Introduction

Interferences between species occurs when more than one species attack the same host species, affecting the distribution and population dynamics of the same (Xu et al. 2013; Holt and Bonsall 2017). These interferences between species arise during adult host searching or during the development of the immature stages of the parasitoids inside the same host (Bográn et al. 2002; Cusumano et al. 2013, 2016; Xu et al. 2013). The outcomes of such interference can affect the ecosystem in different ways such as the reduction or extinction of a specific species, which affects pest control in terms of the depletion, displacement, or extinction of natural enemies (Reitz and Trumble 2002; Wang et al. 2008; Feng et al. 2015; Tan et al. 2016). For example, the ectoparasitoid, Spathius agrili Yang (Hymenoptera: Braconidae) eggs do not develop on Agrilus planipennis (Fairmaire) (Coleoptera: Buprestidae) larvae earlier parasitised by the endoparasitoid, Tetrastichus planipennisi Yang (Hymenoptera: Eulophidae) (Ulyshen et al. 2010). On that same pulse beetle, the endoparasitoid, Dinarmus basalis (Rond.) (Hymenoptera: Pteromalidae) differentiated larvae previously parasitised by the ectoparasitoid, Eupelmus vuilleti Crawford (Hymenoptera: Eupelmidae) (van Alebeek et al. 1993). It is therefore important to understand the possible effects that different parasitoids can have when planning a biological control program that involves more than one parasitoid species (Mills 1992; Cusumano et al. 2016).

Since the invasion of P. absoluta, numerous parasitoids associated with this pest were discovered in invaded countries (Tropea Garzia et al. 2012; Zappalà et al. 2012, 2013; Naselli et al. 2017; Mansour et al. 2018; Ferracini et al. 2019; Salas Gervassio et al. 2019). A combination of natural enemies for control of this pest has been reported in various studies. The parasitoid Stenomesius japonicus Ashamed (Hymenoptera: Eulophidae) released together with the predator Macrolophus pygmaeus Rambur (Hemiptera: Miridae) improved control of P. absoluta Chailleux et al. (2017). The predator Nesidiocoris tenuis (Reuter) (Hemiptera: Miridae) in combination with the egg parasitoid, Trichogramma brassicae Bezdenko (Hymenoptera: Trichogrammatidae) showed the highest tomato yield in a trial compared to where each of the natural enemies was alone (Mirhosseini et al. 2019). When evaluated in combination, females of both the ectoparasitoid, Dineulophus phthorimaeae De Santis (Hymenoptera: Eulophidae) and the endoparasitoid, Pseudapanteles dignus (Muesebeck) (Hymenoptera: Braconidae) changed their searching behaviour for P. absoluta in the presence of the competing species. The efficacy of P. dignus was, however, not affected by the presence of the ectoparasitoid and it achieved the same parasitism rate whether the competitor was present or absent (Savino et al. 2016).

Bracon nigricans Szépligeti (Hymenoptera: Braconidae) is a gregarious, generalist idiobiont larval ectoparasitoid (Yu and Actherberg 2010), recorded in P. absoluta in several invaded areas. For example, its association with P. absoluta was recorded in Italy (Zappalà et al. 2012), Jordan (Al-Jboory et al. 2012), Spain (Gabarra et al. 2013) and Sudan (Idriss et al. 2018). This parasitoid was amongst the three most abundant parasitoids recorded in Italy. Bracon nigricans was reported to co-occur with other larval parasitoid species such as Diadegma pulchripes (Hymenoptera: Ichneumonidae), Elachertus inunctus Nees, Necremnus sp., Neochrysocharis formosa (Hymenoptera: Eulophidae) and Elasmus sp. (Hymenoptera: Elasmidae). In Spain, B. nigricans was reported to co-occur with Hockeria unicolor Walker (Hymenoptera: Chalcididae), Pnigalio cristatus (Ratzeburg), and Neochrysocharis formosa (Westwood) (Hymenoptera: Braconidae), while a Cotesia sp., and Dolichogenidea litae (Nixon, 1972) (Hymenoptera: Braconidae) has also been recorded in a nearby locality two years before (Gabarra et al. 2013). Bracon nigricans has also recently been recorded in Kenya, and several areas in Africa are suitable for the persistence of this parasitoid (Mama Sambo et al. 2022b). Approximately 21% parasitism of T. absoluta by B. nigricans was reported from open field tomatoes in central Kirinyaga county, Kenya where the exotic D. gelechiidivoris was released (Shiraku 2020; Mama Sambo et al. 2022b).

Dolichogenidea gelechiidivoris is a specialist solitary endoparasitoid of some Gelechiidae species, including P. absoluta (Bajonero et al. 2008; Mujica and Kroschel 2017; Aigbedion-Atalor et al. 2020; Mama Sambo et al. 2022c). The parasitism rate by this parasitoid under laboratory conditions varied between 55 and 87% (Bajonero et al. 2008; Aigbedion-Atalor et al. 2020; Mama Sambo et al. 2022a, c), depending on the host and parasitoid density, as well as P. absoluta larval stage. Dolichogenidea gelechiidivoris was established in Chile 10 years after its release (Desneux et al. 2010). In Kenya, less than 5% parasitism by D. gelechiidivoris on P. absoluta larvae were recorded in open fields, five months after the release of D. gelechiidivoris (Mama Sambo et al. unpublished data). However, D. gelechiidivoris has also been recovered from non-release areas in Europe and Africa (Krache et al. 2021; Denis et al. 2022).

Bracon nigricans hasa very high ability to kill its host (Biondi et al. 2013; Becchimanzi et al. 2017, 2020). This parasitoid prefers 4th-instar P. absoluta larvae (Idriss et al. 2018), while D. gelechiidivoris prefers 1st and 2nd-instar larvae of this host (Aigbedion-Atalor et al. 2020). Parasitism by D. gelechiidivoris in Kenya was reported to be very low in the release area, where B. nigricans was recovered (Shiraku 2020; Mama Sambo et al. 2022b). The aim of this study was to investigate the population dynamics of P. absoluta and two of its parasitoid species, D. gelechiidivoris and B. nigricans in laboratory bioassays, when present alone or in combination.

Materials and methods

Host plants

Tomato (cv. Moneymaker) seedlings were planted in plastic pots (14 cm × 14 cm × 9 cm) containing soil mixed with goat manure. The seedlings were grown in a greenhouse following the procedures described by Mama Sambo et al. (2022c).

Insects rearing

Phthorimaea absoluta

Tomato plants with P. absoluta larvae and eggs were collected from farmers’ plots in the Kirinyaga County (00°37′196'' S, 37°22′615'' E, Elevation ≈ 1200 msl and 00°37′922'' S, 37°22′794'' E, Elevation ≈ 1200), Kenya to initiate a colony. The infested plants were incubated and P. absoluta obtained from these plants was reared on healthy tomato plants following the procedure described by Mama Sambo et al. (2022c).

Dolichogenidea gelechiidivoris

Dolichogenidea gelechiidivoris cocoons were kept in a Perspex cage (40 cm × 20 cm × 40 cm) with 80% honey solution droplets provided on the inside of the top of the cage for the emerged parasitoids to feed on. The adults were kept isolated in the Animal Rearing and Containment Unit (ARCU) at icipe and reared on P. absoluta larvae according to the protocol described by Mama Sambo et al. (2022c).

Bracon nigricans

The B. nigricans colony was initiated from parasitoids that emerged from infested tomato plant material collected from Mwea (00°37′196'' S, 37°22′615'' E, Elevation ≈ 1200 msl and 00°37′922'' S, 37°22′794'' E, Elevation ≈ 1200 msl), Kirinyaga, Kenya. The infested leaves were kept in a transparent plastic lunch box. Upon adult parasitoid emergence, insects were aspirated into a clean Perspex cage (40 cm × 20 cm × 50 cm) and provided with 80% honey solution droplets applied on the top of the cage. The insects were maintained under ambient laboratory conditions (25 ± 1 °C, 70 ± 5 RH, and 12:12 L:D of photoperiod) in the Animal Rearing and Containment Unit (ARCU) at icipe. Third instar P. absoluta larvae in tomato plant leaves were exposed to B. nigricans adults for 48 h. Thereafter, the tomato leaves were transferred to another Perspex cage until B. nigricans and P. absoluta emerged. Two days-old females (F3) were used in this experiment.

Experimental setup

The interaction between D. gelechiidivoris and B. nigricans was investigated under laboratory conditions (25 ± 1 °C, 70 ± 5 RH, and 12:12 L:D of photoperiod). The experiment was conducted in cages (30 cm × 30 cm × 30 cm) and consisted of four treatments. Initially, one uninfested potted tomato plant (3 weeks old) was placed in each cage. Ten newly emerged naïve P. absoluta male–female moth pairs were released per cage. Drops of an 80% honey solution were streaked onto the inner top part of the cage as food for the moths. Four treatments were applied (Table 1). Each treatment was replicated ten times. For each treatment, an additional potted plant was placed into each cage at three-day intervals for the entire duration of the study and supplemented daily with fresh tomato leaves to ensure that enough food was available for P. absoluta larvae.

Table 1 Experimental scenarios used to assess the effect of the co-occurrence of D. gelechiidivoris and B. nigricans
Full size table

Data collection

Parasitism by the two species

To determine the level of parasitism, ten P. absoluta larvae were removed from plants in each cage, two weeks after release of B. nigricans and dissected under a stereomicroscope (Leica EZ4D digital stereomicroscope; Leica Microsystems, Heerbrugg, Switzerland). Additionally, dead P. absoluta larvae inside mines were recorded in each of the treatments.

Population dynamics of the different insect species

The number of P. absoluta mines and eggs was determined by counting the number of eggs and mines in each leaflet on the plant every week from the third-week of monitoring. In addition, the number of D. gelechiidivoris and B. nigricans parasitoids that emerged were counted. The experiment was conducted over a period of 10 weeks.

Data analysis

A Generalised Linear Model (GLM) with a negative binomial was used to determine the difference in percentage parasitism by D. gelechiidivoris and B. nigricans, of dissected larvae, as well as the number of larvae killed by B. nigricans. The numbers of D. gelechiidivoris and B. nigricans adults and mines per treatment were analysed using a mixed-effects model with repeated measures, using lmer function under the lme4 package. All the data were analysed in R (R Core Team 2018).

Results

Parasitism by the two species

The level of D. gelechiidivoris parasitism varied significantly, depending on the presence or the absence of B. nigricans (F1, 18 = 26.69, P ˂ 0.001). Significantly more larvae were parasitised by D. gelechiidivoris, when it was present alone, compared to the level of parasitism when P. absoluta larvae were exposed to both D. gelechiidivoris and B. nigricans (Fig. 1a). However, no difference in parasitism by B. nigricans was observed regardless of the presence or absence of D. gelechiidivoris (F1, 18 = 1, P = 0.33). Similarly, the release of B. nigricans after D. gelechiidivoris was already present did not increase P. absoluta larval mortality (F1, 18 = 0.35, P = 0.56). Although the maximum parasitism by B. nigricans (10%) (Fig. 1b) was lower than that by D. gelechiidivoris (73%) (Fig. 1a), its parasitism was not affected by the presence of D. gelechiidivoris (Fig. 1b).

Fig. 1
figure 1

a Mean percentage (± SE) P. absoluta larvae parasitised by D. gelechiidivoris and b Mean percentage P. absoluta larvae parasitised and killed by B. nigricans, when released as a standalone treatment as well as in combination with a competitor. Bars capped with different upper or lower letters are not significantly different (Tukey's HSD, p < 0.05)

Full size image

Population dynamics of the different insect species

The treatments with an absolute t-value > 2 affected the parameters significantly (Table 2). The D. gelechiidivoris population was negatively affected by the association with B. nigricans (t = 5.70) while no significant effect on the B. nigricans population was observed by the co-occurrence of the two parasitoids species (Table 2). Where the D. gelechiidivoris population was kept separate, the number of adults present was high, with an average of 156 parasitoids five weeks after introduction. In co-occurrence with B. nigricans only 8 D. gelechiidivoris individuals survived at week 5 (Fig. 2). The number of P. absoluta progeny expressed as the number of mines and eggs, was significantly higher compared to all other treatments with parasitoids present in the respective combinations, five weeks after release of the P. absoluta moths (t = 2.72) (Table 2). At week four, the pest level significantly decreased where both B. nigricans occurred alone and where the combination of D. gelechiidivoris and B. nigricans occurred (Fig. 3). From week 5, significantly lower pest levels occurred in all treatments, except for the control (Fig. 3).

Table 2 Parameter estimates for the effect of parasitoid combinations in terms of D. gelechiidivoris and B. nigricans population growth and P. absoluta infestation
Full size table
Fig. 2
figure 2

Mean number (± SE) of a D. gelechiidivoris and b B. nigricans adults, per week in the respective treatments over time

Full size image
Fig. 3
figure 3

Mean number (± SE) of P. absoluta progeny per week in the respective treatments over time

Full size image

Discussion

The efficacy of control of P. absoluta by two parasitoids, viz. the imported endoparasitoid, D. gelechiidivoris, and the endogenous ectoparasitoid, B. nigricans was confirmed in this laboratory study. The negative impact of the co-occurrence of these two species on D. gelechiidivoris was also demonstrated. Multiple parasitism by parasitoid species frequently occur in the absence of interspecific host discrimination (van Alphen and Visser 1990; Hassell and Godfray 1994), and cause competition by the offspring of parasitoid species within a host. It also affects the population growth of a specific parasitoid if, as in most cases with ectoparasitoids, the female kills the host (Biondi et al. 2013; Chailleux et al. 2014; Idriss et al. 2018). Consequently, the chances of survival of a host are low when initially parasitized by an endoparasitoid and subsequently by an ectoparasitic, resulting in the larvae of the endoparasitoid being instantly killed together with the host (Rosenheim et al. 1995). This might be the case with B. nigricans, an ectoparasitoid (Biondi et al. 2013; Becchimanzi et al. 2017) when present in co-occurrence with the endoparasitoid, D. gelechiidivoris (Fernandez-Triana et al. 2020). Bracon nigricans is abundant in Kenya (Mama Sambo et al. 2022b), where D. gelechiidivoris was also released (Shiraku 2020). The level of parasitism by D. gelechiidivoris in the area investigated in this study, was low in comparison to parasitism levels when released in a greenhouse in the absence of B. nigricans (Mama Sambo et al. Unpublished data).

Phthorimaea absoluta larvae were effectively controlled by D. gelechiidivoris when exposed to this species alone (parasitism level of 73%). It was in contrast to the level of parasitism by B. nigricans of only 10%, when exposed under the same conditions. Previously, parasitism levels of 87% (Mama Sambo et al. 2022a), 58% (Mama Sambo et al. 2022c), and 55% Aigbedion-Atalor et al. (2020) were reported when 1st -instar P. absoluta larvae were exposed to female D. gelechiidivoris for 24 h. Biondi et al. (2013) and Idriss et al. (2018) reported that B. nigricans parasitised 4th-instar P. absoluta larvae and that only one or two parasitoids emerged from these larvae per day. Parasitism by B. nigricans, therefore, appears to be insignificant compared to parasitism by D. gelechiidivoris. However, Idriss et al. (2018), reported 54% parasitism of 4th instar larvae under laboratory conditions, and Mama Sambo et al. (2022b), up to 21% parasitism of P. absoluta larvae by this parasitoid species, in recent field studies. It emphasises the effect of other factors such as the host: parasitoid ratio and host stage available that can significantly affect the level of parasitism. A reduction of approximately 50% in D. gelechiidivoris parasitism was recorded in this study when this species co-occurred with B. nigricans. A highly negative impact on the population growth and resultant parasitism level and control of P. absoluta larvae by D. gelechiidivoris can therefore be expected in areas where B. nigricans is abundant.

Bracon nigricans did not discriminate the larvae already parasitised by the endoparasitoid, D. gelechiidivoris. Ectoparasitoids are in general also better competitors than endoparasitoids, and the venom from the females often paralyze the immature stages of the endoparasitoid already present in the host, as well as the host itself (Harvey et al. 2013). However, Savino et al. (2016) reported that the ectoparasitoid, D. phthorimaeae spent more time in general host searching for P. absoluta when in competition with the endoparasitoid P. dignus, while the presence of the ectoparasite did not have any effect in this regard on P. dignus. From the perspective of insect biological control, a superior parasitoid species must have a shorter developmental time, greater searching capability, high host specificity, a positive correlation with host density, good synchronization between the populations of the host and parasitoid as well as a good dispersal ability (Haye et al. 2008). The approximate female developmental time and longevity of B. nigricans are 12 days and between 13–24 at 25 °C respectively (Biondi et al. 2013; Idriss et al. 2018) while at the same temperature, the developmental time of D. gelechiidivoris is approximately between 19–25 days and its longevity, between to 59 days (Bajonero et al. 2008; Aigbedion-Atalor et al. 2020; Mama Sambo et al. 2022c).) The development time of B. nigricans is therefore shorter than that of D. gelechiidivoris, but it lives longer than D. gelechiidivoris. The number of B. nigricans progeny per day was reported to be fewer than five per female (Biondi et al. 2013; Idriss et al. 2018) while the number of D. gelechiidivoris progeny was reported to be host-density dependent (Mama Sambo et al. 2022c). Bracon nigricans is a generalist parasitoid, known to attack different lepidopteran families (Loni et al. 2016; Becchimanzi et al. 2017; Aigbedion-Atalor et al. 2019; Mama Sambo et al. 2022b), while the only family known to be attacked by D. gelechiidivoris to date, is Gelechiidae (Bajonero et al. 2008; Mujica and Kroschel 2017; Aigbedion-Atalor et al. 2020, 2021; Mama Sambo et al. 2022c).

It should, however, be noted that 32% killing of host larvae could have interfered with the population growth of a competing parasitoid as well as with that of the pest population. Several studies documented the performance of B. nigricans on host larval killing. For example, Biondi et al. (2013) reported the killing of approximately 50% of mature P. absoluta larvae by B. nigricans, while Idriss et al. (2018) found 55% mortality of 3rd- and 40% of 4th-instar P. absoluta larvae. The parasitism rates by D. gelechiidivoris might have been consequently affected by the population size of B. nigricans. However, predation by N. tenuis of 1st-instar P. absoluta larvae was not found to affect the progeny production and adult emergence of D. gelechiidivoris (Aigbedion-Atalor et al. 2021), although N. tenuis prefers the egg stage of P. absoluta (Sylla et al. 2016).

The number of P. absoluta mines drastically decreased from the second week after the parasitoids were introduced into the cages where P. absoluta larvae were present, both as standalone treatments as well as in combination, to represent the co-occurrence of the two species. Several studies reported an additional effect of more than one natural enemy on the control of P. absoluta. For example, P. absoluta larval parasitism by the parasitoids D. phthorimaeae and P. dignus resulted in high P. absoluta mortality in the field (Luna et al. 2015). A complementary effect in control of P. absoluta was also reported for the specialist, S. japonicus, and omnivorous natural enemy M. pygmaeus (Chailleux et al. 2017). Combined use of N. tenuis and T. achaeae also resulted in better control compared to N. tenuis alone (Calvo et al. 2012). The combined use of the parasitoid T. achaeae, and the predator M. caliginosus also increased P. absoluta control (Kortam et al. 2014). However, an antagonistic effect has been reported between Spathius agrili Yang (Hymenoptera: Braconidae) and Agrilus planipennis Fairmaire (Coleoptera: Buprestidae) with larvae from the latter being paralyzed after parasitisation. It served as an indicator to Tetrastichus planipennisi Yang (Hymenoptera: Eulophidae) to discriminate between S. agrili parasitized and non-parasitized larvae (Yang et al. 2013). Spathius agrili cannot detect larvae parasitised by T. planipennisi, which is disadvantageous to the parasitoid, since its progeny cannot survive on a host previously parasitized by T. planipennisi (Ulyshen et al. 2010).

Conclusion

The endoparasitoid D. gelechiidivoris and the ectoparasitoid, B. nigricans can co-exist by exploiting the same P. absoluta larvae differently. However, the presence of B. nigricans negatively affected the population growth of D. gelechiidivoris while the presence of D. gelechiidivoris did not affect B. nigricans population growth. Ecological niche differences created by biotic factors such as host plant, host preference, and abiotic factors, specifically climate, may drive the augmentation of these two parasitoids for control of P. absoluta. Additionally, since B. nigricans is a generalist parasitoid of several lepidopteran species such as Spodoptera littoralis Boisduval (Lepidoptera Noctuidae) (Becchimanzi et al. 2017) and Lobesia botrana (Denis & Schiffermüller) (Lepidoptera, Tortricidae) (Loni et al. 2016), a minimal effect is expected on the introduced D. gelechiidivoris under field conditions. However, the host plant species and host pest species preferences of B. nigricans should be studied to determine the optimal conditions for the co-existence of the two P. absoluta parasitoids.

Fundings

This research was supported by the following organizations: Biovision Foundation grant number BV DPP-012/2019–2022, Norwegian Agency for Development Cooperation-(NORAD), the section for research, innovation and higher education grant number RAF-3058 KEN-18/0005, Scaling-up Climate-Smart Pest Management Approaches for Enhanced Maize and Tomato Systems Productivity in Eastern Africa (SCLAMP) grant number 81250394, the Swedish International Development Cooperation Agency (Sida); the Swiss Agency for Development and Cooperation (SDC); the Federal Democratic Republic of Ethiopia; and the Government of the Republic of Kenya. The first author was supported by the German Academic Exchange Service (DAAD) through the African Regional Postgraduate Programme in Insect Science (ARPPIS) tenable at icipe. The views expressed herein do not necessarily reflect the official opinion of the donors.