Fruit flies of the subfamily Tephritinae (Diptera: Tephritidae) are widely distributed in temperate and subtropical regions of the world; however, most studies of their biology and ecology have been focused on Nearctic species inhabiting northern temperate regions (Headrick and Goeden 1998). On the contrary, few studies have documented the relationship between nonfrugivorous tephritids and their host plants in Neotropical species, in many cases, recording the first biological data and the findings of new taxa (Prado et al. 2002, 2004; Almeida et al. 2006; Norrbom and Prado 2006; Norrbom et al. 2010).

In the central region of Veracruz, Mexico, an undescribed species of Dictyotrypeta and Rhynencina spilogaster (Steyskal) (Diptera: Tephritidae) were found feeding on the capitula of Smallanthus maculatus (Cavanilles) Robinson (Asterales: Asteraceae) (V.H.O., unpublished data). The genus Dictyotrypeta comprises six species of exclusive Neotropical distribution stretching from Mexico to Argentina, with the recent inclusion of two further species, Tetreuaresta crenulata (Wulp) and Baryplegma incisa Wulp (Norrbom et al. 1998, 2010). The limited biological data available indicate that Dictyotrypeta atacta (Hendel) feeds on the capitula of Lessingianthus, Proteopsis, and Trichogonia (Asteraceae) in Brazil (Prado et al. 2002), whereas Dictyotrypeta incisa has been reared from the capitula of S. maculatus and Smallanthus riparius (Kunth) Harold Robinson in Mexico and Guatemala (Norrbom et al. 2010). The genus Rhynencina contains four Neotropical and one Nearctic species (Norrbom et al. 1998). The biology of these species is also poorly understood, as only the host plant of Rhynencina emphanes (Steyskal) is known, as these develop on Espeletia sp. flowers in Colombia (Freidberg and Norrbom 2000); likewise, Rhynencina longirostris Johnson feeds on Smallanthus uvedalius (L.) Mackenzie ex Small in North America, whose biology has already been studied (Steck and Sutton 2000, Steck et al. 2003).

The presence of Tephritinae species on their host plants is influenced by several characteristics, for instance, plant height, flowering time, and spatial host abundance (Walton et al. 1990, Rivero-Lynch and Jones 1993, Fondriest and Price 1996, Williams et al. 2001, Walker et al. 2008), in addition to climatic features as precipitation and temperature (Solbreck and Sillén-Tullberg 1986, Milton 1995). Some authors have reported significant differences in the levels of parasitism under natural conditions, suggesting that parasitism rates may be associated with host range and the particular biology of each fly species involved (Duan and Messing 1998, Alyokhin and Messing 2003, Knio et al. 2007, Edwards et al. 2009). Especially, seasonal variation during flowering season and development time of capitula may influence their rates of infestation and life cycles (English-Loeb and Karban 1992, Alyokhin et al. 2002, Woods et al. 2008, Duguma et al. 2009). However, in most of the studies, flower head infestation throughout continuous seasonal cycle has not been assessed, which allows monitoring temporal changes on infestation patterns. The aim of this study was to assess the seasonal dynamics of infestation yielded by Dictyotrypeta sp. and R. spilogaster on the capitula of S. maculatus, as well as recording the occurrence of natural parasitism along its entire flowering season, in a coffee-growing region of central Mexico.

Materials and Methods

Study Area. The study sites were located in the montane cloud forest in the central region of Veracruz. Currently, this kind of vegetation occurs as a fragmented community, composed of natural forest patches and secondary growth species, particularly as a result of the presence of shaded coffee plantations and grasslands (Williams-Linera 2007). Field work was conducted from August to November 2008, in the neighboring towns of Coatepec (19° 27′ N, 96° 06′ W; 1,252 m altitude) and Teocelo (19° 23′ N, 96° 58′ W; 1,218 m altitude), both with similar climatic conditions. Highest rainfall occurred in June (571.5 mm), and the driest month was March (24.2 mm). The highest average temperature occurred in May (24.9°C), while the lowest was recorded in November (18.3°C) (CONAGUA 2008) (Fig. 1).

Fig. 1.

Climatic conditions of monthly rainfall (mm) and average temperatures (°C) in the studied locality during the yearly cycle January–December 2008.

The Host Plant. S. maculatus is a Mesoamerican plant that occurs throughout extensive areas of Mexico, including the state of Veracruz, but its range goes as far as Panama. It is a herbaceous perennial, 1 to 2 m height (exceptionally reaching up to 5 m), with flower heads ranging from 1 to 2 cm in width, found growing in groups or panicles and rarely solitary (McVaugh 1984, Villaseñor and Espinosa 1998, Rzedowski 2001). Previous observations in the study area showed that S. maculatus grows in open disturbed areas, on roadsides, and coffee plantations, appearing in irregular groups from a few individuals up to several dozens of them, with a flowering cycle starting from July–August stretching to November–December (V.H.O., unpublished data).

Flower Head Sampling. During the flowering season of the host plant, 16 weekly collections of mature capitula were done, characterized by their ligulate and disc flowers dry, and by the presence of peripheral mature achenes. Assessing the extent of flower head infestation, collections were random among plants, and sample size was determined by the abundance and availability of mature capitula. Each time, the analyzed samples were obtained from different groups of plants, because S. maculatus is a weed periodically removed by coffee growers. Samples were placed in fine mesh bags and transported to the laboratory, where they were separated and placed in plastic chambers of 250 ml with adequate ventilation, and all information was recorded in a data matrix. Every week, each capitulum and its mature achenes were carefully dissected under the stereomicroscope, to detect and quantify the presence of immature stages. Part of the larvae and pupae sampled were kept in rearing chambers until emergence of adults. Identification of the specimens was done with reference to the characteristic damage into the achenes or receptacle, and by the size and color of pupae. Voucher adult specimens were pinned and deposited in the Entomological Collection of the Instituto de Ecología (IEXA), whereas botanical specimens were placed at the Herbarium of the Instituto de Ecología (XAL), both of these housing Instituto de Ecología A.C. (INECOL) institutional collections (Xalapa, Mexico).

Statistical Analysis. To evaluate the weekly proportion of infested capitula, linear regression of arcsine-transformed data (arcsin √proportion) were performed; because flower head production exhibited a lot of variation along the sampling period, a correction factor was applied as follows: CF = √1–n/N, where n is the weekly sample size and N the population size sampled (Zar 2010). Mean of individuals per sample within capitula was estimated by linear regression analysis (including polynomial of second and third order), and likelihood-ratio tests were used for comparisons of these models (Crawley 2007). Possible relationships between some of the dependent variables along the seasonal cycle (e.g., density of individuals per capitulum and fly species, and number of individuals per fly species) were assessed using Spearman's rank correlations. All tests were performed using the R program (Development Core Team 2011).

Results

During the annual flowering cycle of S. maculatus, 16 weekly samples accounted for 1,017 of mature capitula. Infestations by two species occurred throughout the entire season, occupying 51.5% of the flower heads. However, Dictyotrypeta sp. infested 11.3%, representing about one-fifth of all infested capitula; whereas R. spilogaster resulted more abundant infesting four times as many capitula (42.8%), and both species were found together in only 2.6% of the capitula examined (Table 1).

Seasonality. The accumulated infestation by Dictyotrypeta sp. showed differences in terms of abundance, presenting low rates of infested capitula (13.9%) during the first half of the cycle (August–September), with a notable increase in the second half (October–November), occupying 86.1% of them. On the other hand, in R. spilogaster, the highest infestations occurred during the first half of the season (64.2%), decreasing in the second half to 35.8% (Fig. 2). The linear regression analysis showed that the capitula infested by Dictyotrypeta sp. increased significantly toward the end of the season (R ² = 0.28, F1,14 = 5.55, P = 0.033), contrasting with a decrease in the infestation by R. spilogaster, which also proved to be significant (R ² = 0.34, F1,14 = 7.45, P = 0.016) to the end of the flowering cycle (Fig. 3).

Average density of individuals per capitulum infested by R. spilogaster was 2.83 individuals per capitulum (± 1.8 SD), the most frequent ratio was 2–3 specimens observed in 41.5% of cases, a single specimen in 28.7%, 4–6 specimens in 25%, and 7–10 specimens in only 4.8% of the capitula. High densities of individuals (≥ 3 individuals per capitula) mainly occurred through September and early October. In contrast, lower densities were observed starting August and extending from late October to November. Comparisons of the rate of individuals observed throughout the flowering season, using the third-order polynomial regression analysis, showed significant differences (R ² = 0.55, F3,12 = 4.98, P = 0.017; Fig. 4), and these infestation rates were positively correlated with increases in the capitula infested (Spearman's rank correlation, ρ = 0.70, P = 0.001).

Table 1.

Capitula infested by two tephritid species throughout the flowering season of S. maculatus

Fig. 2.

Proportion of the capitula infested by Dictyotrypeta sp. and Rhynencina spilogaster along the flowering season of S. maculatus.

Meanwhile, the average density of Dictyotrypeta sp. individuals was 1.40 individuals per capitulum (± 0.7 SD); thus, most of the capitula was accounted to be infested by a single specimen in 68.7% of cases, two specimens in 23.5%, and 3–5 specimens in only 7.8% of the capitula. However, tested comparisons of the rate of individuals inside the capitula throughout the flowering season neither showed a correlation (P> 0.1; Fig. 4) nor was there a correlation between average of individuals and the number of capitula infested (Spearman's rank correlation, ρ = 0.31, P = 0.12).

Immature Stages and Infestation. Total infestation by Dictyotrypeta sp. was represented by 162 specimens (larvae 15.4%, pupae 84.6%), and correlation analysis showed that individual abundance increased significantly toward the second half of the season (Spearman's rank correlation, ρ = 0.51, P = 0.019). Larvae were absent from the initial two-thirds of the season, appearing later into this period; meanwhile, pupae were observed throughout the entire cycle, with the highest abundance focused in the second half (Fig. 5A). Both larvae and pupae were found mostly within the mature achenes (82.7%), but they were also discovered inside the receptacle (17.3%). Achenes infested by larvae of this species showed a typical damage, indicated by an orifice covered with waste from feeding, while damage to the receptacle was recognized by the presence of galleries that lead toward an achene.

R. spilogaster was recorded by a total of 1,234 individuals, without significant variations along the entire season (Spearman's rank correlation, δ = -0.41, P = 0.055). In this case, most individuals recovered were larvae (80.2%) appearing throughout the flowering period, while pupae accounted for only 19.8% of the specimens, especially observed in the early weeks of the flowering season (Fig. 5B). Larvae and pupae were found exclusively within the mature achenes without apparent traces of infestation, so its presence was detected just dissecting each one of them.

Parasitism. To assess natural parasitism in Dictyotrypeta sp., 118 pupae were kept alive in emergence chambers. Here, 22 adult flies were obtained in a rate of emergence of 15.7 days (± 3.1 SD), while from remaining pupae 88 Hymenopteran parasitoids were recovered, representing 74.6% of parasitism in the sample. All parasitoid specimens belong to five species from the families Figitidae sp. 1 (48) and Pteromalidae sp. 1 (37), followed by three species of Eupelmidae (1), Eurytomidae (1), and Eulophidae (3), respectively. The average times of emergence were as follows: Figitidae sp. 1 = 79.7 days (± 35.4 SD) and Pteromalidae sp. 1 = 50.9 days (± 34.3 SD), while emergence times of the single specimens of Eulophidae, Eupelmidae, and Eurytomidae ranged from 15 to 19 days (Fig. 6A). All parasitoids were reared from a single puparium, excluding the three Eulophid specimens that showed a gregarious parasitism.

Fig. 3.

Correlation analysis of the proportion of capitula infested by two Tephritinae species throughout the flowering cycle of S. maculatus (August–November 2008).

Fig. 4.

Analysis of the density of tephritid individuals infesting the capitula of S. maculatus (August–November 2008): mean of individuals observed in R. spilogaster showed significant differences (P < 0.05).

Likewise, 135 puparia of R. spilogaster were kept in the emergence chambers, and 69 of them remained alive for 303.9 days (± 71.2 SD), whereas 66 parasitoids of Pteromalidae sp. 2 emerged from the other puparia, which came into diapause because adults lasted for 303 days (± 134.2 SD) after collection, and accounted for 48.9% of parasitism in the sample (Fig. 6B).

Discussion

Insects that share food resources follow two survival strategies, either an avoidance strategy or maximizing their competitiveness (Zwölfer 1979). In Tephritinae infesting a single host, phenomena of interspecific competition have been reported, but there are also evidences of avoidance strategies during the seasonality of the host (Michaelis 1986, Straw 1989a, Knio et al. 2001, Story 2002, Holt and Zwölfer 2007). Results of this study showed differences in temporary patterns of infestation by Dictyotrypeta sp. and R. spilogaster, suggesting distinct strategies of resource exploitation, as only 2.6% of the capitula were simultaneously occupied by two species.

The flowering season of the host offered a constant resource supply for both tephritids; however, R. spilogaster presented populations approximately four times larger than Dictyotrypeta sp. These results can be related to the particular biology of each species, because the mature larvae of the former are small-size specimens, and each individual feeds exclusively on a single achene. Although the larvae of the latter are twice as large, their food requirements are greater, and therefore, it was found in the achenes and in the receptacle. The habit of feeding in the receptacle is a facultative mode of resource exploitation, and it may provide additional nutrients for continued development within the capitulum (Knio et al. 2001).

Fig. 5.

Abundance of larvae and pupae recovered from 16 samples along flowering season of S. maculatus in central Veracruz, México: (A) Dictyotrypeta sp. and (B) R. spilogaster.

The average density of 1.4 individuals observed in Dictyotrypeta sp. suggests that females usually lay 1–2 eggs per capitulum and probably do so during its early development, because damage was always observed in the center of the receptacle, and in most of the cases, a later transferral took place to the achene. Low densities per capitulum indicated a small clutch size as a strategy to maximize the probability that offspring found enough food resources for development, but it may also indicate the existence of a marking pheromone after oviposition, which would be recognized by conspecific females (Straw 1989b; Pittara and Katsoyanos 1990; Lalonde and Roitberg 1992a, b). In the case of R. spilogaster, the average density of 2.8 individuals per capitulum suggests that females made multiple ovipositions, depositing a single egg in each achene, which provides enough resources for individual development, whereas an increase in the density of individuals per capitulum was correlated with population growth along the season.

In Dictyotrypeta sp., only pupae were found at the beginning of the season, suggesting that oviposition may occur during the early development of the capitula. However, to the end of the season (S10–S16), larvae also appeared, indicating a second generation of this species. From the start of the season, some adults emerged in an average time of 15 days (range 9–22), whereas during the second half of the season, the two most abundant parasitoid species emerged in an elapsed time of 50–79 days, suggesting that second generation could enter diapause. Contrastingly, R. spilogaster is clearly a univoltine species with a long life cycle that entered diapause, because larvae as pupae were present throughout the entire season, and larvae were represented by 80.2% of the specimens, while 19.8% were pupae remaining viable for up to 303 days on average. Nearly identical times of emerged parasitoids were also recorded. Similar observations were done in the North American sister species, R. longirostris that breeds in the capitula of S. uvedalius, showing that flies and parasitoids exhibited a single annual generation (Steck et al. 2003).

Mature achenes per capitulum produced by S. maculatus were estimated in 13.03 (± 2.4 SD; N = 276), so the total number of damaged achenes by two tephritids was 20.6% (Dictyotrypeta sp. 2.02% and R. spilogaster 18.6%). It means that nearly one-fifth of the seeds were predated, which represents a relatively low impact on the reproductive success of the host plant. However, the damage to the receptacle by the former species may have an indirect effect on the seed production, because studies on other receptacle feeders, such as Chaetostomella cylindrica and Terellia colon, suggest that they may restrict the flow of nutrients, reducing the fitness and weight of healthy seeds (Kinkorová and Mícková 2006, Koprdova and Martínková 2006).

Fig. 6.

Parasitoids recovered and times of emergence in two Tephritinae species: (A) Figitidae sp. 1 and Pteromalidae sp. 1, parasites of Dictyotrypeta sp. and (B) Pteromalidae sp. 2, parasite of R. spilogaster.

There are evidences of high spatial and temporal variability in predispersal seed predation by insect herbivores (Kolb et al. 2007, Abdala-Roberts et al. 2010). Moreover, the action of the natural parasitism found in some Tephritinae species used for control of weeds could also exert a reduction in their effectiveness, limiting their establishment and impact as agents for biological control (Marshall et al. 2004, Edwards et al. 2009).

Results of multiple interactions between these tephritids and their parasitoids were restricted to single-year sampling at a single locality. Further research on several cycles and localities are required to make generalizations of these assertions. Multiple interspecific interactions may also affect whether there is a reduction or an increase in seed predation (Lamp and McCarty 1982, Groenteman et al. 2007, Tonkel and Piper 2009). Moreover, species of Coleoptera and Lepidoptera were observed either in the achenes, the receptacle, or both, whose trophic interactions are being assessed as a whole (V.H.O., unpublished data). This will provide a comprehensive view of the impact of these herbivores on the reproductive success of the host plant.