Relative Abundance of Ceratitis capitata and Anastrepha fraterculus (Diptera: Tephritidae) in Diverse Host Species and Localities of Argentina

Abstract

Two fruit fly species (Diptera: Tephritidae) of economic importance occur in Argentina, the Mediterranean fruit fly, Ceratitis capitata (Wiedemann), and Anastrepha fraterculus (Wiedemann). Here, we compared the relative abundance of these fruit pests in 26 fruit species sampled from 62 localities of Argentina in regions where C. capitata and A. fraterculus coexist. In general, C. capitata was predominant over A. fraterculus (97.46% of the emerged adults were C. capitata), but not always. Using the number of emerged adults of each species, we calculated a relative abundance index (RAI) for each host in each locality. RAI is the abundance of C. capitata relative to the combined abundance of A. fraterculus and C. capitata. Some families of fruit species were more prone to show high (Rutaceae and Rosaceae) or low (Myrtaceae) RAI values, and also native plants showed lower RAI values than introduced plants. RAI showed high variation among host species in different localities, suggesting a differential use of these hosts by the two flies. There were localities where A. fraterculus was not found in spite of suitable temperature and the presence of hosts. Most host species showed little variation in RAI among localities, usually favoring C. capitata, but peach, grapefruit, and guava showed high variation. This suggests that these fruit species are suitable for both fruit flies but more favorable to one or the other, depending on local environmental conditions (e.g., relative humidity and degree of disturbance) of each locality.

The family Tephritidae includes some of the most important fruit pests worldwide (White and Elson-Harris 1992). The economic damage caused by these flies is two-fold: direct damage to the fruit (larval activity) and limited access to potential markets because of quarantine restrictions imposed by countries that are free of these pests (Malavasi et al. 1994). In the American Continent, flies within Anastrepha, Ceratitis, Rhagoletis, and Toxotrypana cause the most economic damage (Landolt 1985, Enkerlin et al. 1989, Aluja 1994).

In Argentina, there are two quarantine species of fruit flies: the Mediterranean fruit fly, Ceratitis capitata (Wiedemann), and Anastrepha fraterculus (Wiedemann) (Aruani et al. 1996). C. capitata is native to Africa and has a wide distribution, covering many tropical, subtropical, and temperate regions of the world (Copeland et al. 2002). This species shows a high adaptability to diverse climates as well as a large number of host fruit species (>350; Liquido et al. 1991). Its presence in Argentina was first recorded at the beginning of the 20th century in orchards located in the vicinity of Buenos Aires city (Vergani 1952). Later, it was reported in commercial orchards of northeastern and northwestern regions of the country. The last region in which it was reported was northern Patagonia (southern Argentina), where C. capitata was first detected in 1952 (Rial 1997). A. fraterculus is native to South America and is distributed from Mexico to Argentina, but there is morphological and genetic evidence indicating that there are many cryptic species (Steck 1991, Hernández-Ortíz et al. 2004) and that not all the species within this species complex are pests (Aluja et al. 2003). In Argentina, A. fraterculus is mainly distributed in regions with tropical and subtropical climate (Ovruski et al. 2003). It is also a polyphagous species that attacks different families of fruit species, but the number of hosts cited is smaller than that for C. capitata (≈80 species; Norrbom 2004). Both species cause significant annual losses to the fruit production of Argentina and constitute a major barrier to the expansion of this market (Ovruski et al. 1999).

In spite of the importance of these fruit fly species, there is little information published on the relative abundance of C. capitata and A. fraterculus in areas of Argentina where they coexist. Other than Vergani (1956), the only map available showing the distribution of the two species at the national level is hypothetical (Ortiz 1999). Variability in the relative abundance of the two species among different regions was reported by Vattuone et al. (1995) based on average values among different hosts, but the relationship between the two species strongly depended on the fruit considered. Comparisons based on trapping data (FAO 1989, Segade and Polack 1999, Vattuone et al. 1999) are biased by the degree of attraction that adult flies of the two species show to the bait used in the traps. Some local studies provide data based on fruit sampling (Costilla 1967, FAO 1989, Putruele 1993, Vattuone et al. 1999). In Tucumán province, Costilla (1967) found a greater proportion of C. capitata than A. fraterculus in citrus orchards of grapefruit, Citrus paradisi L., and orange, Citrus sinensis (L.) Osbeck. Schliserman and Ovruski (2004), working in areas of the same province with much native vegetation, found mainly C. capitata in bitter orange, Citrus aurantium L. Ovruski et al. (2003) also reported proportions of C. capitata and A. fraterculus adults and larval infestation levels in wild or commercially grown plants, both native and introduced, and emphasized the importance of Citrus spp. as hosts of C. capitata and native fruit species as hosts of A. fraterculus. In Entre Ríos province, Putruele (1993) recovered both species from grapefruit; peach, Prunus persica L.; fig, Ficus carica L.; and mandarin, Citrus reticulata Blanco; but only in grapefruit was A. fraterculus more abundant than C. capitata. In the same province, apple, Malus domestica Borkh, and pear, Pyrus communis L., were infested only by C. capitata, whereas pomegranate, Punica granatum L., and quince, Cydonia oblonga Mill, were attacked only by A. fraterculus (Putruele 1993). In Catamarca province, Vattuone et al. (1999) reported that orange, peach, grapefruit, mandarin, and kumquat, Citrus aurantium variety myrtifolia Ker-Gawl, were infested only by C. capitata. In La Rioja province, Nasca et al. (1996) recorded greater infestation of C. capitata than A. fraterculus in fig; persimmon, Diospyros kaki L.; quince; pomegranate; plum, Prunus domestica L.; and apricot, Prunus armeniaca L. Although these studies are valuable, there has been no comprehensive study comparing the relative abundances of both species along the variety of hosts and regions present in Argentina.

The aim of this study was to provide these comparative data for C. capitata and A. fraterculus. We describe the relative abundance of these flies through the analysis of data obtained from fruit samplings of many different host species. Specifically, we were interested in comparing the possible variability in the abundances of the two fly species 1) among different host species from the same locality and 2) among different localities for the same host. We also examined the influence of plant taxonomy and origin (native or introduced) as well as variations in the climate on the relative abundance of these two species.

Materials and Methods

Sampling.

Fruit sampling was carried out in those political provinces of Argentina where the two fruit flies are reported to coexist (see Appendix 1). The localities ranged from undisturbed areas with most of the vegetation being native to very disturbed systems, such as suburban areas or agricultural landscapes. The climatic characteristics, specifically rainfall and relative humidity, also varied widely.

Appendix 1. Ecological regions (Cabrera and Willink 1973, Cabrera 1976, Anonymous 1992), geographical coordinates, and altitude for each locality sampled

Appendix 1. Continued

Sampling was performed mainly during the fruiting season of 1999–2000. Nevertheless, because of inherent logistical complexities faced in a study covering such a large geographical area, sampling data derived from the 2000–2001 season in some localities (e.g., Concordia), and from the 1998–1999 fruiting season in others (e.g., Posadas). To include as much area as possible, we also present fruit-collecting data from 1993 for three localities in the province of Catamarca. For the localities that were sampled during more than one fruiting season, no substantial differences between years were found in the relative abundance index (RAI) values (not shown), so these data were pooled. Descriptions of each locality are provided in Appendix 1.

All fruit species sampled have been previously cited as hosts for both fruit fly species (Liquido et al. 1991, Norrbom 2004), and they included native and introduced plants. In each locality, at least 10 fruit were sampled for each host species. Only fruit with evident signs of infestation by fruit flies was collected. The sampled fruit was placed in plastic trays over a layer of sand (or vermiculite), which was used by larvae as a pupation substrate after leaving the fruit. Pupae were separated from the sand using a sieve and then transferred to new containers, where they were maintained until emergence. Adults were identified, and the number of individuals of each fruit fly species was recorded. For inclusion in the analysis, at least 10 adults were required for a given host species and locality.

Data Analysis.

To describe the relationship between the abundances of the two fruit fly species, a relative abundance index (RAI) was calculated for each host species in each locality according to the following formula: RAI_xy = Cc/(Cc + Af), where Cc and Af are the number of emerged adults of C. capitata and A. fraterculus, respectively, for host X in locality Y. RAI ranges from 0 (exclusive presence of A. fraterculus) to 1 (exclusive presence of C. capitata). The RAI value for a given host (RAI_X) was estimated as the mean value for all of the localities, where that particular host was sampled. According to the RAI value obtained, hosts and localities were assigned to one of five categories: exclusive presence of one or the other species (RAI = 0 or RAI = 1), both species present but higher abundance of one or the other (0 < RAI < 0.33 or 0.66 < RAI < 1), and intermediate cases (0.33 ≤ RAI ≤ 0.66).

For every host–locality combination, we compared the observed frequencies with expected frequencies estimated multiplying the number of cases by the probability that in a random sample of 10 pupae: all of them were C. capitata; all of them were A. fraterculus; seven to nine pupae were of one of the two flies (and three to one of the other); and four to six pupae were of either of them. A chi-square goodness-of-fit test was performed to compare observed and expected frequencies (StatSoft, Inc. 2000).

A Mann–Whitney test was performed to compare the RAI values recorded for native and introduced fruit species (StatSoft, Inc. 2000). To compare the RAI values found for the main taxonomic families of fruit species (Myrtaceae, Rosaceae, and Rutaceae), a nonparametric analysis of variance (ANOVA) (Kruskal–Wallis test) was performed. When this analysis showed significant differences, nonparametric multiple comparison tests were performed (Analytical Software 2000).

Results

We obtained fruit samples from 62 localities in 12 of the 24 political provinces of Argentina, covering six ecological regions (see Appendix 1) and representing the complete area shared by both species of fruit flies in the country. In total, 30,354 fruit were collected, belonging to 20 host species from eight families of plants. Three of these families are native to South America (Myrtaceae, Caricaceae, and Olaceae), and five are introduced (Actinidiaceae, Ebenaceae, Moraceae, Rosaceae, and Rutaceae) (Table 1). Of 43,142 pupae recovered from the fruit, a total of 27,301 adults emerged: 23,608 (97.46%) were C. capitata, and 3,693 (2.54%) were A. fraterculus. We recovered parasitoids in only seven sampling sites, and in those cases the percentage of parasitism was never higher than 5%.

Table 1.

Host species sampled

Thirty-two localities and 12 host species were sampled in northwestern Argentina (NWA) (Fig. 1; Appendix 1), and 30 localities and 18 host species were sampled in northeastern Argentina (NEA) (Fig. 2; Appendix 1). A general predominance of C. capitata was observed, which was higher in NWA than in NEA (U = 1096.5, P = 0.042; Mann–Whitney test). The proportion of samples in which RAI = 1 was 72% in the NWA and 50% in NEA (Table 2).

Fig. 1.

RAI for each locality and host species sampled in NWA

Fig. 2.

RAI for each locality and host species sampled in NEA

Table 2.

RAI values per localities and hosts

Table 2.

Continued

In some cases, RAI showed a marked variation among hosts in the same locality, e.g., in Yuto from NWA and in Posadas and Concordia from NEA (Figs. 1 and 2; Table 2). In those cases, some hosts had a RAI that was very high (mandarin from Yuto; persimmon from Posadas; and peach, guava, and orange from Concordia) or very low (guava, Psidium guajava L., from Yuto; grapefruit and guava from Posadas; and feijoa from Concordia), whereas other hosts showed intermediate values. In other localities, the variation among hosts was considerably lower. For example, in Chilecito (Fig. 1) and in Saenz Peña and San Pedro (Fig. 2), C. capitata was the main fruit fly species found, whereas in Montecarlo (Fig. 2)A. fraterculus was more abundant than C. capitata in almost all host species.

Irrespective of locality, many fruit species showed values of RAI closer to 1 than to 0, e.g., orange and fig. No host species consistently had values of RAI near 0 (Table 3). Other fruit species, such as grapefruit, peach, and guava showed a wide range of RAI values (Table 3).

Table 3.

Number of localities in each RAI categories for each host species

Native host plants showed significantly lower values of RAI than introduced hosts (Table 4). Significant differences were also found among the three most abundant families of host species (Table 4). Nonparametric multiple comparisons showed differences between Myrtaceae and Rutaceae (df = 1, P < 0.05), whereas host species belonging to the family Rosaceae gave intermediate values of RAI that did not differ statistically from the other two.

Table 4.

RAI values by origin and botanic family of host plants

The number of cases in which the two species of fruit flies shared a host in equal abundance (Fig. 3) was lower than expected, whereas the cases in which only one fruit fly species was recovered from a host was much higher than expected (χ² = 326.39, P < 0.001).

Fig. 3.

Number of expected and observed cases (for each combination of host species and locality) for each classification category of RAI (see text for delimitation of the RAI categories). The observed frequencies differed statistically from the expected frequencies (chi-square test: χ² = 326.39, P < 0.001)

Discussion

In some localities, we sampled fruit suitable for both C. capitata and A. fraterculus, and only one fruit fly species was recovered, or one of them occurred in extremely low abundance (Sáenz Peña, Monte Caseros, Montecarlo, San Pedro, and Bella Vista in Fig. 2; Chilecito and Campo Santo in Fig. 1). This fact suggests that factors other than availability of suitable hosts are limiting the establishment of one fruit fly species and not the other. Among these factors, we can postulate abiotic factors, such as temperature, humidity, or rainfall, and biotic factors, such as the duration of periods without mature suitable fruit, the competition between the two fruit fly species (Celedonio-Hurtado et al. 1995, and references therein), and the degree of environmental disturbance. Parasitism could bias the relative abundances of these fruit fly species favoring one of the two species (Ovruski et al. 2004); however, the numbers recovered were so low that parasitism did not significantly affect the RAI values obtained here.

Consideration of specific localities offers some insight into the factors affecting fly distribution. For example, in Saenz Peña, A. fraterculus was absent from all the host species sampled, including guavas, one of the primary host for this species (Putruele 1993, Aluja et al. 2000, Selivon 2000). The thermal regime of Saenz Peña is suitable for this species, but the annual relative humidity is close to 50%, suggesting that this site is too dry for A. fraterculus. Orán, by contrast, has a similar thermal regime but a higher annual relative humidity, and here guavas were heavily infested by A. fraterculus. Environmental disturbance supposedly favors C. capitata (Putruele 1997, Malavasi et al. 2000, Ovruski et al. 2003). This seems to be the case for Saenz Peña with a very extensive agricultural landscape and very little native vegetation, but not for Orán [although currently there is a heavy trend to deforest the Yungas and start soybean, Glycine max (L.) Merr., plantations]. However, in Montecarlo, where the original environment also has been disturbed, there is high abundance of A. fraterculus, indicating that the local environment may have a stronger impact than the disturbance on the relative abundance of this fruit fly. Montecarlo, in Misiones province, with subtropical climate and high relative humidity exhibits a landscape with dense vegetation and backyards with fruit fly host plants, many of them native, that provide excellent refuges for A. fraterculus. In San Pedro, the thermal regime and the relative humidity seem appropriate for the development of A. fraterculus; however, RAI values for this locality are high because this species is present in a small number of hosts and in low abundance. The absence of suitable hosts for A. fraterculus from late autumn to middle spring (Segura et al. 2004) is the most likely explanation for this pattern.

The great variation in RAI among hosts (see Posadas, Concordia, and others) indicates a pattern of differential use of the available hosts. The possible explanations are 1) a different pattern of host preference in adults of each fruit fly species; 2) asymmetric interspecific competition (the result of which depends on the fruit species; Fitt 1989); 3) differential mortality of eggs, larvae, or both in each host species (Carey 1984); or 4) differential ability of each fruit fly species to find and infest different fruit species.

Citrus spp. were found to be better hosts for C. capitata than for A. fraterculus, regardless of the inter-locality variation in biotic and abiotic factors that may favor one of the other species, at least in Argentina (Tables 2 and 3). Mandarin, orange, and bitter orange showed lower variation in RAI than the other fruit, and the most abundant species was always C. capitata. In agreement with other observations (Malavasi and Morgante 1980, Putruele 1993, Nasca et al. 1996, Vaccaro 2000, Ovruski et al. 2003) we found that grapefruit was the only Citrus with values of RAI favoring A. fraterculus. Several studies (Nascimiento et al. 1984, da Silva-Branco et al. 2000, Aluja et al. 2003) have shown the low suitability of Citrus spp. as host for Anastrepha spp., but in the laboratory, forced development on grapefruit, orange, and lemon, Citrus limon L., showed better recovery of A. fraterculus pupae in the case of grapefruit (Gramajo 2004). Ovruski et al. (2003) proposed a stronger attraction of C. capitata toward infochemicals (sensu Dicke and Sabelis 1988) produced by Citrus as another explanation for the high RAI values found in these host species (for example, Howse and Knapp (1996) suggested that some components of male C. capitata pheromone are similar to volatiles emitted by Citrus trees and fruit). Asymmetrical larval competition favoring C. capitata also could be postulated (but then, it is not clear why this asymmetry would be reversed for grapefruit).

From apple and pear, we recovered pupae of C. capitata in San Pedro (Buenos Aires province), as did previous studies in the neighboring province of Entre Ríos (FAO 1989, Putruele 1996). Nasca et al. (1996) recorded A. fraterculus pupae from pear collected in Antinaco-Los Colorados Valley (La Rioja province). Several studies carried out in Brazil report the presence of A. fraterculus in these two fruit species (Malavasi et al. 1980, Kovaleski et al. 2000, Nora et al. 2000). However, Ovruski et al. (2003) did not find infestation by any fruit fly in apple and pear (157 and 196 fruit sampled, respectively) in NWA, questioning the status of these fruit species as hosts for the two fly species. Surveys of these two fruit species should be expanded, with emphasis in NWA.

C. capitata was more abundant in plant species belonging to the family Rutaceae, whereas A. fraterculus was predominant in plants of the family Myrtaceae. Species belonging to the family Rosaceae showed intermediate values of RAI, roughly corresponding to the comparison of RAI between introduced and native species: C. capitata dominates in introduced plants (Rutaceae and Rosaceae, among others), whereas in native plants (Myrtaceae among others) A. fraterculus shows higher abundance (Table 4). Various Brazilian authors also found this (Malavasi and Morgante 1980, de Souza Filho et al. 2000, Malavasi et al. 2000, Veloso et al. 2000). Eskafi and Kolbe (1990) described the same pattern in Guatemala, although their samplings also included different Anastrepha species. Ovruski et al. (2003) also reported that the introduced fruit species favor C. capitata and that the native species serve as a reservoir for A. fraterculus (with two exceptions discussed below). The fact that almost 85% of all fruit species sampled are exotic in Argentina could be responsible, at least to some extent, for the high predominance of C. capitata in our samplings.

As noted above, A. fraterculus showed better yields in native plants, probably because of their common evolutionary history. C. capitata, being such a polyphagous species with such high reproductive capacity (Liquido et al. 1991), may gain advantage in introduced hosts with which A. fraterculus has had less contact in its evolutionary history. Moreover, plant species introduced to the Americas from the same region as C. capitata, for example, coffee, Coffea arabica L., constitute good hosts for this fly (Harris and Lee 1989, Malavasi et al. 2000). Interestingly, Copeland et al. (2002) did not find C. capitata in guavas sampled in Kenya (in a sample of 84 fruit), where this fly is native, and P. guajava is an introduced species. In our study, C. capitata was found infesting guavas in 10 of 14 localities sampled (Table 2; Figs. 1 and 2). Another exception, mentioned by Ovruski et al. (2003), might be peach and plum (both introduced species). But, if we calculate the RAIs from their published data (0.16 and 0.25, respectively) they differ markedly from those found in the current study (average RAI of 0.77 and 0.76 for peach and plum, respectively). In Ovruski et al. (2003), however, the sampling of these two host species occurred in forest areas, scattered among native vegetation, whereas in our study, they occurred in highly disturbed areas, illustrating the strong influence of the environment on the RAI values. RAI values lower than 0.10 for peach, in Itacuruzú, Montecarlo, and San Javier, located in areas with native vegetation and subtropical climate are good examples supporting our explanation.

We found that both species tended to occur alone many more times than they occurred together sharing one host in one locality. We could interpret this pattern as competitive exclusion of one species by the other. It has been suggested that, when two or more fruit fly species coexist, some form of competition for hosts could arise (Duyck et al. 2004). Most examples of interspecific competition among tephritids derive from situations in which a new species has been introduced into a given environment (Duyck et al. 2004). For example, interspecific competition was proposed to explain the displacement of C. capitata by Bactrocera dorsalis (Hendel) (Fitt 1989) in Hawaii and by Bactrocera tryoni (Froggatt) in Australia (Allman 1939, Andrewartha and Birch 1954, Christenson and Foote 1960, Bateman 1971, Fitt 1989). Owing to the short period of coexistence (≈100 yr), mechanisms that tend to minimize the competition for resources (“avoidance” sensu Díaz-Fleischer et al. 2000, Duyck et al. 2004, Sivinski et al. 2004) have probably not yet evolved. However, the patterns of relative abundance give only indirect evidence for the existence of interspecific competition. Fitt (1989) suggests looking for direct evidence of competition, as a modification in the abundance of one species after manipulating the abundance of the other.

In conclusion, this first attempt to analyze the relative abundance of C. capitata and A. fraterculus covering different regions and different hosts in Argentina proved that both species coexist here in several areas and exhibit similar ecological requirements. Therefore, we should expect strong competition between them in habitats where the resources are scarce, as in wild or urban habitats where the density of host plants is usually low. These habitats serve as refuges for small populations that are usually neglected by traditional pest control efforts and may be foci where reinfestation starts. Future studies of interspecific competition between C. capitata and A. fraterculus should focus on these habitats to produce valuable information for area-wide management of these pests. C. capitata is the major fruit fly pest in almost all regions in Argentina. The sterile insect technique (Knipling 1955), successfully implemented in La Rioja, Mendoza, and San Juan provinces, and the Patagonia region (De Longo et al. 2000, Frissolo et al. 2001, Sánchez et al. 2001), aims at the eradication of C. capitata. It would be very useful to be able to predict the response of A. fraterculus populations to a marked decrease in the density of C. capitata and identify areas likely to experience an increase in the A. fraterculus population, thereby avoiding outbreaks of this pest.

Acknowledgments

We thank to M. de la Vega, D. Figueroa, L. Hansen, F. Milla, J. Mousques, N. Petit-Marty, A. Polack, L. Santinoni, G. Segade, and E. Willink, for collaboration during the fruit collection. We also thank E. Vattuone for help during the bibliographic search and D. O. McInnis, M. Viscarret, and S. López for comments on a previous version of this manuscript. We are grateful for revisions by T. Shelly and two anonymous reviewers. This work was supported by Foncyt PID 615 to J.L.C. and PIP-CONICET n° 04973 to S.M.O.