Searching for a common host: parasitoids of Lema daturaphila on Datura stramonium in Central Mexico

Study system

In Mexico, Datura stramonium is commonly found in human-disturbed areas (Núñez-Farfán & Dirzo, 1994), where it is consumed by folivore insects such as Epitrix parvula and Sphenarium purpurascens, as well as by the pre-dispersal seed predator Trichobaris soror (Castillo et al., 2015; De-la-Mora, Piñero & Núñez-Farfán, 2015). However, Lema daturaphila stands out because its gregarious larvae can cause defoliation as high as 100% in some D. stramonium populations (Núñez-Farfán & Dirzo, 1994). Except for the pupa, all the stages of its life cycle occur on the plant, beginning when females oviposit clutches of 15–30 ovoid eggs on the underside of the leaves (Kogan & Goeden, 1970). The bright yellow eggs, which are around 1 mm long, gradually darken until hatching (Omer-Cooper & Miles, 1951). Lema daturaphila larvae then passes through four instars, which are morphologically similar to each other (Kogan & Goeden, 1970). The larvae deposit a mass of feces on their backs, and their oral secretions are dark brown fluids that larvae can suck back (Omer-Cooper & Miles, 1951). In these immature stages, L. daturaphila larvae can be target of unidentified tachinid flies and hymenopteran parasitoids (Omer-Cooper & Miles, 1951; Cabrales-Vargas, 1991; Hernández-Cumplido, 2006). Additionally, an ancient record mentions Emersonella lemae, a chalcid wasp that attacks L. daturaphila eggs on D. stramonium in Washington, D.C. (Girault, 1916; Chittenden, 1924). When the last instar is ready, the larvae leave the plant and build a cocoon in the soil where they pupate until the adult emerges and returns to the plant to feed and reproduce (Kogan & Goeden, 1970).

Sampling

Between August and October 2018 and 2019, we sampled nine natural populations of D. stramonium and L. daturaphila in Central Mexico (Fig. 1). Their geographic location and climatic conditions are detailed in Table S1. The collection of L. daturaphila was authorized by the Secretaría de Medio Ambiente y Recursos Naturales, Gobierno de México (Oficio N° GPA/DGVS/011980/17). In the second year of sampling, we included the populations of Dolores and San Martín, located in the states of Querétaro and Puebla, respectively.

In each population, we randomly collected ±30 L. daturaphila egg clutches and, additionally, ±30 individuals of the 2° or 3° larval instars. All individuals were reared under controlled conditions (22 °C, 12:12 h photoperiod). We recorded the number of eggs per clutch and individual clutches were separated into Petri dishes. Larvae were transferred to plastic containers and fed with leaves from a natural D. stramonium population in the Pedregal de San Angel natural reserve (southern Mexico City). All eggs and larvae were inspected daily for emerging chrysomelids or parasitoids. Emerging parasitoids were preserved in 70% ethanol. They were examined with a LEICA MZ125 binocular microscope and determined using taxonomic keys with assistance from entomologist specialists (Wood, 1987; Hansson, 2002; Fernández & Sharkey, 2006). Representative specimens were photographed using a Nikon D3400 digital camera and with a Hitachi SU1510 scanning electron microscope. Photographs were processed with ON1 Photo Raw 2022 and Inkscape 1.3.

Analysis of oral secretions

We selected non-parasitized egg clutches from Bernal, Querétaro, and reared L. daturaphila larvae under the same environmental conditions mentioned above. Oral secretions were collected by squeezing a larva behind its head, inducing immediate regurgitation. Using a micropipette, we collected 1–5 µL of oral secretion per larva. The crude oral secretions from 10–15 larvae were stored in Eppendorf tubes at −20 °C until further processing.

The chemical composition of the oral secretions was analyzed using a liquid chromatography/time-of-flight/mass spectra (HPLC-TOF-MS) system, following the protocol described for D. stramonium by De-la-Cruz et al. (2020). First, the crude oral secretions were centrifuged at 12,000 rpm for 30 min to remove pieces of plant material and other solids. The supernatant was then centrifuged under the same conditions and filtered through a 0.22 µm sterilized membrane.

The sample was re-suspended in 100 µL of MeOH and centrifuged at 12,000 rpm for 2 min. The supernatant was recovered, re-suspended in 500 µL of MeOH, and centrifuged again. Finally, chromatographic separations were performed on an Agilent ZORBAX HPLC column, following the De-la-Cruz et al. (2020) protocol. Chemical compounds were identified using mass spectrometry, retention times, and molecular formulas obtained from chromatograms with MassHunter Workstation software (version B. 06.00; Agilent Technologies, Santa Clara, CA, USA).

Data analysis

All statistical analyses and graphs were performed using R version 4.3.1 (R Core Team, 2021) and the ggplot2 package (Wickham, 2016).

We first conducted descriptive statistical analysis to examine the variation in the number of eggs and larvae of L. daturaphila, as well as parasitoid individuals. Given that the data were counts and there was high variance in the number of eggs per clutch, we used the MASS package (Venables & Ripley, 2002) to fit a generalized linear model (GLM) with a negative binomial distribution. This model allowed us to assess differences in clutch size across populations, years, and their interaction. Next, to assess the incidence of parasitized egg clutches (binomial response: parasitized or non-parasitized), we performed a GLM with a binomial distribution, including the interaction between population and year.

To determine whether the number of wasps emerging per clutch varies across populations and years, we fit a GLM with a negative binomial distribution, where the number of eggs per clutch and population were used as predictor variables. We also performed a logistic regression to evaluate whether the probability of parasitism depends on clutch size each year. For modeling parasitoidism in L. daturaphila larvae, we used a GLM with a Poisson distribution. Since we only collected a sample of larvae in each locality, the number of larvae per population was the sole predictor variable. Finally, to explore patterns in parasitoid incidence and their relationship with environmental variables, we performed a principal component analysis (PCA), including average climatic conditions per year from each locality.

Parasitoid species

The egg stage of L. daturaphila was parasitized by Emersonella lemae, an idiobiont wasp approximately 1 mm in length. This species belongs to the Eulophidae family (superfamily Chalcidoidea) and exhibits sexual dimorphism in the metasoma: females are round, while males have a more hexagonal shape (Fig. 2). Parasitism occurs shortly after L. daturaphila deposits its eggs (Fig. 3A). Multiple E. lemae females likely parasitize a single clutch, as we observed several wasps visiting the same clutch in the field. During L. daturaphila larval development, the eggshell remains transparent (Fig. 3B), but it turns opaque and golden-brown if parasitized (Fig. 3C). Healthy eggs hatch 5–7 days after oviposition, while wasp development takes over a month. Upon emergence, the adult wasp punctures the egg with its mouthparts, typically at the top of the egg (now a pupa) (Fig. 3D).

In the larval stage of L. daturaphila, we identified four koinobiont flies from the genera Patelloa, Winthemia, Heliodorus, and Pseudochaeta. These parasitoids emerge during the pupal stage and are often found around larvae (Fig. 4A), which respond by releasing a bubble of oral secretions (Fig. 4B). After parasitization, the larvae continue feeding until pupation, at which point flies make a reddish-brown cocoon that hatches 15–20 days later (Figs. 4C, 4D). Only in the Pedregal population, we found a parasitoid wasp from the Ichneumonidae family parasitizing L. daturaphila larvae (Fig. S1A). In the Toluca and Tlaxiaca populations, Mesochorus sp., a hyperparasitoid wasp, emerged from fly cocoons (Figs. S1B, S1C).

Population variability in parasitoid infestation

We observed high and significant variability in the abundance of L. daturaphila egg and larval stages and their parasitoids across populations. At least one parasitoid was associated with each population.

Clutch size significantly varied across populations and years (GLM, p < 0.05, Fig. 5, Table S3). Tlaxiaca and Tzintzuntzan had the highest mean clutch sizes in 2018 and 2019, respectively (23.78 ± 12.03; 33.93 ± 9.35, Table S4). In 2018, 46.15% of the total egg clutches collected were parasitized, and in 2019, this increased to 51.58% (Fig. 6). Requena, Texcoco, Tzintzuntzan and Valsequillo had a significantly higher parasitism (Table S5). In the first year, Tzintzuntzan, Valsequillo, and Requena exhibited the highest parasitism rates (91.6%, 84.3%, and 83.3%, respectively). However, no parasitism was observed in Teotihuacan and Toluca (Fig. 6). In 2019, parasitism of egg clutches was total in the Pedregal, Tzintzuntzan, and Valsequillo populations, while Bernal and Teotihuacan were uninfested. In contrast to 2018, Toluca had 8.5% parasitism of egg clutches (Fig. 6).

Despite the high coverage of L. daturaphila egg clutches by E. lemae, not all eggs in a clutch were killed by parasitism; some completed their development or died from other causes. Egg mortality rates varied among populations. In 2018, Tzintzuntzan, Requena, and Valsequillo exhibited the highest mortality rates (67.25%, 61.67%, and 49.45%, respectively) (Fig. 7). Then in 2019, Pedregal and Requena had the highest egg mortality levels (96.8% and 93.03%, respectively). In Valsequillo, although E. lemae visited all clutches, only 49.45% of the eggs died from parasitism. Dolores and San Martín, the new censused populations for that year, had total parasitoidism rates of 34.42% and 22.01%, respectively (Fig. 7). In 2019, larger clutches of L. daturaphila were significantly preferred by E. lemae (GLM, p = 0.0254; Fig. 8, Table S6). Additionally, there was a significant positive relationship between the number of Lema´s eggs and the emerged parasitoid wasps per clutch in both years (Fig. 9, Table S7).

We detected Tachinidae flies in all populations of L. daturaphila except Pedregal. In some localities, the abundance of larvae was very low, but parasitism still occurred. Larvae were absent in the Tzintzuntzan population in 2018. The highest rates of larval parasitism were observed in Tlaxiaca, Texcoco, and Teotihuacan in both years (Fig. 10). The number of parasitoid flies showed a significant relationship with the number of larvae collected per population (GLM, p < 0.0001, Fig. 11, Table S8).

Analysis of environmental variables revealed divergent climatic conditions that influenced parasitoid presence and L. daturaphila populations. The first three principal components explained most of the variance (36.0%, 23.9%, and 20.76% respectively) (Table 1). Populations such as Pedregal, Tzintzuntzan, Valsequillo, and Requena, characterized by higher temperatures and precipitation, tended to have more parasitized egg clutches. Non-parasitized clutches were typically found at higher altitudes in populations such as Teotihuacan, Tlaxiaca, and Toluca. The number of emerged parasitoid wasps was negatively associated with the emergence of larvae in populations at higher altitudes and lower precipitation (Fig. 12).

Analysis of Lema daturaphila oral secretions

Four tropane alkaloids were detected in the oral secretions of larvae of L. daturaphila (Table 2). These compounds are also present in its host plant, D. stramonium (De-la-Cruz et al., 2020).

Parasitoid species

Accurate species identification is crucial for understanding ecological and evolutionary relationships (Smith et al., 2008). Despite being a ubiquitous and diverse group, parasitoids, especially micro-Hymenoptera and Tachinidae, are still under-described (Forbes et al., 2018; Dindo & Nakamura, 2018). Emersonella lemae parasitizes the eggs of chrysomelid beetles in America, particularly in the Neotropics. This is one of the two species of Emersonella reported in the Neartic Region (Alvarenga et al., 2015), primarily in the United States, and represents a new record for Mexico (Noyes, 2019).

Tachinidae flies, the second largest group of parasitoids after Hymenoptera, remain poorly understood in terms of their biology (Stireman, 2002; Dindo & Nakamura, 2018). Genera such as Patelloa and Pseudochaeta are found throughout America, primarily parasitizing Lepidoptera. Records of Patelloa leucaniae attacking Lema exist in the U.S. and Canada (Arnaud, 1978), but it is a new record for L. daturaphila in Mexico (Zetina et al., 2018). The genus Pseudochaeta contains 26 species mainly from the U.S., and both the genus and host are new records for Mexico (Arnaud, 1978; O’Hara & Wood, 2004). Species of the Winthemia genus are widely used as a biological control in economically important crops. In Mexico, there are species such as W. imitator, W. montana, or W. texana (Guimarães, 1972; Zetina et al., 2018), L. daturaphila is a new record as their host. Finally, two species of Heliodorus (H. cochisensis and H. vexillifer) are known from the U.S. (O’Hara & Wood, 2004), and both the genus and the host L. daturaphila represent new records for Mexico.

Previous studies in the Pedregal population reported a tachinid fly parasitizing L. daturaphila larvae (Cabrales-Vargas, 1991). In our study, we identified an undetermined Ichneumonidae wasp. In the Toluca population, we found a hyperparasitoid wasp from the genus Mesochorus, which has a cosmopolitan distribution. This genus belongs to the Mesochorinae family, known for being hyperparasitoids of Braconidae, Tachinidae, and Ichneumonidae (Dasch, 1974). This is the first record of a Mesochorus species parasitizing L. daturaphila parasitoids, making it a new record.

Population variability in parasitoid infestation

Before this study, there had been no reports of egg parasitoids attacking L. daturaphila in Mexican populations of D. stramonium. Emersonella lemae was especially abundant at Tzintzuntzan, Valsequillo, Pedregal, Requena and Texcoco during both years. High host egg availability and kairomone concentration could explain E. lemae’s preference for these sites (van Alphen, Bernstein & Driessen, 2003; Fatouros et al., 2008; Iranipour et al., 2020). Its high incidence corresponded with a scarcity of L. daturaphila larvae and adults, which were not the main herbivores in previous years (Castillo et al., 2014). In addition to low damage, these D. stramonium populations have shown negative selection for direct defenses, which are ineffective against L. daturaphila (Castillo et al., 2014). Since Emersonella lemae is known to be an effective regulator of L. daturaphila in other populations (Girault, 1916; Chittenden, 1924), D. stramonium could attract this parasitoid when direct defenses are no longer effective (Ballhorn et al., 2008).

It is evident that E. lemae also disperses effectively within Mexican L. daturaphila populations. It was particularly successful at higher temperatures and precipitation, but weather conditions could limit its reproductive success at higher altitudes (Weisser, Volkl & Hassell, 1997; Péré, Jactel & Kenis, 2013), as occurs in Toluca and Tlaxiaca populations. This divergent pattern may also reflect the dispersal abilities of the parasitoid wasp, which because of its small size probably moves passively through the wind and could be limited to access at higher altitudes (Fatouros et al., 2008).

Except for Pedregal, Tachinidae flies were found in all studied populations. Tlaxiaca, Toluca, and Teotihuacan were the most parasitized and historically have been dominated by the herbivory of L. daturaphila (Castillo et al., 2014). Tachinidae flies utilize long-range cues to locate their hosts, employing an efficient searching strategy (Hanyu et al., 2009; Dindo & Nakamura, 2018). Even at a low density, they could detect larvae of L. daturaphila in patches where E. lemae kill most eggs. This seems to be the case in populations with high levels of infested egg clutches. Parasitoids usually compete for the host, which serves as a common resource. Although several species could coexist using different development stages of the same host (as occurs with wasps and flies in L. daturaphila), eventually parasitoids that attack a stage will affect the densities of subsequent stages by increasing host mortality (Briggs, Nisbet & Murdoch, 1993). The decline of larval individuals and its negative relationship with the emergence of wasps was clear with statistical analysis, but it also was a continuous observation in the field. Despite the high number of eggs and the frequent presence of adult L. daturaphila in populations such as Valsequillo, Pedregal, and Requena, it was almost impossible to find larvae and consequently, parasitoid flies. These findings underscore the importance of E. lemae in regulating L. daturaphila populations and their impact on interactions established with other parasitoids. In this complex geographic scenario, some populations (i.e., Tzintzuntzan, Valsequillo, or Requena) could function as hotspots, where the interaction with E. lemae shows strong reciprocal selection, while others are dominated by the interaction with parasitoid flies (i.e., Toluca or Tlaxiaca) (Thompson, 2005).

Analysis of oral secretions

In D. stramonium populations, it is usual to observe tachinid flies approaching L. daturaphila’s larvae but avoiding oral secretions from the head. The chemical composition of these excretions can vary with the insect food and is important for all members of a tritrophic system (Alborn et al., 1997; Sword, 2001; Grant, 2006; Zvereva & Kozlov, 2016). They can function as elicitors for the plant and may be essential to differentiate between mechanical and insect damage (Neveu et al., 2002). However, they can also aid in insect digestion, immunity, the suppression of plant defenses (Rivera-Vega et al., 2017) or be a chemical mechanism of anti-predation for herbivores (Pasteels, Braekman & Daloze, 1988; Karban & Agrawal, 2002).

Oral secretions can contain toxic compounds that insects get from food (Calcagno et al., 2004). Some specialist herbivores release this gut content when they are attacked, affecting natural enemies with plant metabolites (Rowell-Rahier & Pasteels, 1992; Grant, 2006). Because trophic specialization of L. daturaphila to D. stramonium and its resistance to many alkaloids (De-la-Cruz et al., 2020), larvae could use the substances present in its host plant to complement the defense behavior they perform before potential attacks by flies (Omer-Cooper & Miles, 1951). While there may be an intriguing relationship between specialization and the defenses of L. daturaphila, it is crucial to experimentally determine which role oral secretions play in the tri-trophic interaction.