A Comparison of Entomopathogenic Nematode Longevity in Soil under Laboratory Conditions

Abstract

We compared the longevity of 29 strains representing 11 entomopathogenic nematode species in soil over 42 to 56 d. A series of five laboratory experiments were conducted with six to eight nematode strains in each and one or more nematode strains in common, so that qualitative comparisons could be made across experiments. Nematodes included Heterorhabditis bacteriophora (four strains), H. indica (Homl), H. marelatus (Point Reyes), H megidis (UK211), H. mexicana (MX4), Steinernema carpocapsae (eight strains), S. diaprepesi, S. feltiae (SN), S. glaseri (NJ43), S. rarum (17C&E), and S. riobrave (nine strains). Substantial within-species variation in longevity was observed in S. carpocapsae, with the Sal strain exhibiting the greatest survival. The Sal strain was used as a standard in all inter-species comparisons. In contrast, little intra-species variation was observed in S. riobrave. Overall, we estimated S. carpocapsae (Sal) and S. diaprepesi to have the highest survival capability. A second level of longevity was observed in H. bacteriophora (Lewiston), H. megidis, S. feltiae, and S. riobrave (3–3 and 355). Lower levels of survivability were observed in other H. bacteriophora strains (Hb, HP88, and Oswego), as well as S. glaseri and S. rarum. Relative to S. glaseri and S. rarum, a lower tier of longevity was observed in H. indica and H. marelatus, and in H. mexicana, respectively. Although nematode persistence can vary under differing soil biotic and abiotic conditions, baseline data on longevity such as those reported herein may be helpful when choosing the best match for a particular target pest.

Entomopathogenic nematodes in the genera Heterorhabditis and Steinernema are capable of controlling a wide variety of economically important insect pests (Klein, 1990; Grewal and Georgis, 1999; Shapiro-Ilan et al., 2002). In nature, entomopathogenic nematodes are obligate parasites of insects (Poinar, 1990; Adams and Nguyen, 2002). In order to reproduce, infective juvenile nematodes (IJ), the only free-living stage, generally must persist in soil until encountering a host, which they enter through natural openings (mouth, anus, and spiracles) or, in some cases, through the cuticle. After entering the host's hemocoel, the nematodes release their symbiotic bacteria, which are primarily responsible for killing the host (Dowds and Peters, 2002). The nematodes reproduce within the host cadaver and ultimately produce up to several hundred thousand IJ that emerge to search out new hosts (Shapiro et al., 1999a; Shapiro and Gaugler, 2002). The nematodes contribute to biological pest suppression through the actions of endemic populations (Campbell et al., 1998; Duncan et al., 2003) and through purposeful release using classical, inoculative, or inundative approaches (Parkman et al., 1993; Grewal and Georgis, 1999, Shapiro-Ilan et al., 2002). Due to sensitivity to UV light and desiccation, entomopathogenic nematodes are usually applied to protected environments, particularly soil (Kaya and Gaugler, 1993; Grewal and Georgis, 1999).

The efficacy of entomopathogenic nematodes in biological control depends on certain beneficial traits such as host-finding and virulence (Gaugler, 1987; Shapiro-Ilan et al., 2002). The nematode's ability to persist in soil can contribute to pest control efficacy (Kaya, 1990; Kaya and Gaugler, 1993; Smits, 1996). Extended persistence in soil can result in greater cumulative insect mortality and a reduced need for multiple nematode applications. Some factors that affect nematode survival in soil include environmental conditions, such as moisture and temperature (Molyneux and Bedding, 1984; Molyneux, 1985; Kung et al., 1991), antagonists (Kaya, 2002), soil type (Kung et al., 1990; Shapiro et al., 2000), cultural conditions (Shapiro et al., 1996, 1999b; Millar and Barbercheck, 2002), and nematode species or strain (Molyneux, 1985; Ferguson et al., 1995; Grewal et al., 2002b).

Laboratory assays have characterized diverse nematode species and strain differences for a number of traits, such as virulence to particular host species or among host stages (Morris and Converse, 1991; Shapiro and McCoy, 2000; Grewal et al., 2002a; Koppenhöfer and Fuzy, 2004), temperature tolerance (Grewal et al., 1994), and desiccation tolerance (Patel et al., 1997; Liu and Glazer, 2000). These studies, though conducted under controlled conditions, provide baseline data that can be useful in choosing which entomopathogenic nematode might be the best match for a particular target pest. Similar experiments addressing variation in overall nematode persistence or longevity may be considered lacking in that they have generally been limited to comparisons of only one or a few species (Molyneux, 1985; Kung et al., 1990, 1991; Hass et al., 1999; Shapiro et al., 2000; Hass et al., 2001, 2002; Grewal et al., 2002b;). Our objective was to conduct a broader comparison of intra- and inter-specific variation in entomopathogenic nematode longevity in soil.

Materials and Methods

All nematodes were cultured in the last instar of the greater wax moth Galleria mellonella according to procedures described by Kaya and Stock (1997). Nematodes were stored at 13 °C for up to 3 wk prior to use in experiments. All experiments were conducted at 25 °C and repeated once in entirety.

A series of five experiments was conducted to compare intra- and inter-specific variation in entomopathogenic nematode longevity. The nematode strains and species used in each experiment are listed in Table 1. Experiments I and II focused on intra-species variation in S. carpocapsae and S. riobrave, respectively. Experiments III to V focused primarily on inter-species variation, except four different strains of H. bacteriophora were tested and an additional S. riobrave strain was added. To facilitate qualitative comparisons of inter-species variation across experiments III to V, several nematode strains were included in more than one experiment. The Sal strain of S. carpocapsae, which exhibited the highest overall survival in experiment I, was included in all three inter-species comparisons. Additionally, three other strains that exhibited relatively high survival, i.e., S. riobrave strain 3–3 and 355 and the SN strain of S. feltiae, were each included in two of the inter-species comparison experiments.

Table 1.

Entomopathogenic nematode strains used in longevity experiments I to V and their extraction efficiencies in soil.

Nematode survival in soil was measured by dilution and direct counting (Shapiro and Glazer, 1996; Shapiro and Lewis, 1999; Perez et al., 2003). Experimental units consisted of 30-ml plastic cups (3 to 4-cm-diam., 3.5-cm-deep), each containing 5 g of loamy sand soil (84% sand, 10% silt, 6% clay, pH 6.1, and organic matter = 2.8%). Nematodes were applied in 0.5 ml tap water, and the final moisture level in each cup was brought to field capacity (14%). One d after inoculation, the initial number of extractable live nematodes was determined by rinsing the entire contents from a cup into a 1-liter beaker and adding water to a final volume of 304 ml. The suspension was stirred on a stir-plate with a 5-cm stir bar at constant speed. One ml of suspension was removed and placed on a 60-mm-diam. petri dish lid that was marked with lines 1 cm apart to facilitate counting. Additional water was added to a depth of approximately 0.5 cm, and the number of live IJ was determined using a dissecting microscope (nematodes were considered live if they were moving naturally or if they responded to probing with a needle). To avoid basing survival estimates on small sample sizes, a minimum count of 30 IJ per sample was considered acceptable. If fewer than 30 IJ were observed from a single 1-ml sample, additional samples from the beaker were added to the petri dish. Based on the dilution factor, the total number of live nematodes was then estimated for the experimental unit. The process to determine number of surviving nematodes was repeated 7, 14, 28, 42, and 56 d post-inoculation (except experiment II ended on d 42 and experiment IV had the last sample date on d 53 rather than 56). For the first two experiments, 10,000 IJ were initially added to each cup, but toward the end of these experiments numerous 1-ml samples were required to achieve a count of 30 IJ. Therefore, the number of IJ initially added to each cup increased to 15,000 for experiments III and IV, and 20,000 for experiment V. Extraction efficiency for each nematode in each experiment was calculated based on the percentage of IJ extracted 1 d post-inoculation relative to the number applied. In each trial, there were five replicate cups for each treatment (10 replicates total).

Nematode longevity over time was estimated based on the total number of live nematodes remaining in the cups. The percentage survival for each nematode was determined by dividing the number of live nematodes detected on a particular sample date by the number initially extracted 1 d post-inoculation (similar to Hass et al., 2002). Even when applying the same technique, soil extraction efficiency can vary for different nematode species or strains (Curran and Heng, 1992) (Table 1). Therefore, our approach of relating longevity to the number of nematodes initially extractable (1 d post-inoculation) was designed to minimize differences in extraction efficiency. Further, basing longevity on the number of nematodes extractable 1 d post-inoculation emphasizes longevity once the nematodes are established in the soil and eliminates mortality that might have occurred during application.

Analysis of variance (ANOVA) was used to detect treatment differences in nematode longevity for each sample date (starting 7 d post-application), as well as over the whole experimental period (i.e., for the 42-, 53-, or 56-d period) (Shapiro et al. 2000) (SAS Software, version 9.1, 2001, SAS Institute, Cary, NC). In the analyses covering the entire experimental period, a conservative error term (replicate*treatment) was used to test for treatment effects (Cochran and Cox, 1957). ANOVA was also used to compare extraction efficiencies among treatments. When ANOVA detected statistical significance (P ≤ 0.05), treatment differences were further elucidated with the SNK test. In all analyses, percentage survival or extraction efficiency was arcsine transformed (arcsine of square root) prior to analysis (Southwood, 1978); untransformed means are presented in figures.

Results

Differences in nematode extraction efficiencies were detected among nematode strains and species (Table 1). Specifically, differences in extraction efficiencies were detected in experiments I (F = 3.68, df = 7,72, P = 0.002), IV (F = 3.94, df = 5,54, P = 0.004), and V (F = 2.64, df = 6,63, P = 0.024). Note that the Ftest was significant in experiment V, but the SNK test did not differentiate among treatments (Table 1). No differences in extraction efficiency were detected among treatments in experiments II (F = 0.88, df = 7,72, P = 0.527) and III (F = 0.60, df = 6,63, P = 0.728). Extraction efficiencies throughout the study ranged from 57% to 96% (Table 1).

Intra- and inter-specific differences in longevity were detected in all five experiments (Figs. 1–6) (Table 2). In experiment I, which focused on intra-specific variation among S. carpocapsae strains, Agriotos and Sal strains tended to exhibit the highest survival on sample dates 7 to 28, and Sal strain had the highest survival in the latter part of the experiment (i.e., on days 42 and 56) (Fig. 1). When longevity was analyzed over the entire experimental period, survival was greatest in the Sal strain followed by Agriotos, and among the lowest in Kapow, Breton, and All strains (Fig. 6A).

Fig. 1.

Average longevity of Steinernema carpocapsae strains in soil 7, 14, 21, 28, 42, and 56 d after inoculation (D7-D56). Different letters above bars within each day of evaluation indicate statistical differences (SNK test, P ≤ 0.05).

Fig. 6.

Longevity of entomopathogenic nematodes in soil averaged over a 42- to 56-d experimental period. Each graph represents a separate experiment. A) Steinernema carpocapsae strains. B) S. riobrave strains. C-E) Various strains and species: Hb = Heterorhabditis bacteriophora (HP88, Hb, Lewiston and Oswego strains), Hmeg = H. megidis (UK211), ScSal = S. carpocapsae (Sal), Sr = S. riobrave (3–3 and 355 strains), Hmex = H. mexicana (MX4 strain), Sd = S. diaprepesi, SfSn = S. feltiae (SN), Srar = S. rarum (17C&E), HiHoml = H. indica (Homl strain), Hmpoi = H. marelatus (Point Reyes), SgNJ43 = S. glaseri (NJ43). Different letters above bars within each experiment indicate statistical differences (SNK test, P ≤ 0.05).

Table 2.

Statistics generated from entomopathogenic nematode experiments.

Little intra-specific variation in longevity was observed among S. riobrave strains compared with variation among S. carpocapsae strains (compare Figs. 2, 6B to Figs. 1, 6A). On sample dates 7, 14, and 42, strains 9–5, 3–3, and 3–8a exhibited greater survival than four, five, and one other strain, respectively (Fig. 2). No other differences were detected on individual sample dates. Additionally, no differences were detected among strains when longevity was analyzed over the entire experimental period (Fig. 6B).

Fig. 2.

Average longevity of Steinernema riobrave strains in soil 7, 14, 21, 28, and 42, d after inoculation (D7-D42). Different letters above bars within each day of evaluation indicate statistical differences (SNK test, P ≤ 0.05).

In experiment III, S. carpocapsae Sal strain exhibited the greatest longevity relative to other nematodes; the Sal strain's survival was greater than S. riobrave (3–3), H. megidis, and the four H. bacteriophora strains tested at 14 d post-inoculation, and when compared over the whole experimental period (Figs. 3, 6C). Steinernema riobrave 3–3 had greater survival than all heterorhabditids on day 28 and all heterorhabditids except H. megidis on day 21. When analyzed over the entire period, survival of the 3–3 strain was not different from H. megidis or H. bacteriophora (Lewiston); H. megidis exhibited greater survival than all of the H. bacteriophora strains except Lewiston (Fig. 6C).

Fig 3.

Average longevity of entomopathogenic nematodesin soil 7,14, 21, 28, 42, and 56 d after inoculation (D7–D56). Hb = Heterorhabditis bacterioph&ra (HP88, Hb, Lewiston and Oswego strains), Hmeg = H. megidis (UK211), ScSal = Steinernema carpocapsae (Sal), Sr3–3 = S. riobrave (3–3). Different letters above bars within each day of evaluation indicate statistical differences (SNK test, P ≤ 0.05).

In Experiment IV, inter-species comparisons revealed no significant differences 7 d post-inoculation, yet differences were detected subsequently (Fig. 4). On day 14 and thereafter, S. carpocapsae (Sal) remained in the highest-ranked group for survival throughout the experimental period. Steinernema diaprepesi survival was lower than S. carpocapsae (Sal) on day 21, but exhibited similar longevity thereafter. Over the whole experimental period, survival of S. carpocapsae (Sal) and S. diaprepesi was the highest, followed by S. feltiae, which was greater than S. rarum but not different from S. riobrave (355) (Fig. 6D). Survival of H. mexicana was the lowest among the nematodes tested in experiment IV 14 d post-inoculation and thereafter as well as when averaged over the whole experimental period (Figs. 4, 6D).

Fig. 4.

Average longevity of entomopathogenic nematodes in soil 7, 14, 21, 28, 42, and 53 d after inoculation (D7-D53). Hmex = Heterorhabditis mexicana (MX4 strain), ScSal = S. carpocapsae (Sal), Sd = Steinernema diaprepesi, SfSn = S. feltiae (SN), Sr355 = S. riobrave (355), Srar = S. rarum (17C&E). Different letters above bars within each day of evaluation indicate statistical differences (SNK test, P ≤ 0.05).

In experiment V, the two heterorhabditids, H. marelatus and H. indica, exhibited lower survival than all steinernematids within the first 4 wk except S. glaseri, and survival of these heterorhabditids was lower than all steinernematids in the latter part of the experiment (Fig. 5). Similarly, over the entire experimental period H. marelatus and H. indica had the lowest survival, followed by S. glaseri (Fig. 6E). Also, when averaged over the experimental periods, no difference in survival was detected among S. carpocapsae Sal, S. feltiae (SN), or S. riobrave (strains 3–3 and 355) (Fig. 6E).

Fig. 5.

Average longevity of entomopathogenic nematodes in soil 7, 14, 21, 28, 42, and 56 d after inoculation (D7-D56). HiHoml = Heterorhabditis indica (Homl strain), Hmpoi = H. marelatus (Point Reyes), ScSal = S. carpocapsae (Sal), SfSn = S. feltiae (SN), SgNJ43 = S. glaseri (NJ43), Sr = S. riobrave (3–3 and 355 strains). Different letters above bars within each day of evaluation indicate statistical differences (SNK test, P ≤ 0.05).

As an illustrative summary of the inter-species comparisons, a schematic diagram was generated based on experiments III to V (Fig. 7). The relative rankings in the diagram were derived from nematode survival when averaged over the duration of the experiments, i.e., the entire 53 or 56 d (Fig. 6C–E). In general, to construct the diagram nematode species or strains estimated to have greater survival abilities were placed above those estimated to have lower survival. Specifically, if a nematode strain was found to have greater survival than another strain in at least one experiment, then that relationship of superior survival was represented in the diagram by a solid vertical line. Strains on the same level connected through horizontal lines were estimated to have approximately equivalent survival based on direct comparison or by inference from comparisons made among strains common to more than one experiment. For example, H. megidis and S. feltiae were not directly compared but were placed at the same level because (in separate experiments) their survival capabilities were both similar to S. riobrave (3–3). In direct comparisons where intermediate survival relationships were indicated, dotted lines were used to show where a lack of significance occurred. For example, survival of H. megidis was greater than H. bacteriophora (Oswego) but similar to the Lewiston strain, and Lewiston was not found to be different from the Oswego strain (Fig. 6C); thus, dotted lines were used between H. bacteriophora Oswego and Lewiston strains (Fig. 7). Nematode strains not connected through one or more series of vertical or horizontal lines were not considered to be subject to comparison, e.g., relative survival between S. glaseri vs. S. rarum or H. indica vs. H. mexicana was not subject to ranking (even qualitatively).

Fig. 7.

A schematic summarizing a series of laboratory experiments comparing survival among entomopathogenic nematodes in soil. The illustrative summary is based on average survival over a 53- or 56-d period (see Fig. 6C-E). In general, nematode species or strains higher up on the diagram (indicated by solid vertical lines) exhibited greater survival than those below. Strains on the same level connected by horizontal lines had approximately equivalent survival (based on direct comparison or inference from comparisons made among strains common to more than one experiment). Dotted lines between two strains indicate that a statistical difference was not detected.

Based on the ranking approach described above, a summary of experiments III to V indicates the highest survival capability in S. carpocapsae (Sal) strain and S. diaprepesi (Fig. 7). A second level of survivability was observed in H. bacteriophora (Lewiston), H. megidis, S. feltiae, and S. riobrave (3–3 and 355 strains). Lower levels of survivability were observed in H. bacteriophora (Hb, HP88, and Oswego strains), S. glaseri and S. rarum. Relative to S. glaseri and S. rarum, a lower tier of survival was observed in H. indica and H. marelatus, and H. mexicana, respectively (Fig. 7).

Discussion

In general, treatment effects of inundatively applied entomopathogenic nematodes have been observed to persist for a relatively short term, e.g., over 2 to 6 wk (Duncan et al., 1996; Smits, 1996, McCoy et al., 2000). In some instances, however, longer-term persistence and insect control have been reported to occur over several seasons or years (Klein and Georgis, 1992; Parkman et al., 1993; Shields et al., 1999). Endemic nematode populations also can contribute to long-term suppression of insect populations (Duncan et al., 2003). Long-term persistence of entomopathogenic nematodes depends to a large extent on their ability to find and reproduce in suitable hosts (Kaya, 1990; Smits, 1996), yet the probability of recycling and potential for prolonged insect control depends on how long the nematodes can survive without a host. We observed substantial difference among various nematode Species and strains in their innate abilities to survive in soil. Some nematodes, such as S. diaprepesi and S. carpocapsae (Sal), displayed superior potential for persistence, whereas others, such as H. mexicana and H. indica, exhibited relatively poor longevity. Laboratory comparisons such as those reported herein may be helpful when considering which nematode strain or species to use in field applications.

Our findings on relative entomopathogenic nematode longevity are consistent with some previous laboratory studies but not with others. For example, our findings of superior survival of S. carpocapsae relative to S. glaseri are similar to those of Kung et al. (1990). Also consistent with our study, Molyneux (1985) reported greater longevity in S. carpocapsae and S. glaseri relative to two heterorhabditids but, unlike our study and that of Kung et al. (1990), survival of S. glaseri in Molyneux's study was greater than that of S. carpocapsae. One possible cause for these discrepancies is strain variation. We observed considerable variation in longevity among S. carpocapsae strains. Similarly, Grewal et al. (2002b) reported significant variation in longevity among 15 populations of H. bacteriophora. However, we observed little difference among S. riobrave strains, indicating that within-species variation in longevity is greater for some nematodes than others. Possibly, the different degrees of intra-species variation observed is due to the substantially wider geographical range in which the S. carpocapsae strains originated compared with the relatively narrow range of the S. riobrave strains (Poinar, 1990; Stuart et al., 2004).

Our rankings of nematode longevity in the laboratory are consistent with observations on nematode persistence in certain field studies. For example, our observations of superior longevity in S. diaprepesi are consistent with prior evidence of this nematode's persistence capabilities under field conditions, e.g., relative to S. riobrave (Duncan et al., 2003). Also analogous to our study, persistence of H. bacteriophora under field conditions was reported to be lower than that of S. riobrave and S. carpocapsae (Duncan et al., 1996). Jansson et al. (1993), however, reported greater field persistence with S. feltiae than with S. carpocapsae strains (a trend we did not observe). Additionally, we observed relatively poor longevity in H. bacteriophora (Oswego), yet this strain was reported to exhibit substantial persistence and infectivity for 700 d or more in the field (Ferguson et al., 1995; Shields et al., 1999). Thus, the Oswego strain's persistence under field conditions was likely due to its ability to efficiently find and reproduce in insect hosts rather than its innate longevity. Indeed, a variety of factors can contribute to variations in nematode longevity and persistence under field conditions (Kaya, 1990; Smits, 1996; Kaya, 2002).

A physiological basis for differences in longevity observed in this study is not clear. Grewal et al. (2002b) observed a strong positive correlation between H. bacteriophora longevity in sterile water at 25 °C and tolerance to heat, UV, and hypoxia, indicating a possible genetic link between these stress factors and longevity. Lipid content or energy reserves have been related to entomopathogenic nematode longevity (Selvan et al., 1993; Abu Hatab and Gaugler, 2001). In that case, one might have expected the longevity of S. glaseri to have been relatively higher in our study due to its greater energy content (Selvan et al., 1993). On the other hand, Hass et al. (2002) also reported a correlation between entomopathogenic nematode persistence and energy reserves but concluded that the effect was not directly responsible for reductions in nematode survival. The authors suggested differences in longevity were more likely attributable to other factors, such as activity of antagonists in nonsterile soil (as was used in this study as well).

Our laboratory comparison of entomopathogenic nematode longevity included 29 strains representing 11 species. At least eight of the species tested are or have been commercially available (H. bacteriophora, H. indica, H. megidis, H. marelatus, S. carpocapsae, S. feltiae, S. glaseri, and S. riobrave). Although a number of species and numerous strains were not included (Poinar, 1990; Adams and Nguyen, 2002), our study was the broadest comparison to date. Due to limitations in space and labor requirements, it was not feasible to compare all nematodes simultaneously. However, incorporation of strains that were common among the experiments (which were conducted under essentially identical conditions) allows for substantially more valid qualitative comparisons than when comparing across studies conducted under varying conditions. It is likely that the relative longevities of nematodes we tested will differ to some (or a large) extent when tested under varying temperatures, soil types, etc., or under the complexities of diverse field conditions. Additionally, it must be noted that a nematode's ability to survive for a certain period does not necessarily mean it will remain infective during that time; our focus was on longevity, not persistence of infective abilities. Clearly, further research is needed to address entomopathogenic nematode longevity and long-term efficacy in various environments. Nonetheless, our study provides an initial assessment and baseline data from which expanded studies can be made.