In Brazil, Spodoptera frugiperda (J. E. Smith) is a common species in corn plantations. It is a pest of great economic importance, with production losses reaching 34% (Carnevalli & Florcovski 1995). It is found in corn from plant emergence up to the ear stage, within which it is often sheltered, and in many cases has become as important a pest as the corn earworm, Helicoverpa zea (Boddie) (Parra et al. 1995).

Morphological and physiological variations occur during the development and maturation of plants, often provoking changes in S. frugiperda egg distribution and egg mass characteristics, such as number of egg layers and scale density. These may be due to nutritional factors that vary as a consequence of physiological changes in the plants. S. frugiperda egg mass distribution on the host plant is influenced by the phenological stage of the crop, and can be concentrated within the lower, middle or upper plant regions and on different parts of the leaves and fruiting structures (Ali et al. 1989, Sifontes et al. 1988, Meneses et al. 1991). Sifontes et al. (1988) found that as the rice plant host grew older, S. frugiperda egg masses had a higher scale density, as well as a larger number of layers and eggs.

A possible control for S. frugiperda is the use of egg parasitoids like Trichogramma. These species, however, have difficulties in parasitizing egg masses of this pest because they are covered in scales and the eggs are deposited in layers (Toonders & Sánchez 1987, Cortez & Trujillo 1994). Thus, optimal use of Trichogramma spp. to control S. frugiperda requires information on egg placement by S. frugiperda during development of the corn crop, especially egg distribution in layers and the presence of scales, as well as the degree of natural parasitism by Trichogramma spp. The aim of this research was to evaluate the egg distribution and the natural parasitism of S. frugiperda eggs at different phenological stages of corn.

Materials and Methods

The trials were conducted at the Areão Farm, in Piracicaba, state of São Paulo, Brazil, at 22°42’00” South latitude, 47°38’00” West longitude and 546 meters altitude. Data were collected from March—July, 1998 and 1999, October—December, 1998, and from February—April, 2000. Each time a 2-hectare field, with approximately 50,000 corn plants per hectare was studied. The crop received conventional cultural practices recommended for corn, except that no insecticides were applied after the first leaves appeared.

Weekly, 250 plants were sampled, randomly distributed among 10 sampling spots, approximately 30 meters apart. In each spot, 25 plants in an “X” distribution were evaluated, with 5 plants in the center and 5 at each end of the “X”. Each group thus included 5 plants, approximately 5 meters from the next group. The plants were evaluated thoroughly and the plant height, the number of leaves per plant and the egg-laying sites were recorded. The eggs were grouped according to the plant region (lower, middle or upper) and leaf surface (abaxial or adaxial) and distribution and natural parasitism were studied at the corn plant phenological stages of 4-6, 8-10 and 12-14 leaves (Cruz & Turpin 1982). The eggs were collected and placed in labeled plastic tubes (3.5 × 1.0 cm), taken to the laboratory, and then kept in glass tubes (8.5 × 2.5 cm). Data were collected on the number of egg layers, the presence or absence of scales and the number of eggs, based on the number of larvae that hatched. Natural parasitism by Trichogramma spp. was determined by counting the number of parasitized eggs. The S. frugiperda eggs were placed in glass tubes (8.5 × 2.5 cm) in the laboratory for the observation of parasitoid emergence.

The results on the distribution and number of eggs found, the number of eggs and egg layers in each egg mass, and the number of parasitized eggs and their distribution at different phenological stages and on different parts of the corn plant were subjected to an analysis of variance with the means compared by the PLSD (Protected Least Significant Difference) test (alpha = 0.05). The treatments were arranged in a factorial scheme with two factors: plant age and plant region; the number of parasitized eggs and percentage S. frugiperda egg parasitism at different phenological stages were compared by the Tukey test (alpha = 0.05) with a completely randomized design, using SAS software for analysis.

Results

The S. frugiperda eggs were found at every phenological stage of the corn plant, and over 99% were found on leaves with only two egg masses found on the stem. A significant interaction was found between plant phenological stage and region (lower, middle and upper) (F = 2.79; P = 0.02) and corn leaf surface (abaxial and adaxial) (F = 7.79; P = 0.009). That is, the egg distribution among the different parts of the plant varied with the age and development of the corn plant.

At the early developmental stages of the crop (4-6 leaves), the majority of oviposition occurred on the 1st and 2nd leaves (60.4%) and on the abaxial surface of the leaves (83.4%). There were significant differences between the upper, medium and lower regions and the abaxial versus the adaxial leaf surface at the same developmental stage and among the various developmental stages (4-6, 8-10 and 12-14 leaves) (Table 1). As the plant developed to the 8-10 leaf stage, the preferred oviposition site changed and the eggs became more concentrated on the middle and upper plant regions (73.5%) and on the adaxial leaf surface (66.9% of the egg masses). At the 12-14-leaf stage, the greatest concentration of egg masses (61.4%) was observed on the middle region of the plant, significantly greater than on the lower and upper regions. However, there was no significant difference between the abaxial and adaxial leaf surfaces in terms of egg distribution. The middle part of the plant was, to a degree, constant as an oviposition site, with no significant difference between the phenological stages (Table 1).

The mean number of egg masses was significantly (F = 4.09; P = 0.01) different among the three developmental stages of the corn, with the highest concentration of egg masses at the early developmental stages (4-6 leaves) until approximately 60 days after planting, when the plants had 8-10 leaves, decreasing at the 12-14 leaf stage. However, the mean number of eggs did not differ among the three stages (F = 2.76; P = 0.01) (Table 2).

A significant interaction was found between plant-age and the number of egg layers in each S. frugiperda egg mass (F = 2.72; P = 0.01). At every phenological stage, most egg masses had more than one layer; the most frequent configuration was three layers, with 91.56% covered by scales. At the 4-6-leaf stage the eggs predominantly had 2 or 3 layers and at the following stages - 8-10 and 12-14 leaves, those with 3 or 4 layers prevailed (Table 2).

With respect to parasitism, no significant interaction was observed between plant phenology and plant region (F = 1.49; P = 0.21), and no significant differences were found among the three developmental stages in the number (F = 0.3; P = 0.74) and percentage (F = 1.37; P = 0.25) of S. frugiperda egg masses parasitized by Trichogramma spp. (Table 3). However, significant differences in parasitism were detected among the three parts of the plant (F = 2.91; P = 0.05); the greatest number of parasitized eggs occurred on the lower and middle regions of the plant (Table 3). Also the number (F = 0.98; P = 0.38) and percentage (F = 1.84; P = 0.17) of parasitized S. frugiperda eggs did not differ among the three phenological stages of the corn (Table 4), showing that the distribution and degree of parasitism by Trichogramma spp. on different regions of the plant were independent of the developmental stage of the crop.

Overall natural parasitism was very low with a maximum of 2.21% of the eggs parasitized (Table 4), despite the fact that S. frugiperda eggs were available at every plant phenological stage. Parasitism remained constant over all densities of eggs observed at the 4-6 and 8-10 leaf stages (Table 2). Few Trichogramma species were found parasitizing the S. frugiperda eggs. Trichogramma pretiosum Riley was most abundant, comprising 93.79%. Trichogramma atopovirilia Oatman & Platner occurred in 2.07% of the samples. Multiparasitism—a frequent phenomenon in Trichogramma—was observed, as T. pretiosum and T. atopovirilia appeared simultaneously in various egg masses. The remaining infested eggs (4.14%) were parasitized by other species of the genus Trichogramma. Only one individual of another genus of the family Trichogrammatidae was found. It was not identified because only one female emerged (the identification is based on males). In addition, this female failed to reproduce in eggs of the factitious host Anagasta kuehniella Zeller.

Discussion

Changes in the oviposition behavior of S. frugiperda occurred as the host plant developed. Egg masses were most abundant on the lower region of the plant, and on the abaxial leaf surface, at the early 4-6-leaf stage; and on the middle and upper regions and the adaxial leaf surface at the subsequent stages (8-10 and 12-14 leaves). Pitre et al. (1983) showed that while in grasses such as sorghum and corn, S. frugiperda egg masses are more numerous and larger than in other hosts such as soybean and cotton. They also showed that the abaxial leaf surface, and the plant regions with greatest leaf mass, are preferred oviposition sites since they constitute a protected environment. The oviposition strategy of S. frugiperda on corn may focus on protection from natural enemies or suitable feeding sites for progeny, since females oviposit in areas near the feeding sites of the larvae. As the plant develops, egg laying shifts to the upper region of the plant, close to the spindle. When this site is not available after bolting, the middle region of the plant, close to the ear, becomes the main oviposition site. Labatte (1993) found that larvae migrate towards the upper region of the corn plant to feed at the 4-6 leaf stage. He indicated that up to the 10-leaf stage feeding occurs preferably among the whorl of young leaves, where the worms remain until the end of their development. After the 12-14-leaf stage, the worms migrate to the tassel and the middle region of the plant, during tasselling and ear formation, remaining in the leaves and ears. This egg laying and larva distribution pattern was also observed for Ostrinia nubilalis (Hüb.) by Shelton et al. (1986) and Labatte (1991), who stated that the phenological stage of the corn is the main factor affecting the pest’s behavior.

The decreased number of egg masses observed in our study as the plant grew older coincides with the population fluctuation of S. frugiperda adults and larvae observed by Mitchell et al. (1984) and Gutiérrez-Martinéz et al. (1989). Surveys conducted by Gutiérrez-Martinéz et al. (1989) showed that the first adults appear right after plant emergence, with the highest frequency from the 10th through the 41st day of emergence, when the plant is more susceptible, decreasing considerably up to the 72nd day, when fewer attacks on the corn crop occur. This phase also coincides with the period of greatest production losses, which may reach 19% when the plant achieves the 8-10 leaf stage (Cruz & Turpin 1982).

Although the number of egg masses decreased with plant age, the number of eggs found on the 8-10 and 12-14 leaf stages was higher than at the 4-6-leaf stage (Table 4). Meneses et al. (1991) also observed a high frequency of egg masses with three layers and an average of 385.6 eggs/egg mass in rice. Sifontes et al. (1988) found S. frugiperda with a higher number of layers and eggs in egg masses, and a higher scale density on older rice plants.

Studies by García & Sifontes (1987), Toonders & Sánchez (1987) and Hoffmann et al. (1995) revealed that Trichogramma spp. have difficulty in parasitizing S. frugiperda eggs, which is in accordance with our results. García & Sifontes (1987) reported that 46.9% of 47 egg masses parasitized by Telenomus and 4.3% parasitized by Trichogramma spp. together with Telenomus. Moreover, they considered only the number of egg masses parasitized and neglected the number of eggs parasitized, which obviously would give an even lower percentage of parasitism. Toonders & Sánchez (1987), on the other hand, counted the number of parasitized eggs, and observed that natural parasitism by Trichogramma spp. varied from 0 to 10% in samples examined in six different fields and when 30,000 parasitoids were released in 1.5 ha of corn the parasitism rate was only 4%.

The differences found in the degree of parasitism observed among the three parts of the plant (upper, middle and lower) does not indicate a preference of Trichogramma spp. for parasitizing the host eggs in the lower and middle portions, rather it may be associated with the oviposition behavior of S. frugiperda which overall oviposited more in these regions (average 5.37, 5.64 and 3.29 egg masses for the lower, middle and upper regions, respectively), allowing an increase in the number of parasitized eggs, although there were no differences in the percentage of parasitism (Table 3). These results are in accordance with those reported by Wang et al. (1997), who observed a greater preference of T. ostriniae for parasitizing O. nubilalis eggs on the lower and middle regions of the corn plant, as these are the sites with the greatest concentration of the host’s egg masses. However, a greater availability of eggs in a given region of the plant does not always result in increased parasitism by Trichogramma. For example, Romeis et al. (1998, 1999) observed that Trichogramma chilonis Ishii parasitized more Helicoverpa armigera eggs on pigeonpea leaves than on flowers and pods, even though more H. armigera eggs were available on the latter structures. These differences, in this case, can be explained by the presence of trichromes, which damage the parasite.

Records of parasitism by Trichogramma on S. frugiperda eggs are scarce and usually indicate low intensities, as found by García & Sifontes (1987) in the Sancti Spiritus region of Cuba, with 4.3% parasitized egg masses, similar to our data.

Knowledge about changes in the oviposition behavior of S. frugiperda at the various phenological stages could help in the planning of field-sampling methods, location of ovicide applications, and releases of Trichogramma spp. in augmentative biological control. However, doubts remain on the effectiveness of using Trichogramma to control this lepidopteran pest as natural infestation rates are quite low. This problem needs to be addressed before this parasitoid can be used as a viable alternative to chemical controls.