Abstract

In tropical America the American palm weevil, Rhynchophorus palmarum is a significant pest of palms through direct attack and as a vector of the red ring nematode ( Rhadinaphelenchus cocophilus ). The objective of the present study was to reduce the incidence of the red ring disease in a commercial oil palm ( Elaeis guineensis ) plantation through mass trapping R. palmarum , using bucket traps containing carbofuran treated sugar-cane and baited with the principal component of its male produced aggregation pheromone, 6- methyl-2(E)-hepten-4-ol. Trapping in a 30 ha plot that was highly infected with red ring disease was effective over several months in lowering weevil population and the rate of disease infection.

During the year in which the mass trapping was conducted over 60,000 weevils were removed from the test site. This number exceeds by one to two orders of magnitude the number estimated to be present by mark-release-recapture experiments (23-57/ha), and suggests that weevils were drawn from adjacent stands into the traps.

Survey trapping over 1,256 ha revealed that the usual practices of removal of old stands by poisoning or bulldozing caused emigration of large number of resident weevils to adjacent plots. Trapping around such areas when these practices are done is likely to lessen the spread of red ring disease to adjacent plots.

Introduction

The American palm weevil, Rhynchophorus palmarum (L) (Wattanapongsiri, 1966) is a major pest of cultivated oil and coconut palm in the tropical Americas. Weevils attack healthy coconut palms but in oil palm mechanical injury or rot usually precede weevil attack (Chinchilla, 1988). Damage is caused by extensive tunneling of larvae in the trunk or the whorl (Griffith, 1987). Although R. palmarum can kill palms through direct attack, its major threat to palm plantations lies in its being the vector of the nematode Rhadinaphelenchus cocophilus (Cobb, 1922; Tidman, 1951; Hagley, 1963, Martyn, 1953; Griffith, 1967 and 1978) which causes red ring disease (RRD). Palm losses due to RRD are commonly between 0.1 to 15% (Chinchilla, 1988). With several hundred thousand hectares of commercial oil and coconut palm plantations in the tropical Americas, it is estimated that yearly losses due to this disease amount to ten of millions dollars.

In oil palm, infection by other mechanisms such as via nematode carrying harvest knives or soil (Fenwick, 1968) are considered insignificant in comparison to inoculation by the weevil (Schuiling and Van Dinther, 1981; Chinchilla, 1988). Red ring symptoms do not become evident until 2-3 months after infection and nematicidal treatments at this point have proved fruitless to salvage oil palms with the classical red ring symptoms. Low percentage of recuperation with systemic nematicide treatment occurs when symptoms are solely of the little leaf type (Chinchilla, 1988).

The most widely recommended strategy to lower the incidence of the RRD is to reduce the population of the palm weevil (Griffith, 1987; Chinchilla, 1988), by removing breeding sites and by trapping adult insects. It has been long known that weevils are attracted to wounds in trunks of several palms including oil palm and coconut. Treatments of host palms (Fenwick, 1967) with insecticide and removal of red ring-diseased trees, coupled with trapping using insecticide-laden palm stem pieces have been considered appropriate phytosanitary practices (Griffith, 1969). The latter technique has been used in the Caribbean since the early 1970's (Griffith, 1987; Morin, 1986) and several designs (Moura et al , 1990 and 1991) have been found to be effective for trapping adult weevils. Present traps in use utilize insecticide treated palm stem (Mariau, 1968; Griffith, 1969; Morin et al , 1986; Chinchilla et al , 1990; Morales y Chichilla, 1990) although some tropical fruits are marginally effective as trap baits.

We recently reported an efficient trap for R. palmarum that utilizes the major component of the male produced aggregation pheromone 6-methyl-2(E)-hepten-4-ol (Rochat et al , 1991) in combination with insecticide laden palm stem or sugar cane (Oehlschlager et al , 1993, in press). The objective of the present study was to examine the efficacy of this trap in the management of the population of  R. palmarum and hence of RRD in a commercial oil palm plantation. This paper reports preliminary results of a planned one year study of the effect on red ring incidence of mass trapping of R. palmarum in approximately 30 ha of a commercial oil palm plantation. In addition, we report the results of coincident survey trapping over 1,256 ha of oil palm.

Materials and Methods

Experiments were conducted in the 6,600 ha Palma Tica oil palm plantation in the Coto river valley situated on the Southern Pacific coast of Costa Rica. Commencing in 1989, surveys for RRD were conducted on a regular basis. Survey crews located red ring cases using the criteria described by Chinchilla (1988) which consisted of identifying those palms with any combination of the classical red ring symptoms or the "little leaf" condition. A large majority of the palms determined to be afflicted with red ring had the symptoms of extensive yellowing and petiole breakage of the lower leaves. The whorl of these palms normally had a compact appearance due to the presence of leaves that were shorter than normal. These little leaves usually had a pale or yellow coloration. Palms with RRD were eliminated by poisoning with a systemic herbicide (MSMA) or felling. Cut surfaces of felled palms were sprayed with an insecticide such as carbofuran.

Traps consisted of white 19 l plastic buckets modified by cutting holes in the bottom and entry slots in the sides and top as described in Oehlschlager et al , 1993, in press. Traps contained a device that released 6-methyl-2(E)-hepten-4-ol at 20 mg per 24 h at 30ºC. Additionally, traps contained 15 pieces of halved sugar cane stalk that had been immersed in a suspension containing 1.5 cc/l carbofuran. Sugar cane was changed and captured weevils counted weekly. Traps were hung on palms at aprox.1.7 m above the ground. This trapping system and the pheromone lure are now available from ASD de Costa Rica (P. O. Box 30, 1000 San José, Costa Rica, ph. (506) 257-2666, Fax (506) 257-2667). Synthetic 6-methyl-2(E)-hepten-4-ol was pre pared as previously described (Oehlschlager et al . 1992) and was of 95% chemical purity.

Experiments

Mass trapping

In September and October 1991 traps were placed in a 30 ha of oil palm (DelixAVROS, Malaysia, planted in 1971) that was diagnosed to have a high incidence of RRD compared to other stands of similar age and origin. The test area was surrounded on three sides by stands of palms of comparable age and on the fourth side by a new plantation (1989 planting). Trap captures were determined weekly during the entire test period. The mean weekly capture per trap (8.65 weevils/trap/week) as well as the standard deviation (7.26) were determined for the initial four month period. Using these values we defined high trap captures as those that were above the mean by more than one half a standard deviation, and low trap captures as those that were below the mean by more than one half a standard deviation.

Survey trapping

In December 1991, 24 traps were placed six each equidistant from each other in three concentric circles of 500, 1000 and 2000 meters over 1,256 hectares of the plantation. Two weeks after commencement of the experiment one additional trap was placed at the center of the smallest circle of traps. There were 16 traps in plots planted between 1983 and 1984, 7 traps in plots planted between 1975 and 1977 and 2 traps in plots planted in 1986. At least 4 km separated any survey trap from the mass trap site. Weevils were collected and sugar cane renewed weekly.

Mark-release-recapture experiments

In an experiment designed to estimate population over a large area and judge the dispersion capabilities of the weevil, 535 freshly field captured weevils were marked and released (November 30, 1991) at the center position of the survey traps. Captured marked and unmarked weevils were counted on days 2, 4, 6, 8, 11, 13 and 15. Thereafter, traps were left in place and treated as survey traps.

Data treatment

Assumption of normality and homogeneity of variance were tested on all data by graphical assessment of log variance versus log mean and Bartlett's test, respectively (SAS Institute, 1985). Some data were transformed by (X + 0.5) 0.5 to eliminate heteroscedasticity (Zar, 1984) and subjected to an analysis of variance using PROC GLM (SAS Institute, 1985). In cases were data did not approach a normal distribution, data were ranked non parametrically using the chi-square approximation method of a Kruskal-Wallis or Wilcoxon test (SAS Institute, 1985). Nonparametric multiple comparisons were separated using the Q test statistic (Zar, 1984).

RESULTS AND DISCUSSION

Beginning in 1989 periodic (3-4 month) plantation surveys for RRD were conducted and these became bimonthly in 1991. In 1990 the most highly infected stand of any age was a 1971 planting of Deli xAVROS, Malaysia covering 39 ha and designated as lot 50. This area was chosen to examine the efficacy of pheromone enhanced trapping as a method to reduce the weevil population and the red ring disease.

Mass trapping experiment

In September and October 1991, 219 traps were placed in the chosen area, and 24 traps were placed in the adjacent stand of newly planted and unsusceptible palms bordering one side of lot 50. Trap captures steadily declined throughout the year, except for the period December 1991 to Feb ruary 1992 (Fig.1).

According to an earlier study conducted (Morales and Chinchilla, 1990) in this plantation, the population of R. palmarum reached a low in April and increased by a factor of 3 to reach a maximum in January and February. As revealed by captures in survey traps placed in the plantation in December 1991, the population of R. palmarum was steady between December and March, declining only thereafter (Fig.2).

During the period September 1991 to July 1992 over 60 000 weevils were captured in the 219 traps located within lot 50. This would correspond to approximately 2000 weevil/ha or about 17 weevils/palm in the site. Given the large size of the weevil population, we consider that a significant portion of the resident population was captured and that additional weevils were drawn from adjacent stands and intercepted during migration. As suggested below we estimated the population of weevils over a large area of the plantation to be in the range of 23-57 weevils/ha. Given that the area in which the mass trapping was conducted is heavily infected with the RRD it is probable that the population is in the high end of this estimate.

Supporting the hypothesis that weevils were being captured from adjacent stands, capture rates for traps occupying peripheral positions were consistently and significantly higher that captures rates for traps occupying internal positions. Indeed, capture rates of peripheral traps varied with their immediate surroundings. Those peripheral traps placed in the 1989 planting bordering lot 50 consistently captured significantly more (2 to 2.5 times) weevils than those traps bordering the 1968 and 1976 plantings. This, even though the latter stands had appreciable RRD and were assumed to harbor a larger weevil populations than the 1989 plantings (Chinchilla et al , 1990). We consider that the higher capture rates in the young planting were due to commercial activity in the plantation (see below Survey trapping) that caused dispersion. When disturbed weevils migrate from areas undergoing harvesting, pruning or palm elimination and finding themselves in an area of young palm, they continue to search for older stands that have more suitable sites for feeding and oviposition. Placement of traps baited with the aggregation pheromone in young stands presents the weevils with the signal of a suitable site in which to feed and mate. By this reasoning traps in such environments are likely to be effective over larger distances than those in more hospitable environments.

To determine if trapping was having a measurable effect on the population of weevils resident within the trapping area we defined areas of low and high capture rate according to the criteria outlined previously. These criteria were applied to traps within lot 50. The initial area of low capture defined in September constituted only 28% of the lot 50 trap area, but this increased to 83% by December 1991. Average captures in the defined low capture areas were consistently between 4 and 5 wee vils /trap /week. Whereas those in the high capture area steadily decreased from 18 weevils /trap /week at the commencement of the experiment (September) to approximately 10 weevils/trap/week in December. Between January and March overall trap captures increased in the test area but the increases were confined to the peripheral traps and those located in the December high capture area. After March 1992 trap captures in the site steadily decreased until June, the average trap capture over the entire site was less than two weevils/trap/week. A noticeable event during the latter half of the test period was the occurrence of pockets of infestation, which would appear month after month as high capture rates (12/week) in specific isolated traps. The trend in capture rates observed over time suggests that a significant portion of weevils resident within the test area were captured, and that periphery traps were effective in preventing migrating weevils from reinfecting the trapping area.

Mark-release-recapture experiments

To gain insight into the dispersal characteristics and obtain estimates of the population of weevils in the plantation, we released 535 marked weevils at the center of the survey trap area. Captured weevils were counted at day 2 and at 2-3 day intervals for two weeks. The initial count taken 48 h after release revealed 53 marked weevils in traps 500 meters from the release point and 13 in each of the sets of traps located 1 and 2 km from the release point. Thus, the average flight of recaptured weevils was about 500 meters/day and a small portion (16%) migrated up to 1 km/day.

Using data from the initial weevil collection (256 unmarked insects collected in the 500 m traps; 364 in 1 km traps and 360 in 2 km traps) taken two days after release it was estimated that within the 500 m radius of traps the population was 3,120 ±1,344 or 23-57 weevils/ha. During the initial week of the experiment a total of 3,274 unmarked weevils were collected from all 25 traps covering 1,256 ha. This represents an average of 2.6 weevils/ha during the period and represents 5-11% of the population estimated by the above calculation. By the end of the second week the captured insects were estimated to represent 7-18% of the total weevil population.

Effect of mass trapping on incidence of RRD

The rate of red ring infection in oil palm plantations has been suggested to vary with seasonal changes in palm weevil populations and with the population of nematode-infected weevils (Hagley, 1963; Blair, 1970a and 1970b; Griffith, 1978; Schuiling and Van Dinther, 1981; Chinchilla et al , 1990; Morales y Chinchilla, 1990).

In the Coto plantation infection by RRD is comparatively low during March and October and reaches a maximum in January and February (Morales and Chinchilla, 1990). It is assumed that the period between infection and exhibition of noticeable symptoms of RRD is two to three months. Disease incidence steadily decreased during the course of the trapping, in a pattern that followed trap captures by 2-3 months. Regression analysis revealed that the trap capture correlated best with RRD within the test site if an incubation period of two months (R²=0.698) or a three months (R²=0.761) was assumed. Trap capture did not correlate well if a one (R²=0.458) or four (R²=0.341) month incubation periods were assumed.

Clearly the most important aspect of the mass trapping experiment deals with the question of the impact of this practice on the rate of new RRD infections. The stand in which the trapping was conducted was surrounding on three sides by stands of comparable age, and on one side by a 1989 planting that was not susceptible to RRD. For the purposes of comparison, we utilized the existing history of RRD infections for the test area and surrounding stands available from surveys for RRD commencing in 1990. As shown in figure 3, disease incidence for the surrounding stands increased steadily between 1990 and 1992 (July). During 1990 and 1991 the experimental area followed RRD incidence trends of the surroundings, but disease incidence dropped in the test area once the experiment was well underway in 1992 (Fig.3). Thus, lot 50 was transformed from a lot with surging RRD (between 1989 and 1991 the rate of new infections increased by a factor of 10) to one in which the rate of new RRD infection was one of the lowest in the plantation. A more definite comparison in RRD behavior between lot 50 and surrounding stands can be done at the end of the year 1992, when the rate of RRD is expected to increase seasonally in the plantation (Morales y Chinchilla, 1990).

A high density of traps (near 6/ha) was used in the mass trapping experiment. As judged from trap captures this density seemed to rapidly capture a sizable fraction of the resident population, while peripheral traps apparently were effective in capturing intruders. We suspect that a lower density of traps would be just as effective. Thus, the number of weevils captured weekly per hectare in the mass trapping experiment was very similar to the average weekly capture in the widely spaced (minimum distance about 523 m between traps) survey traps. It is reasonable to assume that if a density of one or two traps/ha would have been used in the present study that an equivalent result would have been obtained but with higher average trap captures.

Survey trapping

Capture rates for survey traps fluctuated with commercial activity in the plantation. Three noticeable increases in trap captures were observed for the weeks ending January 6, January 29 and April 1992 (Fig.2). The increase noted on the January 6 trap capture coincides with a partial elimination of nearby areas (70 ha) between December 23 and 26, 1991. These areas had been planted in 1968 and alternate rows of palms were treated with a systemic herbicide to reduce the original stand in 50%. Browning of leaves occurred within one week and by the end of the second week browning was complete in the majority of palms. The idea that weevils resident in the treated area were captured in the survey traps, as they dispersed from their dying host is evident from a detailed examination of trap captures as a function of distance from the poisoned palms. Thus, the captures in traps within 1,000 m of the poisoned palms increased proportionally more than captures in traps at distances greater than 3,000 m (Fig.4).

The second and very distinct increase in capture rates of survey traps (January 29, 1992) coincides with elimination of several sections (177 ha) of palms planted also in 1968. These palms were felled by bulldozing January 6-16. Fresh wounds caused during bulldozing would act to maintain weevils in the area, but within 4-5 days wounds would no longer be attractive to weevils. Thus, dispersal of dislocated insects was noted in increased trap captures on January 29. As expected captures in traps within 1,000 m of the elimination site increased proportionally more than captures in traps a distances greater than 2,000 m (Fig.4). The highest number of weevils (420) to be captured during a single week in a single trap was captured during this week in a trap within 1,000 m of the eliminated area.

It is clear that the elimination of old stands either by poisoning or bulldozing causes dispersal of R. palmarum and that dispersing weevils can be captured. This fact was also noticed in another study on population dynamics of R. palmarum carried out in Honduras (Chinchilla et al . 1990). It would appear prudent to place perimeter traps around plots targeted for elimination prior to the disturbing the plots and leave traps in place at least two weeks after the elimination.

Conclusions

This paper reports some preliminary results which can be applied to the development of management technics for American palm weevil populations in an oil palm plantation using a pheromone-based trapping system.

The above experiments suggest that pheromone based trapping can be used to effectively manage R. palmarum populations and hence RRD, in oil palm and probably coconut plantations. In the heavily RRD infested stand investigated, trapping at a density of about 6 traps/ha effectively reduced resident weevil population which was accompanied with a reduction of red ring incidence. Several experiments at different trapping densities suggest that 1-2 traps/ha would achieve similar results. Periphery traps seem to be effective in preventing migrating weevils from reinfecting a stand. Periphery trapping would also be very beneficial when executing elimination of older stands. Indeed, it would seem prudent to conduct trapping within stands destined for elimination prior to commencement of this activity to reduce resident populations before disturbing them. After commencement of the elimination, moving traps to the periphery of this area to additionally contain migrating weevils would seem to be a beneficial practice. We are currently examining these strategies in two commercial oil palm plantations in Costa Rica.

Acknowledgments

The authors would like to thank the Palma Tica company for permission to publish these results. Special thanks to D. L. Richardson for invaluable support to this research; G. Castrillo for coordination of field work; professor L. Jirón of the University of Costa Rica for help in the initial phases of some of the field work; R. S. Mc. Donald of the Department of Biology Sciences at Simon Fraser University for statistical analysis; the Natural Sciences and Engineering Research Council of Canada for Operating and International Collaboration Grants to A.C.O. An administrative leave granted to A.C.O. by Simon Fraser University is also gratefully acknowledged.

References

Blair, G. 1970a. Studies on red ring disease of the coconut palm.Oleagineux 25: 19-22.

Blair, G. 1970b. Studies on red ring disease of the coconut palm.Oleagineux 25: 79-83.

Chinchilla, C. M. 1988. The red ring-little leaf syndrome in oil palm and coconut. Bol. Técnico OPO-CB 2(4): 113-136 and Agricultural Services and Development, Technical Bull. 1,1991, San José, Costa Rica.

Chinchilla, C. M.; Menjivar, R. y Arias, E. 1990. El picudo de la palma y la enfermedad del anillo rojo/hoja pequeña en una plantación comercial de palma aceitera en Honduras. Turrialba,40: 471-477.

Cobb, N. A. 1922. A note on the coconut nema of Panama. J.Parasitology, 9: 44-45.

Fenwick, D. W. 1968. Red ring disease of the coconut palm. In. G. C.Smart, Jr. and V.G.V. Preey (eds) Tropical Nematology. Center for Tropical Agriculture. University Agriculture Press,Gainesville, FL, pp 38-48.

Griffith, R. 1967. Progress on the entomological aspects of red ring disease of coconuts. J. Agric. Soc. Trin. Tob. 67: 209-218.

Griffith, R. 1969. A method of controlling red ring disease of coconuts. J. Agric. Soc. Trin. Tob. 67: 827-845.

Griffith, R. 1978. Epidemiology of red ring disease of coconuts inTrinidad and Tobago. J. Agric. Soc. Trin. Tob. 78: 200-217.

Griffith, R. 1987. Red ring disease of coconut palm. Plant Dis. 71:193-196.

Hagley, E. A. C. 1963. The role of the palm weevil, Rhynchophorus palmarum as a vector of red ring disease of coconuts. 1 Results of preliminary investigations. J. Econ. Entomol. 56:375-380.

Mariau, D. 1968. Mettodes de lutte contre le Rhynchophore.Oleagineux, 23: 443-446.

Martyn, E. B. 1953. Red ring disease of coconuts in Trinidad and Tobago. Tropical Agriculture, Trinidad, 30: 43-53.

Morales, J. L. y Chinchilla, C. Ml. 1990. Estudios poblacionales en Rhynchophorus palmarum y su relación con la enfemedad del anillo rojo/hoja pequeña en palma aceitera en Costa Rica.Turrialba, 40(4): 478-485.

Morin, J. P.; Luchini, F.; Araujo, J. C. A.; Ferreira, J. M. S. and Fraga, L. S. 1986. Rhynchophorus control using traps made from oil palm cubes. Oleagineux, 41: 57-62.

Moura, J. I. L.; Resende, M. L. V.; Sgrillo, R. B.; Nascimenti, L.A. and Romano, R. 1990. Diferentes tipos de armadilhas e iscasno controle de Rhynchophorus palmarum . Agrotropica 2(3): 165-169.

Moura, J. I. L.; Resende, M. L. V.; Lima, M. F. and Santana, D. L.de Q. 1991. Taticas para o controle integrado de Rhynchophorus palmarum . CEPLAC/CEPEC, Bahia, Brazil, 16p.

Oehlschlager, A. C.; Pierce, H. D.; Morgan, B.; Wimalaratne, K.;Slessor, N.; King, G. G. S.; Gries, G.; Gries, R.; Borden, J.H.; Jirón, L.F.; Chinchilla, C. M. and Mexzón R. 1992. Chirality and field testing of rhynchophorol, the aggregation pheromone of the American palm weevil. Naturwissenschaften,79: 134-135.

Oehlschlager, A. C.; Chinchilla, C. M.; González, L. M.; Jirón, L.F.; Mexzón, R. and Morgan, B. 1993. Development of a pheromone-based trapping system for the American palm weevil, Rhynchophorus palmarum . J. Econ. Entomol. (in press).

Rochat, D. C.; Lettere, M.; Ducrot, P. H.; Zagatti, P.; Renou, M.and Descoins, C. 1991. Male-produced aggregation pheromone of the American palm weevil, Rhynchophorus palmarum : collection, identification, electrophysiological activity and laboratorybioassay. J. Chem. Ecol. 17: 2127-2141.

SAS Institute. 1985. SAS User's guide: Statistics. SAS Institute,Cary North Carolina.

Schuiling, M. and Dinther, J. M. B. 1981. Red ring disease in the Paricatuba oil palm estate, Para, Brazil. Z. Angew. Ent. 91:154-169.

Tidman, D. A. 1951. Agricultural and horticultural problems of Brazil. World Crops. 3: 341-344.

Wattanapongsiri, A. A revision of the genera Rhynchophorus and Dynamis (Coleoptera, Curculionidae). Dept. Agric. Sci. Bull., Bangkok. 1: 1-328.

Zar, J. H. 1984. Biostatistical Analysis. 2nd. ed. Prentice-Hall,Englewood Cliffs, N.J.