated with cultivated rooibos, Aspalathus linearis (Fabaceae), in South Africa 1 2 3‡ J.L. Hatting *, J. Brand & M.R. Damavandian 1
Agricultural Research Council-Small Grain Institute, Private Bag X29, Bethlehem, 9700 South Africa Rooibos Ltd., P.O. Box 64, Clanwilliam, 8135 South Africa 3 Department of Conservation Ecology and Entomology, University of Stellenbosch, Private Bag X1, Matieland, 7602 South Africa 2
Production of ‘rooibos’ or red bush tea from Aspalathus linearis (Burman f.) Dahlgren (Fabaceae) is unique within the greater Cederberg region, Western Cape Province, South Africa. One of the principal insect pests associated with A. linearis production is the root-boring clearwing moth, Monopetalotaxis candescens Felder & Felder (Lepidoptera: Sesiidae). Annual moth flights peak during November/December, resulting in young plantations being targeted during the hot, dry summer conditions prevailing at this time. Average infestation levels progressively increased from 28 ± 2 % in year 1, to 53 ± 2 % and 72 ± 1 % in years 2 and 3, respectively. A second, 4–6 years old plantation during the three-year survey period was found to be severely infested (average 92 ± 1 %) with virtually no additional infestation over time. Estimated losses to borer infestation were calculated at 4, 24, 27 and 36 % in plantations that were 1–4 years old, respectively. Significantly more eggs were laid on leaves (66 ± 4 %) compared to woody tissue, stems and twigs. Most eggs hatched between 02:00 and 04:00, with neonates migrating down to the stem base and entering the root just below the soil surface. A control strategy, exploiting aspects of neonate migratory and feeding behaviour, should result in both yield increases and prolonged survival of plantations beyond the typical 4–5-year cycle. Key words: agricultural insect pests, red bush tea, borer, infestation, natural prevalence, pest behaviour, phenology.
INTRODUCTION Production of ‘rooibos’ or red bush, Aspalathus linearis (Burman f.) Dahlgren (Fabaceae), is unique within the greater Cederberg region in the Western Cape Province of South Africa (Van Rooyen et al. 1999). The needle-like leaves are harvested and oxidized before being steeped in hot water to produce a herbal tea (Cheney & Scholtz 1963; Morton 1983) with a characteristic reddish-brown or amber colour, hence the name ‘red bush’ tea. Commercial production commenced in the early 1930s and increased from 17 500 ha in 1981 (Stassen 1987) to a current 84 000 ha (Rooibos Ltd., unpubl.). Although several insect species are associated with A. linearis in its region of origin (Annecke & Moran 1982; Rust & Myburg 1989; Stals 1997; Picker et al. 2003), the three principal pest species are a clearwing moth, Monopetalotaxis candescens Felder & Felder (Lepidoptera: Sesiidae) ‡
Present address: Department of Plant Protection, P.O. Box 578, Sari Agricultural Sciences and Natural Resources University, Sari, Iran.
*To whom correspondence should be addressed. E-mail:
[email protected]
(Fig.1A–D), a leafhopper, Molopopterus theae Theron (Hemiptera: Cicadellidae) and a looper, Isturgia exerraria Prout (Lepidoptera: Geometridae). Research was initiated by the South African Agricultural Research Council (ARC) on the development of an integrated pest management (IPM) programme against these pests, with special emphasis on the root-boring sesiid, M. candescens. Formulation of an IPM approach, based on the judicious use of chemical insecticides, called for an understanding of the phenology and behaviour of this cryptic pest, the pest status of which is largely unknown due to its subterranean habitat. The purpose of this study, therefore, was to investigate: (1) the natural prevalence of M. candescens in A. linearis plantations of different ages over time; (2) the potential impact of M. candescens in terms of plants lost per hectare; and (3) oviposition locality and larval behaviour following egg-hatch and subsequent entry into the host root system, the hypothesis being that most eggs are deposited African Entomology 19(1): 1–10 (2011)
2
African Entomology Vol. 19, No. 1, 2011
Fig. 1. A, Monopetalotaxis candescens (female, yellow form); B, male; C, female (black form); D, larva in an infested root of one-year-old Aspalathus linearis.
on the stem base near the root zone to facilitate entry and subsequent survival of larvae. The window period from egg hatch to physical entry into the root system represents the most likely opportunity for the control of the clearwing larval stage. Effective control could be accomplished if the cryptic larval stage is denied entry into the plant or eliminated soon thereafter. None of these aspects have previously been studied and this information is crucial in developing a control strategy (Hatting, in prep.) with limited environmental impact on the sensitive Fynbos biome in which A. linearis is typically cultivated. This paper provides a summary of several years of research on the biology of M. candescens.
MATERIAL AND METHODS
Natural prevalence Surveys were conducted on the farm Ysterfontein situated c. 8 km west of Clanwilliam. Two plantations of different age were surveyed, namely plantation A (32°08’S 18°48’E; seedlings established autumn/winter 1999) and plantation B (32°09’S 18°48’E; seedlings established autumn/ winter 1996). Surveys were conducted once a month within each plantation from April to December 2000, February to December 2001, February to November 2002 (plantation A) and February to September 2002 (plantation B; ploughed out during October 2002 due to age).
Hatting et al.: The clearwing moth Monopetalotaxis candescens associated with cultivated rooibos
During each survey, c. 100 plants were selected within crisscross transects with one plant uprooted every 3 m (transect lines of c. 30 m; 10 plants selected per line). The taproot of each plant (dead or alive) was split in two up to the base of the stem (c. 50 mm above the soil surface) using a sharp knife and recorded as either infested (i.e. presence of insect stages and/or physical borer damage) or uninfested. Special attention was directed towards the presence of pupae as indicative of pending moth eclosion and flights. Borer larvae present in the root and/or stem base were retained for identification to family level (Scholtz & Holm 1985; Solomon 1995).
Data analysis Total yearly proportions (binomial) were compared by means of the normal deviate Z derived from the normal approximation to the binomial and the progression of infestation was subjected to pairwise comparisons (Snedecor & Cochran 1967). Potential loss of plants to borer infestation Four plantations of different age were surveyed during June 2000 on the farm Blomfontein (31°58’S 19°06’E), situated c. 30 km northeast of Clanwilliam. These plantations were established during the autumn/winter of 1996 (plantation A), 1997 (B), 1998 (C) and 1999 (D). Within each plantation, c. 100 live and 100 dead plants were selected at random (at least 3 m apart) and uprooted, totalling c. 800 plants. Root infestation was determined as described above. Calculation of the percentage loss of plants to borer infestation was based on the assumption that dead plants with borer infestation died as a result of such infestation. The percentage loss was calculated for each of the four age-related plantations using the formula: S = [d/t × 100] × [f/d], where S = estimated percentage of plants lost to borer infestation, t = total number of plants (dead and alive) in the random sample, d = number of dead plants in the random sample, and f = number of dead plants infested or damaged by borers.
Neonate behaviour During 6–12 December 2004, studies on neonate host-entry behaviour were conducted onsite at the farm Rondefontein (31E49’S 18E43’E), c. 43 km northwest of Clanwilliam. Ten 6-month-old plants were extracted from a plantation and transplanted
3
into 3 l plastic pots. These plants were maintained under prevailing ambient conditions for later recording of neonate entry into the root system. In addition, 50 eggs were randomly collected from plants and maintained as a separate batch under similar conditions to facilitate monitoring of egg-hatch (i.e. duration and time of day/night). Eggs were closely monitored at hourly intervals using a Nikon SMZ-2T stereo microscope and newly hatched larvae were photographed with a Nikon Coolpix E4500 digital camera fitted to this microscope. Newly hatched larvae were individually transferred onto leaves, twigs or main stems to simulate the natural position in which eggs were noted during collection. The migratory behaviour and route from oviposition site to host root was captured visually using a Samsung VP-A12 home cam-recorder.
Oviposition sites and impact on host entry by neonates During December 2005, a visual inspection of young plants was performed at three sites, namely Rondefontein, De Hangen (32°01’S 18°55’E; 17 km northwest of Clanwilliam), and Ysterfontein. During these inspections, plants were selected at random and the position of any eggs relative to leaves, stems or twigs were noted. Also, their height above ground was recorded as either bottom, middle or top. At the Rondefontein and De Hangen sites, a total of 30 and 49 egg-harbouring plants were characterized, respectively. In addition to the number and position of eggs encountered at Ysterfontein, the actual position of eggs deposited on woody tissue was differentiated as either main stem or twig. On 7 December 2005, a total of 53 egg-harbouring plants were characterized and tagged at this site and these were destructively sampled on 13 March 2006 to quantify the percentage infestation relative to the number and position of eggs. The frequency at which eggs were laid on either leaf or woody tissue at these sites was compared using Chi-square tests at the 5 % test level (FREQ.EXE Version 2.1; Van Ark 1981). RESULTS Natural prevalence Borer presence or damage was evident throughout all months sampled in both plantations (Figs 2, 3) with >90 % infestation recorded among plants in the older (B) plantation. The younger (A)
4
African Entomology Vol. 19, No. 1, 2011
Fig. 2. Monthly infestation of Monopetalotaxis candescens recorded among live Aspalathus linearis plants in 1–3-year-old plantations at Ysterfontein.
plantation suffered a lower level of infestation, although infestation increased significantly (P < 0.05) from 28 to 72 % over the three-year period (Table 1). The plant material inspected comprised two possible areas of infestation, namely root and/or stem base. Inclusion of the stem base inadvertently included woody tissue typically frequented also by coleopteran borers such as cerambycids and buprestids. However, based on the actual number of family-specific borers identified from a sample of 1269 larvae, the prevalence ratio of Sesiidae:Cerambycidae:Buprestidae was 16:3:1. This 80 % prevalence underscores the primary pest status of sesiid borers in the root system of A. linearis.
Sesiid pupae (Fig. 4A) were noted from October to early December (data not shown). By feeding upwards and laterally into the base of the stem, the final instar larva created an emergence tunnel protruding as a unique chimney-like structure or tube from the side of the stem base (Fig. 4B) through which the pupa later appeared (Fig. 4C). Some larvae emerged through subterranean tissue resulting in an extended ‘chimney’ protruding several centimetres from the soil surface directly adjacent to the stem (Fig. 4D). These ‘chimneys’ were mainly constructed of silk, frass and finely chewed woody material, appearing rigid and adequately protective. Directly before eclosion, pupae exposed themselves by pushing through
Table 1. Percentage infestation of live Aspalathus linearis plants by Monopetalotaxis candescens in 1–6-year-old plantations at Ysterfontein. Age (years) 1 2 3 4 5 6
n*
% Infestation (mean ± S.E.)
Z (age compared)
P
885 1087 942 757 793 477
28.0 ± 1.51 52.6 ± 1.51 71.7 ± 1.47 91.3 ± 1.02 90.9 ± 1.02 92.7 ± 1.19
– –11.50 (1yr vs 2yr) –902.25 (2yr vs 3yr) –10.97 (3yr vs 4yr) 0.37 (4yr vs 5yr) –1.13 (5yr vs 6yr)
–
