- Top of page
- Materials and methods
Soil faunal biodiversity is an important determinant of key ecosystem functions (Hendrix et al. 1986; Moore & de Ruiter 1991; Lavelle et al. 1992). Control of insect pests can influence the densities of target and non-target organisms and the associated indirect interactions in the soil food web, potentially leading to a reduction in species richness and diversity in native communities. Recent reviews of the ecological consequences of the introduction of biological control agents (Follet & Duan 2000; Strong & Pemberton 2000; Wajnberg, Scott & Quimby 2001) suggest that biological control carries risks as well as benefits. These reviews emphasize the need for understanding any collateral damage to native fauna resulting from biological control introductions. Most studies on non-target effects of biological control introductions have been focused on above-ground flora and fauna with emphasis on direct (parasite–host or predator–prey) relationships. Little is known about the effects of introduction or augmentation of biological control agents on the diversity of native fauna in below-ground food webs. We addressed this issue for the case of entomopathogenic nematodes, considering the response of the naturally occurring nematode community in a turfgrass ecosystem. Further, we compared the impact of entomopathogenic nematodes on the soil nematode community with that of trichlorfon, a commonly used insecticide in turfgrass.
Nematodes are ubiquitous soil fauna that interact in ecosystems directly as herbivores on plants and indirectly as consumers of microflora and fauna, thus playing a significant role in regulating primary production, predation, decomposition of organic matter and nutrient cycling (Coleman, Cole & Elliott 1984; Ingham et al. 1985; Griffiths 1990). Nematodes possess attributes that make them useful ecological indicators (Bernard 1992; Ritz & Trudgill 1999). Analyses to determine the effect of agricultural management practices on nematode community structure and function are generally based on nematode species, generic or trophic group abundance, diversity indices and maturity indices (Heip, Herman & Soetaert 1988; Ludwig & Reynolds 1988; Bongers 1990; Yeates et al. 1993). Nematode community indices have been used for monitoring the changes in both natural ecosystems and agroecosystems induced by a variety of disturbances (Semoiloff 1987; Wasilewska 1989; Bongers, Alkemade & Yeates 1991; Ettema & Bongers 1993; Freckman & Ettema 1993; Ferris, Venette & Lau 1996; Yeates, Wardle & Watson 1999).
Entomopathogenic nematodes are currently marketed world-wide for the biological control of insect pests (Grewal & Georgis 1998). Wide host range, high efficacy, lack of mammalian toxicity and the availability of techniques for economic mass production have led to the rapid increase in the use of these biological control agents in recent years (Grewal & Georgis 1998). Entomopathogenic nematodes in the genera Steinernema Travassos and Heterorhabditis Poinar (Nematoda: Steinernematidae and Heterorhabditidae) are lethal insect parasites (Kaya & Gaugler 1993). The non-feeding third-stage infective juveniles penetrate into the haemocoel of host insects through natural openings and release symbiotic bacteria (Xenorhabdus spp. Thomas and Poinar for Steinernematidae, and Photorhabdus spp. Boemere et al. for Heterorhabditidae). Toxins produced by the developing nematodes (Burman 1982; Ehlers, Wulff & Peters 1997) and bacteria (Dunphy & Webster 1988; Bowen et al. 1998) cause septicaemia and kill the insect host usually within 48 h of infection. Nematodes complete two to three generations inside the host. When the host cadaver is consumed, the next group of infective juveniles is produced, which leaves the cadaver in search of new hosts in the soil.
The susceptibility of non-target organisms to entomopathogenic nematodes has been tested in several laboratory and glasshouse experiments, but only a few studies have considered their impact on non-target soil fauna under field conditions. No significant adverse effects of entomopathogenic nematodes were observed on populations of collembolans and mites (Ishibashi et al. 1987; Georgis, Kaya & Gaugler 1991) or non-target insects in the families Carabidae, Histiridae, Staphylinidae and Gryllidae (Georgis, Kaya & Gaugler 1991; Koch & Bathon 1993) under field conditions. Some field studies have indicated that inundative applications of entomopathogenic nematodes suppress populations of plant-parasitic nematodes in soil (Smitley, Warner & Bird 1992; Grewal et al. 1997), but little is known about their impact on the structure and function of the rest of the nematode community.
In this study we tested the following hypotheses: (i) inundative application of entomopathogenic nematodes changes the structure of the nematode community in soil; and (ii) entomopathogenic nematodes and chemical insecticides differ in their impact on the soil nematode community. We addressed these hypotheses by using nematode community indices to quantify the changes in the composition of the soil nematode fauna in response to inundative application of entomopathogenic nematodes and a chemical insecticide.
- Top of page
- Materials and methods
Our results indicate that inundative application of entomopathogenic nematodes changes the structure of the nematode community in a turfgrass ecosystem. Total nematode abundance significantly decreased in all the treatments relative to the untreated control, reflecting differences that could be attributed to the disturbance induced by pest control treatments. The abundance of plant-parasitic nematodes was significantly reduced in all the entomopathogenic nematode treatments while the abundance of free-living nematodes was not affected. In contrast with the nematode treatments, trichlorfon reduced the abundance of both plant-parasitic and free-living nematodes (including bacterial feeders, fungal feeders, predators and omnivores). These results agree with earlier observations that free-living nematodes were relatively more sensitive to chemical pesticides compared with plant-parasitic nematodes (Ishibashi, Kondo & Ito 1983; Smolik 1983; Yardim & Edwards 1998).
The entomopathogenic nematodes considered as a group significantly reduced the species richness and diversity, and maturity of the nematode community, in soil by reducing the number and abundance of plant-parasitic nematode genera. However, the free-living nematode groups in soil remained unaffected by the entomopathogenic nematode treatments. In contrast to the entomopathogenic nematodes, trichlorfon reduced the species richness and diversity of the nematode community by adversely affecting both plant-parasitic and free-living nematode genera, indicating the differential response of nematode trophic groups to biological and chemical pest control treatments. None of the experimentally applied species were recovered 30 or 60 days after the treatment. This may be because of the extremely low population densities compared with the other nematodes present in the soil (Wasilewska 1979). Previous studies have also shown that populations of entomopathogenic nematodes usually decline to below detectable levels within 4 weeks of inundative application to the soil (Georgis 1992; Smits 1996). Therefore, it is less likely that these species will be recovered in the soil samples collected 30 or 60 days after the application.
Nematode diversity indices based on all the nematodes are considered to be indicators of ecosystem status and are commonly used for comparing nematode communities in ecosystems of varying human intervention (Freckman & Ettema 1993; McSorley 1997; Yeates, Wardle & Watson 1999). The nematode community in the soil is composed of plant-parasitic and free-living nematode trophic groups that play mutually contrasting roles in ecosystem functioning and often differ in their response to disturbances in the ecosystem (Bongers 1990). Therefore, diversity indices based on total nematode taxa alone may not be sufficient to assess the impact of disturbance on the nematode community. For instance, in this study both insecticide and non-native entomopathogenic nematode treatments had lower values of total nematode abundance and diversity indices but they differed in abundance and diversity within plant-parasitic and free-living nematode groups. We suggest that measuring changes in trophic group composition and diversity within plant-parasitic and free-living nematode groups in addition to the indices based on total nematode taxa would yield a better indication of how management practices impact upon the structure and function of the soil nematode community.
Our results are the first to indicate selective suppression of plant-parasitic nematodes by the entomopathogenic nematodes H. bacteriophora and H. indica, with no adverse effect on free-living nematodes under field conditions. Suppression of plant-parasitic nematodes in response to the application of entomopathogenic nematodes Steinernema spp. has been documented in several glasshouse and field studies (Bird & Bird 1986; Ishibashi & Kondo 1986; Ishibashi & Choi 1991; Smitley, Warner & Bird 1992; Grewal et al. 1997; Perry et al. 1998; Lewis, Grewal & Sardanelli 2000). Contrary to conclusions based on diversity and maturity indices, it could be assumed that plant production would be enhanced by the entomopathogenic nematodes due to the lower number, abundance and c-p-values of plant-parasitic nematode taxa in these treatments.
Various mechanisms have been proposed to explain the suppression of plant-parasitic nematodes by entomopathogenic nematodes: (i) crowding of entomopathogenic nematodes along the plant roots forces plant-parasitic nematodes away (Bird & Bird 1986); (ii) massive doses of entomopathogenic nematodes lead to the build-up of nematode antagonistic organisms in the soil resulting in nematode suppression (Ishibashi & Kondo 1986; Ishibashi & Choi 1991); and (iii) allelochemicals like ammonium and indole produced by entomopathogenic nematodes and/or their symbiotic bacteria (Xenorhabdus spp. and Photorhabdus spp.) inhibit egg hatching and repel or intoxicate plant-parasitic nematodes (Grewal, Lewis & Venkatachari 1999; Hu, Li & Webster 1999).
In this study, the impact of entomopathogenic nematodes on the soil nematode community can be interpreted as a beneficial non-target effect based on the importance of plant-parasitic nematodes as agricultural pests. However, we cannot predict a similar effect on other soil fauna or flora. Furthermore, the metabolic products of symbiotic bacteria (Xenorhabdus spp. and Photorhabdus spp.) of entomopathogenic nematodes were reported to possess a broad spectrum of biological activity including insecticidal, nematicidal anti-mycotic, anti-carcinogenic and antibiotic properties (Webster, Chen & Li 1998). Therefore, the exact mechanisms that result in this effect need to be resolved in order to ascertain the safety of entomopathogenic nematodes to other organisms in the soil food webs.
Introduced biological control agents may have a short-term and/or long-term impact on native communities over a smaller or larger geographical area (Simberloff & Stiling 1996). Entomopathogenic nematodes do not have the ability to disperse long distances as they can only move up to a few centimetres per day through moist soil. (Strong et al. 1996). Furthermore, field releases of entomopathogenic nematodes quickly come to reflect natural densities and distributions (Georgis 1992; Ehlers & Hokkanen 1996; Smits 1996; Campbell et al. 1998), thus long-term negative effects on the local environment of entomopathogenic nematodes are likely to be negligible (Gaugler, Lewis & Stuart 1997). Therefore, the impact of entomopathogenic nematodes on the soil nematode community structure observed in this study could be a short-term effect confined to the application sites. However, the impact of entomopathogenic nematodes, even if it is a short-term effect, is significant because entomopathogenic nematodes are applied often as inundative biological control agents, and repeated applications of these nematodes to control recurring pest populations may sustain the effect. Therefore, the long-term ecological consequences of inundative applications of entomopathogenic nematodes need to be investigated in future studies.
Our results indicate that introduced biological control agents have the potential to affect the diversity of native fauna in soil ecosystems even though they do not have any direct parasite/host or predator/prey relationship. Therefore, the risks of biological control programmes need to be assessed in the broader context of their impact on all native fauna in the target ecosystems. Furthermore, our results emphasize the need for research to understand collateral damage to native fauna resulting from introduced species to ensure the safety and public confidence in biological control programmes.