Hidden Host Mortality from an Introduced Parasitoid: Conventional and Molecular Evaluation of Non-Target Risk

Abstract

Hidden trophic interactions are important in understanding food web ecology and evaluating the ecological risks and benefits associated with the introduction of exotic natural enemies in classical biological control programs. Although non-target risk is typically evaluated based on evidence of successful parasitism, parasitoid-induced host mortality not resulting in visible evidence of parasitism (i.e., nonreproductive effects) is often overlooked. The adventive establishment of Trissolcus japonicus, an exotic parasitoid of the introduced stink bug Halyomorpha halys, provides an opportunity to investigate the total impact of this parasitoid on target and non-target hosts in the field. We developed a new methodology to measure nonreproductive effects in this system, involving a species-specific diagnostic PCR assay for T. japonicus. We applied this methodology to field-deployed eggs of four pentatomid species, coupled with traditional rearing techniques. Nonreproductive effects were responsible for the mortality of an additional 5.6% of H. halys eggs due to T. japonicus, and were even more substantial in some of the non-target species (5.4-43.2%). The observed hidden mortality of native non-target species from an introduced parasitoid could change predictions about direct and indirect ecological interactions and the efficacy of biological control of the target pest.

Keywords: Halyomorpha halys; Trissolcus japonicus; hidden mortality; indirect effects; molecular forensics; non-target; nonreproductive effects; parasitoid.

Figure 1

Occurrence of reproductive and nonreproductive effects during the study period. The colored portions in each bar represent the number of eggs in individual egg masses for each of four pentatomid species (Chinavia hilaris, Euschistus conspersus, Halyomorpha halys, and Podisus maculiventris) on a given date.

Figure 2

Reproductive and nonreproductive impacts of T. japonicus on four pentatomid host eggs. Reproductive effects: F = 15.79, p < 0.001; nonreproductive effects (upper estimate): F = 55.84, p < 0.001; nonreproductive effects (lower estimate): F = 48.15, p < 0.001; total impact (upper estimate of nonreproductive effects): F = 57.46, p < 0.001; total impact (lower estimate of nonreproductive effects): F = 49.65, p < 0.001; df all analyses 3, 80. Upper case letters above bars are for total impact (reproductive + nonreproductive) using the upper estimate of the nonreproductive effects, and lower case letters are for total impact using the lower estimate of nonreproductive effects. Numbers beneath the species names are the number of eggs evaluated. Error bars represent the standard error of the mean of total parasitoid impact (reproductive + nonreproductive).

Figure 3

Simulated direct and indirect effects of T. japonicus on non-target and target stink bug populations in the absence and presence of nonreproductive effects (NREs), based on the population dynamics model of Kaser et al. (2018), and parameterized with data from the current study. See Table 2 for parameter values used. Upper estimates of nonreproductive effects (NRE_U) was used, but similar results were obtained using lower estimates (NRE_L) (Figure 4). Simulations were performed for three two-host, one-parasitoid systems consisting of T. japonicus, the target host H. halys, and one of three non-target species. For each host species pair, simulated equilibrium population densities of the non-target and target stink bug in the presence versus absence of measured nonreproductive effects, taking into account both direct and indirect (i.e., via the other host species) effects of T. japonicus, are shown as barplots. Simulated target and non-target population densities in the presence and absence of nonreproductive effects considering only direct effects of T. japonicus (i.e., in the absence of the other host species) are shown as points. When grey bars (including NERs) show values lower than paired white bars (not including NREs), the combined direct and indirect impact of nonreproductive effects in the system is to decrease stink bug population levels, and vice versa. When points are above bars, the host populations are being reduced as a consequence of the presence of the other host species in the system. When points are below bars, the host populations are higher as a result of the other host species being present in the system.

Figure 4

Simulated direct and indirect effects of T. japonicus on non-target and target stink bug populations in the absence and presence of nonreproductive effects (NREs), using lower estimates (NREL). See Table 3 for parameter values used. Simulations were done for three two-host, one-parasitoid systems consisting of T. japonicus, the target host H. halys, and one of three non-target species. For each host species pair, simulated equilibrium population densities of the non-target and target stink bug in the presence versus absence of measured non-reproductive effects, taking into account both direct and indirect (i.e., via the other host species) effects of T. japonicus, are shown as barplots. Simulated target and non-target population densities in the presence and absence of non-reproductive effects considering only direct effects of T. japonicus (i.e., in the absence of the other host species) are shown as points. When grey bars show values lower than paired white bars, the combined direct and indirect impact of non-reproductive effects in the system is to decrease stink bug population levels, and vice versa. When points are above bars, the host populations are being reduced as a consequence of the presence of the other host species in the system. When points are below bars, the host populations are higher as a result of the other host species being present in the system.

Figure 5

Simulated direct and indirect effects of T. japonicus on non-target and target stink bug populations if hidden host mortality (due to nonreproductive effects, NREs) is excluded, or is included as attack resulting in parasitoid reproduction (reproductive effects) (to contrast with Figure 2 above where the same amount of mortality is added as nonreproductive effects). See Table 4 for parameter values used. Simulations were done for three two-host, one-parasitoid systems consisting of T. japonicus, the target host H. halys, and one of three non-target species. For each host species pair, simulated equilibrium population densities of the non-target and target stink bug in the presence versus absence of additional mortality due to parasitism, taking into account both direct and indirect (i.e., via the other host species) effects of T. japonicus, are shown as barplots. Simulated target and non-target population densities in the presence and absence of additional mortality due to parasitism considering only direct effects of T. japonicus (i.e., in the absence of the other host species) are shown as points. When grey bars show values lower than paired white bars, the combined direct and indirect impact of additional host mortality in the system is to decrease stink bug population levels, and vice versa. When points are above bars, the host populations are being reduced as a consequence of the presence of the other host species in the system. When points are below bars, the host populations are higher as a result of the other host species being present in the system.

Figure 6

Modified Levins’ Diagrams showing examples of hypothesized indirect and direct effects of Trissolcus japonicus on target and non-target stink bugs, and indirect interactions between stink bug species pairs, based on predictions from population dynamics models (see Figure 3). Dots at the end of lines indicate a negative effect on the stink bug species shown; arrows indicate a positive effect. Solid lines are direct effects and dashed lines are indirect effects. For all effects of T. japonicus on stink bugs, the length of the line indicates magnitude of the interaction. The gray portion of the line corresponds to the nonreproductive contributions and the black portion corresponds to the reproductive effect. Where the dotted line is missing in one direction, no indirect effect was observed. Effects of stink bugs on T. japonicus populations were not evaluated in this study, and are not shown here. (A) Halyomorpha halys (target) and Podisus maculiventris (non-target); apparent amensalism. (B) Halyomorpha halys (target) and Euschistus conspersus (non-target); apparent predation. (C) Halyomorpha halys (target) and Chinavia hilaris (non-target); apparent amensalism.