- Email: [email protected]
Aggregative group behavior in insect parasitic nematode dispersal
International Journal for Parasitology 44 (2014) 49–54
Contents lists available at ScienceDirect
International Journal for Parasitology journal homepage: www.elsevier.com/locate/ijpara
Aggregative group behavior in insect parasitic nematode dispersal David I. Shapiro-Ilan a,⇑, Edwin E. Lewis b, Paul Schliekelman c,⇑ a
USDA-ARS, SEFTNRL, 21 Dunbar Road, Byron, GA 31008, USA University of California Davis, Department of Nematology and Entomology, University of California, Davis, CA 95616, USA c Department of Statistics, University of Georgia, Athens, GA 30602, USA b
a r t i c l e
i n f o
Article history: Received 26 July 2013 Received in revised form 1 October 2013 Accepted 5 October 2013 Available online 31 October 2013 Keywords: Behavior Dispersal Entomopathogenic nematode Group foraging Heterorhabditis Steinernema
a b s t r a c t Movement behavior of foraging animals is critical to the determination of their spatial ecology and success in exploiting resources. Individuals sometimes gain advantages by foraging in groups to increase their efficiency in garnering these resources. Group movement behavior has been studied in various vertebrates. In this study we explored the propensity for innate group movement behavior among insect parasitic nematodes. Given that entomopathogenic nematodes benefit from group attack and infection, we hypothesised that the populations would tend to move in aggregate in the absence of extrinsic cues. Movement patterns of entomopathogenic nematodes in sand were investigated when nematodes were applied to a specific locus or when the nematodes emerged naturally from infected insect hosts; six nematode species in two genera were tested (Heterorhabditis bacteriophora, Heterorhabditis indica, Steinernema carpocapsae, Steinernema feltiae, Steinernema glaseri and Steinernema riobrave). Nematodes were applied in aqueous suspension via filter paper discs or in infected insect host cadavers (to mimic emergence in nature). We discovered that nematode dispersal resulted in an aggregated pattern rather than a random or uniform distribution; the only exception was S. glaseri when emerging directly from infected hosts. The group movement may have been continuous from the point of origin, or it may have been triggered by a propensity to aggregate after a short period of random movement. To our knowledge, this is the first report of group movement behavior in parasitic nematodes in the absence of external stimuli (e.g., without an insect or other apparent biotic or abiotic cue). These findings have implications for nematode spatial distribution and suggest that group behavior is involved in nematode foraging. Published by Elsevier Ltd. on behalf of Australian Society for Parasitology Inc.
1. Introduction Behaviors involving movement are intimately tied to the spatial distribution and foraging success of most animals. When animals increase their individual fitness by exploiting resources in groups, the spatial distribution of their populations should be overdispersed. There are many examples of group foraging behavior, or cooperative behaviors, demonstrated by organisms of diverse trophic levels (Nøttestad et al., 2002; Fourcassié et al., 2010; Bonnell et al., 2013; Cook et al., 2013). Various biotic and abiotic factors can affect movement behavior (Perony and Townsend, 2013) such as group size (Bonnell et al., 2013), environmental spatial structure and resource availability (Patterson and Messier, 2001; Reeve and Cronin, 2010), and avoidance of predators (Srinivasan et al., 2010; De Vos and O’Riain, 2013). But if a massattack strategy is necessary for efficient garnering of resources, then we would hypothesise that even in the absence of these factors, populations should remain overdispersed, and thus their ⇑ Corresponding authors. Tel.: +1 478 956 6441; fax: +1 478 956 2929. E-mail addresses: [email protected] (D.I. Shapiro-Ilan), pdschlie@ uga.edu (P. Schliekelman).
movement should be en masse. In this study we investigated the group movement and aggregation behavior of insect parasitic nematodes in an arena that was environmentally homogeneous and devoid of resources. Some species follow temporal patterns of random movement while foraging whereas others move in directed spatial patterns, alone or in aggregate (Bonnell et al., 2013). Here we are interested in how movement in groups relates to the behavior and fitness of the individual. For group behaviors to persist in any animal population there must be fitness gains for the individuals who form the group; the individuals within the group must win some advantage over individuals on their own (Clark and Mangel, 1986). This is especially true in groups composed of unrelated or distantly related individuals. These advantages potentially include increased probability of finding resources, better ability to overcome defensive actions of hosts or prey, or the mitigation of risk posed by engaging with the resource. Entomopathogenic nematodes in the genera Heterorhabditis and Steinernema (the subjects of this study) are obligate parasites of insects. These nematodes are important natural regulators of insect populations and are also applied as biological control agents to suppress pest populations (Lacey and Shapiro-Ilan, 2008). The
0020-7519/$36.00 Published by Elsevier Ltd. on behalf of Australian Society for Parasitology Inc. http://dx.doi.org/10.1016/j.ijpara.2013.10.002
50
D.I. Shapiro-Ilan et al. / International Journal for Parasitology 44 (2014) 49–54
developmentally arrested infective juvenile nematodes (IJs), the only free-living stage, typically occupy soil habitats until they infect an insect, which they must do in order to resume development and reproduce. Generally, an insect host is infected by tens to hundreds of individual IJs, and a single IJ is seldom able to initiate a successful infection (Koppenhöfer et al., 2006; Nguyen et al., 2006), which confers a fitness advantage to individuals that invade hosts as part of a group. A few hours after the nematodes enter the host, the nematodes release symbiotic bacteria, which are the primary agents in killing the host (Kaya and Gaugler, 1993). Infected hosts generally die within 24–72 h, and the nematodes feed on the symbiotic bacteria and insect tissues, reproducing for one to three generations (Kaya and Gaugler, 1993). As food resources dwindle, a new cycle of IJs is produced and they emerge to seek new hosts. The number of IJs exiting a host can range from tens of thousands to hundreds of thousands (Shapiro-Ilan and Gaugler, 2002). Foraging strategies among entomopathogenic nematode species vary along a continuum between ambushers, which generally sit and wait for a passing host, and cruisers that actively search for hosts (Lewis et al., 1992). Entomopathogenic nematodes can exhibit a combination of these behaviors to locate hosts and although some species exhibit primarily ambush-type behaviors and others are mainly cruisers, others are considered intermediate in their foraging behavior (Campbell and Kaya, 1999; Lewis, 2002). Foraging strategies may contribute to spatial distribution of entomopathogenic nematode populations in the environment. Several studies indicate that the distribution of entomopathogenic nematode populations is aggregated in nature (Stuart and Gaugler, 1994; Campbell et al., 1996). The clumped distribution may largely be due to a sparse distribution of infection loci and subsequent mass emergence from host cadavers, thus creating nematode hot spots within the soil ecosystem (Campbell et al., 1996; Spiridonov et al., 2007). However, we suggest that aggregate distributions may also result from group nematode movement when dispersing from hosts or points of application. Group movement of entomopathogenic nematodes may have adaptive value, e.g., in maximising chances for successful host infection. Similar to some other predators and parasites, group attack is generally required by entomopathogenic nematodes to overcome host defenses (Li et al., 2007). Although entomopathogenic nematode dispersal has been studied in terms of biotic and abiotic factors that affect movement, or distance travelled within the soil profile (Stuart et al., 2006; Jabbour and Barbercheck, 2008; El-Borai et al., 2011), there is a dearth of knowledge on more general patterns of movement, especially when biotic and abiotic drivers of movement are not confounding results. Recently, Fushing et al. (2008) described mathematically the temporal aspects of group infection behavior in certain entomopathogenic nematode species. In a similar fashion, we hypothesise that entomopathogenic nematodes also display group movement behavior, even in the absence of hosts. In this study that hypothesis was tested with six species of entomopathogenic nematodes.
Nematodes were applied to experimental arenas using two approaches: application in infected host cadavers or in aqueous suspension. The cadaver approach was intended to represent natural emergence from the host, whereas the aqueous suspension method was intended to mimic artificial introduction of nematodes to the environment for purposes of biological pest control. Experimental arenas consisted of polypropylene boxes (20.5 20.5 cm, 9 cm deep) containing sand filled to a depth of 2.5 cm. In all arenas, the final total moisture level in sand was at approximate field capacity (10%). For the cadaver application approach, three nematode-infected hosts (at 3 days p.i.) were placed in the center of each arena. The cadavers were observed daily for the initiation of nematode emergence. The number of IJs that emerged per cadaver was estimated based on separate yield assessments (Kaya and Stock, 1997) conducted on 10 replicate insects each. The results indicated average (±S.E.M.) IJ production by the species as follows: H. bacteriophora produced 244,390 ± 28,419.1 IJs; H. indica produced 112,210 ± 11,183.9 IJs; S. carpocapsae produced 178,300 ± 15,507.9 IJs; S. feltiae produced 153,700 ± 14,272 IJs; S. glaseri produced 76,650 ± 5,299.9IJs; S. riobrave produced 126,080 ± 10,788.9 IJs. For application of aqueous suspensions, 200,000 IJs from culture flasks (except 300,000 IJs were used for S. riobrave) were vacuum filtered onto filter paper (60 mm, Whatman No. 1); the filter paper was then placed upside down in the center of a sand arena. Arenas were stored inside plastic bags at 25 °C until nematode movement was assessed. The plastic bags contained moist paper towels to ensure that high levels of relative humidity (approximately 100%) were maintained. There were four replicate arenas for each species and application approach. Experiments addressing movement of each nematode species were conducted separately, and all experiments were repeated once in time (i.e., run two times in total) using different batches of nematodes and insects. In arenas with host cadavers, the results of nematode movement were assessed 3 days after IJ emergence began. In arenas receiving aqueous nematode applications, resulting nematode movements were assessed 3 days post-treatment. At the time of assessment, cadavers and filter papers were removed and soil cores measuring 2 cm diameter and 2.5 cm depth were taken at four points that were 4.5 cm from the center of the arena (diagonally and equidistant from each other). The sand from each core was then placed in 100 ml of tap water in a 250 ml beaker, and the number of nematodes in the core was determined by dilution counts under a stereomicroscope (Shapiro-Ilan et al., 2006). For the aqueous suspension approach, the total number of IJs entering sand was also estimated by washing any remaining IJs from the filter paper and subtracting that number from the number originally applied. 2.2. Analysis of movement patterns
2.1. Experimental approach
Our goal was to determine whether there is evidence that the post-emergence or post-application distribution of nematodes indicate aggregated movement patterns. The null hypothesis was that the nematodes move independently versus the alternate hypothesis that they tend to aggregate. Two different approaches to the analysis were taken.
All nematodes were cultured in vivo at 25 °C in commercially obtained final instar Galleria mellonella according to procedures described in Kaya and Stock (1997). Nematode species included Heterorhabditis bacteriophora (Oswego strain), Heterorhabditis indica (HOM1 strain), Steinernema carpocapsae (All strain), Steinernema feltiae (SN strain), Steinernema glaseri (NJ43 strain) and Steinernema riobrave (355 strain). The nematodes were stored in culture flasks at 13 °C for less than 2 weeks prior to experimentation.
2.2.1. First approach: Index of Dispersion If the null hypothesis is further specified to be that the nematodes undergo independent diffusive movement, then the number of IJs that is found in a core of a fixed size should follow a Poisson distribution with a mean proportional to the area of the core. The data for each trial consisted of the four cores of equal size that were equidistant from the center. Thus, under the null hypothesis each core follows a Poisson distribution with the same mean.
2. Materials and methods
D.I. Shapiro-Ilan et al. / International Journal for Parasitology 44 (2014) 49–54
The Poisson distribution has the property that its mean is the same as its variance. Thus, if the null hypothesis is correct, then the mean of the counts across the four cores should be approximately equal to the variance of counts across the quadrants. The ratio of the variance to the mean across the cores is known as the Index of Dispersion (see Krebs, 1999 for overview). Thus, an index greater than one indicates more aggregation than expected under the null hypothesis; less than one indicates a more uniform distribution than expected under the null hypothesis. If n is the number of cores and D is the Index of Dispersion, then (n 1)D has a chi-square distribution under the null hypothesis. A significant deviation of this quantity from its null hypothesis expectation indicates that the counts do not follow a Poisson distribution, giving evidence of aggregative movement. This test was applied to each replicate in each combination of the conditions separately. 2.2.2. Second approach: simulation approach for calculating P-values We have four replicates of each set of experimental conditions. The Index of Dispersion approach described above is a classical approach to the problem of testing for deviations from independent diffusive movement. However, it does not allow combining data over multiple replicates. Thus, we calculated the values separately for each replicate in our data. In order to do a joint test over replicates, we used a test statistic consisting of the product of the Index of Dispersion across replicates and used a simulation approach to calculate P values. We wrote an R program that generated 10,000 simulated null hypothesis data sets for each set of experimental conditions. Under the null hypothesis, the counts in each core follow a Poisson distribution. Each simulated data set consisted of four sets of four counts (one for each core) generated from a Poisson distribution, corresponding to the four replicates with four cores each. The dispersion statistics were calculated for each group of four cores and the test statistic was the product of the dispersion statistics across the four replicate data sets. The simulation P value is thus the proportion of the 10,000 simulated data sets in which the product of D across the four replicates was greater than the product observed in the actual data. A third approach for calculating P values over replicates based on the multinomial distribution is shown in Supplementary Table S2. This analysis was only conducted for the aqueous suspension experiments because the specific number of nematodes in each arena was not known for the cadaver experiments. There was good agreement between these results and the results above. 3. Results Table 1 shows the Index of Dispersion calculated for each arena in the study. This has an expected value of 1 for independent diffusive movement and larger values indicate evidence of aggregative movement. Values greater than 2.6 show statistically significant deviation (a = 0.05) from the null expectation (P-values are given in Supplementary Table S1). The majority of arenas indicated aggregative movement (Fig. 1). In most treatments, aggregative movement was indicated in >60% of the arenas (Fig. 1) and for five treatments aggregative movement was observed in P87.5% of the of the arenas (Fig. 1). In contrast, for the S. glaseri-cadaver treatments none of the arenas showed significant departure from random movement (Fig. 1; Supplementary Table S1). The S. feltiaecadaver treatment and the S. glaseri-filter paper treatments both only indicated aggregative movement in 50% of the arenas. Next, we analysed nematode distribution across replicates for both the filter paper and cadaver applications. We evaluated overall evidence for aggregative movement using the product of the Index of Dispersion over the four replicates for each experimental condition and simulated P values as described in the Section 2.2.2.
51
These results also indicated aggregated movement in most of the analyses (Table 1; Supplementary Table S1). Specifically, 20 of the 24 analyses indicated non-random movement (Table 1). Particularly high levels of aggregation were observed in some treatments such as H. bacteriophora and H. indica when applied in host cadavers. Neither of the S. glaseri-cadaver trials indicated that non-random movement had taken place, and only one of the two trials was significantly aggregated for the S. feltiae-cadaver and H. bacteriophora-filter paper treatments (Table 1). The outcome of the third approach (Supplementary Table S2) shows close agreement between the P values calculated by simulation and those calculated by the method of the first approach. This is evidence that both sets of P values are being calculated correctly. In summary, evidence for aggregative movement was observed in all treatments except in the S. glaseri-cadaver treatment, for which none of the statistical tests suggested non-random movement. In two other treatments, evidence for aggregative movement was detected although the effect was less pronounced than in other treatments. Specifically, aggregative movement was lower in the S. feltiae-cadaver treatment, for which only 50% of individual arenas exhibited a significant effect in the first statistical approach, and only one of the two trials exhibited a significant effect in the second approach. Additionally, evidence for aggregative movement was somewhat lower for the H. bacteriophora-filter paper treatment, as only one of the two trials indicated significance in the second statistical approach, yet aggregative movement was detected in 62.5% of the individual arenas using the first statistical approach.
4. Discussion Our findings provided evidence for aggregated group behavior in entomopathogenic nematode movement (in all treatments except in the S. glaseri-cadaver treatment). Although the study included an array of nematode species, all treatments exhibited aggregated movement patterns regardless of genus (Heterorhabditis or Steinernema) or foraging type (ambusher, cruiser or intermediate). The movement patterns resulted in a patchy nematode distribution in the sand arenas that was not due to the nematodes responding to any extrinsic stimulus. Similar to most soil organisms, entomopathogenic nematodes are in fact generally found in a clumped distribution in nature (Stuart and Gaugler, 1994; Giller, 1996; Campbell et al., 1996, 1998; Convey et al., 2002; Spiridonov et al., 2007). Even when applied to the soil environment in a uniform distribution, soon after application, entomopathogenic nematodes tend to be found in a patchy distribution (Campbell et al., 1998). A number of extrinsic factors have been postulated to contribute to patchiness in entomopathogenic nematode distributions including a patchy distribution of hosts, mass emergence from infected cadavers, variable persistence due to biotic and abiotic factors in the soil (following emergence) and local extinctions (Stuart and Gaugler, 1994; Campbell et al., 1998; Stuart et al., 2006). Based on our findings, we contend that aggregated movement behavior may further contribute to clumping of natural and applied entomopathogenic nematode populations. The possibility of group movement in S. riobrave was suggested by El-Borai et al. (2011) after observing that IJs tended to move toward one side in columns of sand that were supposedly uniform; in the current study we have demonstrated experimentally that entomopathogenic nematodes do indeed move in aggregate patterns, thereby providing an explanation for the previous observation. The group movement behavior observed in this study may have been continuous from the point of origin, or it may have been triggered by a propensity to aggregate after a short period of random movement. Given that we only measured movement from the inoculation site to one location within each quadrant, we cannot
52
D.I. Shapiro-Ilan et al. / International Journal for Parasitology 44 (2014) 49–54
Table 1 Statistical values resulting for entomopathogenic nematode movement in a sand arena. Trial
N
Test
D1
D2
D3
D4
Dprod
P
1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2
Sr Sr Sr Sr Sf Sf Sf Sf Sg Sg Sg Sg Sc Sc Sc Sc Hi Hi Hi Hi Hb Hb Hb Hb
FP FP Cad Cad FP FP Cad Cad FP FP Cad Cad FP FP Cad Cad FP FP Cad Cad FP FP Cad Cad
5.58 16.86 4.99 16.38 14,971.7 12.44 0.53 1.08 2.42 5.33 1.00 0.67 3.92 5.06 1.62 8.96 14.47 2.31 69.78 11,860.3 69.30 0.70 54.70 14,561.2
9.55 3.71 0.89 0.93 32.38 8.75 4.45 3.22 2.27 2.13 1.59 0.20 9.38 2.18 33.32 11.72 7.59 1.79 16.56 13,314.84 15.07 3.44 47.54 25,998.64
25.78 27.93 20.07 17.95 5.43 2.75 0.67 25.81 4.20 0.86 2.00 0.52 4.99 6.50 1.83 2.25 8.71 4.56 37.91 28,410.70 40.86 1.11 23,284.40 13,158.79
3.17 3.99 11.95 14,709.41 2.01 2.84 0.33 3.42 5.02 3.21 0.20
4,347.70 6,983.86 1,062.91 4,038,821.73 5,300,759.70 849.72 0.53 306.49 115.91 31.25 0.64 0.07 642.64 1,712.30 548.14 879.34 708.28 205.22 534,900,612.34 70,454,974,403,547,160.00 1,063,559,253.61 0.90 1,017,372,714.10 73,264,704,285,453.59
0.00 0.00 0.00 0.00 0.00 0.00 0.36 0.00 0.00 0.00 0.32 0.84 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.26 0.00 0.00
3.50 23.88 5.54 3.72 0.74 10.84 12,213.71 15,703.57 24,935.13 0.33 16.80 14.71
Nematodes (N) were applied to sand arenas on filter paper (FP) or in infected Galleria mellonella cadavers (Cad). Soil cores were taken to determine nematode distribution within the arena. Each D product and P value is derived from the combination of four replicates; the experiments were repeated (two trials). When P is listed as zero this means that it was less than 10 324. This table reports the result from the second statistical approach based on simulation of P values (see Section 2.2.2. in the text and Supplementary Table S1 for details). A statistically significant value indicates an aggregated distribution. Hb, Heterorhabditis bacteriophora; Hi, Heterorhabditis indica; Sc, Steinernema carpocapsae; Sf, Steinernema carpocapsae; Sg, Steinernema glaseri; Sr, Steinernema riobrave.
Fig. 1. Percentage of arenas indicating a significant departure from random movement by entomopathogenic nematodes (thus indicating aggregative movement). Nematodes were applied to sand arenas on filter paper (Fp) or in infected Galleria mellonella cadavers (Cad). Soil cores were taken to determine nematode distribution within the arena. Hb, Heterorhabditis bacteriophora; Hi, Heterorhabditis indica; Sc, Steinernema carpocapsae; Sf, Steinernema feltiae; Sg, Steinernema glaseri; Sr, Steinernema riobrave.
rule out the possibility that the nematodes initially moved randomly, but subsequently (within the 3 day interval) moved in group fashion to form an aggregate pattern. Aggregation behavior, such as in response to food sources, has been observed in the nematode Caenorhabditis elegans (Srinivasan et al., 2012; Ludewig and Schroeder, 2013). In the current study, one might argue that continuous group movement was more likely because random movement followed by aggregation would result in unnecessary energy expenditure and would not provide any obvious advantage (such as coming to a food source). Nonetheless, additional research is needed to elucidate whether the group behavior is continuous or temporal in nature. Movement behavior in many animals (including nematodes) is directionally influenced by various stimuli such as physical or chemical gradients and environmental structure (Boender et al., 2011; Bonnell et al., 2013; Perony and Townsend, 2013). Entomopathogenic nematodes respond to a variety of stimuli such as CO2
(Lewis et al., 1993; Lewis, 2002), vibration (Torr et al., 2004), temperature (Burman and Pye, 1980; Byers and Poinar, 1982), chemical compounds (Pye and Burman, 1981; Shapiro et al., 2000), and electromagnetic stimuli (Shapiro-Ilan et al., 2012; Ilan et al., 2013). Aggregative movement behavior has been observed in certain nematode species in response to certain stimuli such as adverse environmental conditions (e.g., to avoid desiccation) (Croll and Matthews, 1977). However, in the current study, arenas were arranged such that all potential internal or external stimuli were held constant under conditions conducive to nematode survival and therefore environmental factors did not cause bias in nematode movement. Accordingly, we conclude that the group movement behavior observed was not due to a response to stimuli, but reflects the nematodes’ natural propensity to aggregate. To our knowledge, this is the first report of group movement behavior in parasitic nematodes in the absence of external stimuli (e.g., without an insect or other apparent biotic or abiotic cue). Furthermore, our observations were made in the absence of hosts. Future studies will investigate the dynamics of entomopathogenic nematode group movement behavior in the presence of varying biotic and abiotic stimuli. Entomopathogenic nematodes exhibit group infection behavior (Hay and Fenlon, 1995; Fushing et al., 2008). Infection behavior was postulated to be consistent with risk sensitive foraging patterns (Fushing et al., 2008). In this ‘‘follow-the-leader’’ or herding paradigm, risk-prone nematodes infect the host, followed by nematodes that are more risk averse (Fushing et al., 2008). Followthe-leader behavior has been observed in group movement patterns of both vertebrates and invertebrates (Kaur et al., 2012; Perony and Townsend, 2013). Conceivably, the aggregative group movement observed in this study may also be due to a propensity toward herding or follow-the-leader behavior. Group movement behavior combined with group infection has adaptive value for individual group members in terms of enhancing potential for foraging success, i.e., resulting in a greater chance of finding a host in sufficient numbers to successfully infect and overcome the host
D.I. Shapiro-Ilan et al. / International Journal for Parasitology 44 (2014) 49–54
immune system. In the case that individual IJs come into contact with a host, even if the host supports reproduction, without other IJs to follow on in infection, these individuals are unlikely to have reproductive success. Thus, natural selection should favor individuals that are part of a group over individuals on their own. The causal mechanisms behind the group dispersal and aggregation behavior observed in this study are unknown. Kaplan et al. (2012) reported that ascarosides caused dispersal in several entomopathogenic nematode species as well as other nematode species. Indeed, the impact of ascarosides on aggregation and other behaviors in C. elegans has been reported in a number of studies (Srinivasan et al., 2012; Ludewig and Schroeder, 2013). Conceivably, similar compounds could signal aggregative group movement behavior in entomopathogenic nematodes. It is interesting that S. glaseri, when applied in an infected host, was the only treatment that did not exhibit aggregative movement. The reasons for this finding are unclear, more so because the nematode did exhibit aggregative movement when applied in aqueous suspension via filter paper. Given that chemical factors associated with the nematode-infected host can elicit nematode behaviors such as infection or dispersal (Shapiro and Lewis, 1999; Kaplan et al., 2012), and these factors and their impact can be diluted when nematodes are placed in aqueous suspension (Shapiro and Lewis, 1999), it is possible that compounds associated with the S. glaseri-infected cadaver differed in quality or quantity from those present in the aqueous treatment. Furthermore, relative to other entomopathogenic nematode species, S. glaseri is highly mobile and competitive (Grewal et al., 1994, 1997; Koppenhöfer et al., 1995), and therefore the need for aggregative behavior in freshly emerging (highly fit) S. glaseri may be reduced. On the other hand, it is possible that the highly mobile nature of S. glaseri IJs results in their aggregations developing on a spatial scale that was not detected in the small arenas used in our experiments; a relatively low average number of IJs were found in S. glaseri-cadaver treatment cores (<10 IJs compared with >35 IJs in all other species), indicating that possibly these nematodes had already moved to the edge of the arena. Additional research is needed to elucidate the basis for group movement behavior in entomopathogenic nematodes.
Acknowledgments We thank Wanda Evans and Kathy Halat for technical assistance.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ijpara.2013.10. 002.
References Boender, A.J., Roubos, E.W., van der Velde, G., 2011. Together or alone? Foraging strategies in Caenorhabditis elegans. Biol. Rev. 86, 853–862. Bonnell, T.R., Dutilleul, P., Chapman, C.A., Reyna-Hurtado, R., Hernández-Sarabia, R.U., Sengupta, R., 2013. Analysing small-scale aggregation in animal visits in space and time: the ST-BBD method. Anim. Behav. 85, 483–492. Burman, M., Pye, A.E., 1980. Neoaplectana carpocapsae: movement of nematode populations on a thermal gradient. Exp. Parasitol. 49, 258–265. Byers, J.A., Poinar Jr., G.O., 1982. Location of insect hosts by the nematode Neoaplectana carpocapsae, in response to temperature. Behaviour 79, 1–10. Campbell, J.F., Kaya, H.K., 1999. How and why a parasitic nematode jumps. Nature 397, 485–486. Campbell, J.F., Lewis, E., Yoder, F., Gaugler, R., 1996. Entomopathogenic nematode (Heterorhabditidae and Steinernematidae) spatial distribution in turfgrass. Parasitology 113, 473–482.
53
Campbell, J.F., Orza, G., Yoder, F., Lewis, E., Gaugler, R., 1998. Spatial and temporal distribution of endemic and released entomopathogenic nematode populations in turfgrass. Entomol. Exp. Appl. 86, 1–11. Clark, C.W., Mangel, M., 1986. The evolutionary advantages of group foraging. Theor. Popul. Biol. 30, 45–75. Convey, P., Pugh, P.J.A., Jackson, C., Murray, W., Ruhland, C.T., Xiong, F.S., Day, T.A., 2002. Response of Antarctic terrestrial microarthropods to long-term climate manipulations. Ecology 83, 3130–3140. Cook, Z., Franks, D.W., Robinson, E.J.H., 2013. Exploration versus exploitation in polydomous ant colonies. J. Theor. Biol. 323, 49–56. Croll, N.A., Matthews, B.E., 1977. Biology of Nematodes. John Wiley and Sons, New York, NY. De Vos, A., O’Riain, M.J., 2013. Movement in a selfish seal herd: do seals follow simple or complex movement rules? Behav. Ecol. 24, 190–197. El-Borai, F.E., Campos-Herrera, R., Stuart, R.J., Duncan, L.W., 2011. Substrate modulation, group effects and the behavioral responses of entomopathogenic nematodes to nematophagous fungi. J. Invertebr. Pathol. 106, 347–356. Fourcassié, V., Dussutour, A., Deneubourg, J.L., 2010. Ant traffic rules. J. Exp. Biol. 213, 2357–2363. Fushing, H., Zhu, L., Shapiro-Ilan, D.I., Campbell, J.F., Lewis, E.E., 2008. State-space based mass event-history model I: many decision-making agents with one target. Ann. Appl. Stat. 2, 1503–1522. Giller, P.S., 1996. The diversity of soil communities, the ‘poor man’s tropical rainforest’. Biodivers. Conserv. 5, 135–168. Grewal, P.S., Lewis, E.E., Gaugler, R., Campbell, J.F., 1994. Host finding behaviour as a predictor of foraging strategy in entomopathogenic nematodes. Parasitology 108, 207–215. Grewal, P.S., Lewis, E.E., Gaugler, R., 1997. Response of infective stage parasites (Nematoda: Steinernematidae) to volatile cues from infected hosts. J. Chem. Ecol. 23, 503–515. Hay, D.B., Fenlon, J.S., 1995. A modified binomial model that describes the infection dynamics of the entomopathogenic nematode Steinernema feltiae (Steinernematidae; Nematoda). Parasitology 111, 627–633. Ilan, T., Kim-Shapiro, D.B., Bock, C.H., Shapiro-Ilan, D.I., 2013. The impact of magnetic fields, electric fields and current on the directional movement of Steinernema carpocapsae. Int. J. Parasitol. 43, 781–784. Jabbour, R., Barbercheck, M.E., 2008. Soil and habitat complexity effects on movement of the entomopathogenic nematode Steinernema carpocapsae in maize. Biol. Control 47, 235–243. Kaplan, F., Alborn, H.T., von Reuss, S.H., Ajredini, R., Ali, J.G., Akyazi, F., Stelinski, L.L., Edison, A.S., Schroeder, F.C., Teal, P.E., 2012. Interspecific nematode signals regulate dispersal behavior. PLoS ONE 7, e38735. http://dx.doi.org/10.1371/ journal.pone.0038735. Kaur, R., Anoop, K., Sumana, A., 2012. Leaders follow leaders to reunite the colony: relocation dynamics of an Indian queenless ant in its natural habitat. Anim. Behav. 83, 1345–1353. Kaya, H.K., Gaugler, R., 1993. Entomopathogenic nematodes. Annu. Rev. Entomol. 38, 181–206. Kaya, H.K., Stock, S.P., 1997. Techniques in insect nematology. In: Lacey, L. (Ed.), Manual of Techniques in Insect Pathology. Academic Press, San Diego, pp. 281– 324. Koppenhöfer, A.M., Kaya, H.K., Shanmugam, S., Wood, G.L., 1995. Interspecific competition between steinernematid nematodes within an insect host. J. Invertebr. Pathol. 66, 99–103. Koppenhöfer, A.M., Grewal, P.S., Fuzy, E.M., 2006. Virulence of the entomopathogenic nematodes Heterorhabditis bacteriophora, Heterorhabditis zealandica, and Steinernema scarabaei against five white grub species (Coleoptera: Scarabaeidae) of economic importance in turfgrass in North America. Biol. Control 38, 397–404. Krebs, C.J., 1999. Ecological Methodology, second ed. Benjamin Cummings, Menlo Park, California. Lacey, L.A., Shapiro-Ilan, D.I., 2008. Microbial control of insect pests in temperate orchard systems: potential for incorporation into IPM. Annu. Rev. Entomol 53, 121–144. Lewis, E.E., 2002. Behavioral ecology. In: Gaugler, R. (Ed.), Entomopathogenic Nematology. CABI, New York, pp. 205–224. Lewis, E.E., Gaugler, R., Harrison, R., 1992. Entomopathogenic nematode host finding: response to host contact cues by cruise and ambush foragers. Parasitology 105, 309–315. Lewis, E.E., Gaugler, R., Harrison, R., 1993. Response of cruiser and ambusher entomopathogenic nematodes (Steinernematidae) to host volatile cues. Can. J. Zool. 71, 765–769. Li, X.-Y., Cowles, R.S., Cowles, E.A., Gaugler, R., Cox-Foster, D.L., 2007. Relationship between the successful infection by entomopathogenic nematodes and the host immune response. Int. J. Parasitol. 37, 365–374. Ludewig, A.H., Schroeder, F.C., 2013. Ascaroside signaling in C. elegans. WormBook, ed. The C. elegans Research Community, WormBook, Available from:, . Nguyen, K.B., Shapiro-Ilan, D.I., Fuxa, J.R., Wood, B.W., Bertolotti, M.A., Adams, B.J., 2006. Taxonomic and biological characterization of Steinernema rarum found in the Southeastern United States. J. Nematol. 38, 28–40. Nøttestad, L., Fernö, A., Axelsen, B.E., 2002. Digging in the deep: killer whales’ advanced hunting tactic. Polar Biol. 25, 939–941. Patterson, B.R., Messier, F., 2001. Social organization and space use of coyotes in eastern Canada relative to prey distribution and abundance. J. Mammal. 82, 463–477.
54
D.I. Shapiro-Ilan et al. / International Journal for Parasitology 44 (2014) 49–54
Perony, N., Townsend, S.W., 2013. Why did the meerkat cross the road? Flexible adaptation of phylogenetically-old behavioural strategies to modern-day threats. PLoS ONE 8, e52834. Pye, A.E., Burman, M., 1981. Neoaplectana carpocapsae: Nematode accumulations on chemical and bacterial gradients. Exp. Parasitol. 51, 13–20. Reeve, J.D., Cronin, J.T., 2010. Edge behaviour in a minute parasitic wasp. J. Anim. Ecol. 79, 483–490. Shapiro-Ilan, D.I., Gaugler, R., 2002. Production technology for entomopathogenic nematodes and their bacterial symbionts. J. Ind. Microbiol. Biotech. 28, 137– 146. Shapiro, D.I., Lewis, E.E., 1999. A comparison of entomopathogenic nematode infectivity from infected hosts versus aqueous suspension. Environ. Entomol. 28, 907–911. Shapiro-Ilan, D.I., Stuart, R.J., McCoy, C.W., 2006. A comparison of entomopathogenic nematode longevity in soil under laboratory conditions. J. Nematol. 38, 119–129. Shapiro, D.I., Lewis, E.E., Paramasivam, S., McCoy, C.W., 2000. Nitrogen partitioning in Heterorhabditis bacteriophora-infected hosts and the effects of nitrogen on attraction/repulsion. J. Invertebr. Pathol. 76, 43–48. Shapiro-Ilan, D.I., Campbell, J.F., Lewis, E.E., Kim-Shapiro, D.B., 2012. Directional movement of entomopathogenic nematodes in response to electrical field:
effects of species, magnitude of voltage, and infective juvenile age. J. Invertebr. Pathol. 109, 34–40. Spiridonov, S.E., Moens, M., Wilson, M.J., 2007. Fine scale spatial distributions of two entomopathogenic nematodes in a grassland soil. Appl. Soil Ecol. 37, 192–201. Srinivasan, M., Grant, W.E., Swannack, T.M., Rajan, J., 2010. Behavioral games involving a clever prey avoiding a clever predator: an individual-based model of dusky dolphins and killer whales. Ecol. Model. 221, 2687–2698. Srinivasan, J., von Reuss, S.H., Bose, N., Zaslaver, A., Mahanti, P., Ho, M.C., O’Doherty, O.G., Edison, A.S., Sternberg, P.W., Schroeder, F.C., 2012. A Modular library of small molecule signals regulates social behaviors in Caenorhabditis elegans. PLoS Biol. 10 (1), e1001237. http://dx.doi.org/10.1371/journal.pbio.1001237. Stuart, R.J., Gaugler, R., 1994. Patchiness in populations of entomopathogenic nematodes. J. Invertebr. Pathol. 64, 39–45. Stuart, R.J., Barbercheck, M.E., Grewal, P.S., Taylor, R.A.J., Hoy, C.W., 2006. Population biology of entomopathogenic nematodes: concepts, issues, and models. Biol. Control 38, 80–102. Torr, P., Heritage, S., Wilson, M.J., 2004. Vibrations as a novel signal for host location by parasitic nematodes. Int. J. Parasitol. 34, 997–999.
Contents lists available at ScienceDirect
International Journal for Parasitology journal homepage: www.elsevier.com/locate/ijpara
Aggregative group behavior in insect parasitic nematode dispersal David I. Shapiro-Ilan a,⇑, Edwin E. Lewis b, Paul Schliekelman c,⇑ a
USDA-ARS, SEFTNRL, 21 Dunbar Road, Byron, GA 31008, USA University of California Davis, Department of Nematology and Entomology, University of California, Davis, CA 95616, USA c Department of Statistics, University of Georgia, Athens, GA 30602, USA b
a r t i c l e
i n f o
Article history: Received 26 July 2013 Received in revised form 1 October 2013 Accepted 5 October 2013 Available online 31 October 2013 Keywords: Behavior Dispersal Entomopathogenic nematode Group foraging Heterorhabditis Steinernema
a b s t r a c t Movement behavior of foraging animals is critical to the determination of their spatial ecology and success in exploiting resources. Individuals sometimes gain advantages by foraging in groups to increase their efficiency in garnering these resources. Group movement behavior has been studied in various vertebrates. In this study we explored the propensity for innate group movement behavior among insect parasitic nematodes. Given that entomopathogenic nematodes benefit from group attack and infection, we hypothesised that the populations would tend to move in aggregate in the absence of extrinsic cues. Movement patterns of entomopathogenic nematodes in sand were investigated when nematodes were applied to a specific locus or when the nematodes emerged naturally from infected insect hosts; six nematode species in two genera were tested (Heterorhabditis bacteriophora, Heterorhabditis indica, Steinernema carpocapsae, Steinernema feltiae, Steinernema glaseri and Steinernema riobrave). Nematodes were applied in aqueous suspension via filter paper discs or in infected insect host cadavers (to mimic emergence in nature). We discovered that nematode dispersal resulted in an aggregated pattern rather than a random or uniform distribution; the only exception was S. glaseri when emerging directly from infected hosts. The group movement may have been continuous from the point of origin, or it may have been triggered by a propensity to aggregate after a short period of random movement. To our knowledge, this is the first report of group movement behavior in parasitic nematodes in the absence of external stimuli (e.g., without an insect or other apparent biotic or abiotic cue). These findings have implications for nematode spatial distribution and suggest that group behavior is involved in nematode foraging. Published by Elsevier Ltd. on behalf of Australian Society for Parasitology Inc.
1. Introduction Behaviors involving movement are intimately tied to the spatial distribution and foraging success of most animals. When animals increase their individual fitness by exploiting resources in groups, the spatial distribution of their populations should be overdispersed. There are many examples of group foraging behavior, or cooperative behaviors, demonstrated by organisms of diverse trophic levels (Nøttestad et al., 2002; Fourcassié et al., 2010; Bonnell et al., 2013; Cook et al., 2013). Various biotic and abiotic factors can affect movement behavior (Perony and Townsend, 2013) such as group size (Bonnell et al., 2013), environmental spatial structure and resource availability (Patterson and Messier, 2001; Reeve and Cronin, 2010), and avoidance of predators (Srinivasan et al., 2010; De Vos and O’Riain, 2013). But if a massattack strategy is necessary for efficient garnering of resources, then we would hypothesise that even in the absence of these factors, populations should remain overdispersed, and thus their ⇑ Corresponding authors. Tel.: +1 478 956 6441; fax: +1 478 956 2929. E-mail addresses: [email protected] (D.I. Shapiro-Ilan), pdschlie@ uga.edu (P. Schliekelman).
movement should be en masse. In this study we investigated the group movement and aggregation behavior of insect parasitic nematodes in an arena that was environmentally homogeneous and devoid of resources. Some species follow temporal patterns of random movement while foraging whereas others move in directed spatial patterns, alone or in aggregate (Bonnell et al., 2013). Here we are interested in how movement in groups relates to the behavior and fitness of the individual. For group behaviors to persist in any animal population there must be fitness gains for the individuals who form the group; the individuals within the group must win some advantage over individuals on their own (Clark and Mangel, 1986). This is especially true in groups composed of unrelated or distantly related individuals. These advantages potentially include increased probability of finding resources, better ability to overcome defensive actions of hosts or prey, or the mitigation of risk posed by engaging with the resource. Entomopathogenic nematodes in the genera Heterorhabditis and Steinernema (the subjects of this study) are obligate parasites of insects. These nematodes are important natural regulators of insect populations and are also applied as biological control agents to suppress pest populations (Lacey and Shapiro-Ilan, 2008). The
0020-7519/$36.00 Published by Elsevier Ltd. on behalf of Australian Society for Parasitology Inc. http://dx.doi.org/10.1016/j.ijpara.2013.10.002
50
D.I. Shapiro-Ilan et al. / International Journal for Parasitology 44 (2014) 49–54
developmentally arrested infective juvenile nematodes (IJs), the only free-living stage, typically occupy soil habitats until they infect an insect, which they must do in order to resume development and reproduce. Generally, an insect host is infected by tens to hundreds of individual IJs, and a single IJ is seldom able to initiate a successful infection (Koppenhöfer et al., 2006; Nguyen et al., 2006), which confers a fitness advantage to individuals that invade hosts as part of a group. A few hours after the nematodes enter the host, the nematodes release symbiotic bacteria, which are the primary agents in killing the host (Kaya and Gaugler, 1993). Infected hosts generally die within 24–72 h, and the nematodes feed on the symbiotic bacteria and insect tissues, reproducing for one to three generations (Kaya and Gaugler, 1993). As food resources dwindle, a new cycle of IJs is produced and they emerge to seek new hosts. The number of IJs exiting a host can range from tens of thousands to hundreds of thousands (Shapiro-Ilan and Gaugler, 2002). Foraging strategies among entomopathogenic nematode species vary along a continuum between ambushers, which generally sit and wait for a passing host, and cruisers that actively search for hosts (Lewis et al., 1992). Entomopathogenic nematodes can exhibit a combination of these behaviors to locate hosts and although some species exhibit primarily ambush-type behaviors and others are mainly cruisers, others are considered intermediate in their foraging behavior (Campbell and Kaya, 1999; Lewis, 2002). Foraging strategies may contribute to spatial distribution of entomopathogenic nematode populations in the environment. Several studies indicate that the distribution of entomopathogenic nematode populations is aggregated in nature (Stuart and Gaugler, 1994; Campbell et al., 1996). The clumped distribution may largely be due to a sparse distribution of infection loci and subsequent mass emergence from host cadavers, thus creating nematode hot spots within the soil ecosystem (Campbell et al., 1996; Spiridonov et al., 2007). However, we suggest that aggregate distributions may also result from group nematode movement when dispersing from hosts or points of application. Group movement of entomopathogenic nematodes may have adaptive value, e.g., in maximising chances for successful host infection. Similar to some other predators and parasites, group attack is generally required by entomopathogenic nematodes to overcome host defenses (Li et al., 2007). Although entomopathogenic nematode dispersal has been studied in terms of biotic and abiotic factors that affect movement, or distance travelled within the soil profile (Stuart et al., 2006; Jabbour and Barbercheck, 2008; El-Borai et al., 2011), there is a dearth of knowledge on more general patterns of movement, especially when biotic and abiotic drivers of movement are not confounding results. Recently, Fushing et al. (2008) described mathematically the temporal aspects of group infection behavior in certain entomopathogenic nematode species. In a similar fashion, we hypothesise that entomopathogenic nematodes also display group movement behavior, even in the absence of hosts. In this study that hypothesis was tested with six species of entomopathogenic nematodes.
Nematodes were applied to experimental arenas using two approaches: application in infected host cadavers or in aqueous suspension. The cadaver approach was intended to represent natural emergence from the host, whereas the aqueous suspension method was intended to mimic artificial introduction of nematodes to the environment for purposes of biological pest control. Experimental arenas consisted of polypropylene boxes (20.5 20.5 cm, 9 cm deep) containing sand filled to a depth of 2.5 cm. In all arenas, the final total moisture level in sand was at approximate field capacity (10%). For the cadaver application approach, three nematode-infected hosts (at 3 days p.i.) were placed in the center of each arena. The cadavers were observed daily for the initiation of nematode emergence. The number of IJs that emerged per cadaver was estimated based on separate yield assessments (Kaya and Stock, 1997) conducted on 10 replicate insects each. The results indicated average (±S.E.M.) IJ production by the species as follows: H. bacteriophora produced 244,390 ± 28,419.1 IJs; H. indica produced 112,210 ± 11,183.9 IJs; S. carpocapsae produced 178,300 ± 15,507.9 IJs; S. feltiae produced 153,700 ± 14,272 IJs; S. glaseri produced 76,650 ± 5,299.9IJs; S. riobrave produced 126,080 ± 10,788.9 IJs. For application of aqueous suspensions, 200,000 IJs from culture flasks (except 300,000 IJs were used for S. riobrave) were vacuum filtered onto filter paper (60 mm, Whatman No. 1); the filter paper was then placed upside down in the center of a sand arena. Arenas were stored inside plastic bags at 25 °C until nematode movement was assessed. The plastic bags contained moist paper towels to ensure that high levels of relative humidity (approximately 100%) were maintained. There were four replicate arenas for each species and application approach. Experiments addressing movement of each nematode species were conducted separately, and all experiments were repeated once in time (i.e., run two times in total) using different batches of nematodes and insects. In arenas with host cadavers, the results of nematode movement were assessed 3 days after IJ emergence began. In arenas receiving aqueous nematode applications, resulting nematode movements were assessed 3 days post-treatment. At the time of assessment, cadavers and filter papers were removed and soil cores measuring 2 cm diameter and 2.5 cm depth were taken at four points that were 4.5 cm from the center of the arena (diagonally and equidistant from each other). The sand from each core was then placed in 100 ml of tap water in a 250 ml beaker, and the number of nematodes in the core was determined by dilution counts under a stereomicroscope (Shapiro-Ilan et al., 2006). For the aqueous suspension approach, the total number of IJs entering sand was also estimated by washing any remaining IJs from the filter paper and subtracting that number from the number originally applied. 2.2. Analysis of movement patterns
2.1. Experimental approach
Our goal was to determine whether there is evidence that the post-emergence or post-application distribution of nematodes indicate aggregated movement patterns. The null hypothesis was that the nematodes move independently versus the alternate hypothesis that they tend to aggregate. Two different approaches to the analysis were taken.
All nematodes were cultured in vivo at 25 °C in commercially obtained final instar Galleria mellonella according to procedures described in Kaya and Stock (1997). Nematode species included Heterorhabditis bacteriophora (Oswego strain), Heterorhabditis indica (HOM1 strain), Steinernema carpocapsae (All strain), Steinernema feltiae (SN strain), Steinernema glaseri (NJ43 strain) and Steinernema riobrave (355 strain). The nematodes were stored in culture flasks at 13 °C for less than 2 weeks prior to experimentation.
2.2.1. First approach: Index of Dispersion If the null hypothesis is further specified to be that the nematodes undergo independent diffusive movement, then the number of IJs that is found in a core of a fixed size should follow a Poisson distribution with a mean proportional to the area of the core. The data for each trial consisted of the four cores of equal size that were equidistant from the center. Thus, under the null hypothesis each core follows a Poisson distribution with the same mean.
2. Materials and methods
D.I. Shapiro-Ilan et al. / International Journal for Parasitology 44 (2014) 49–54
The Poisson distribution has the property that its mean is the same as its variance. Thus, if the null hypothesis is correct, then the mean of the counts across the four cores should be approximately equal to the variance of counts across the quadrants. The ratio of the variance to the mean across the cores is known as the Index of Dispersion (see Krebs, 1999 for overview). Thus, an index greater than one indicates more aggregation than expected under the null hypothesis; less than one indicates a more uniform distribution than expected under the null hypothesis. If n is the number of cores and D is the Index of Dispersion, then (n 1)D has a chi-square distribution under the null hypothesis. A significant deviation of this quantity from its null hypothesis expectation indicates that the counts do not follow a Poisson distribution, giving evidence of aggregative movement. This test was applied to each replicate in each combination of the conditions separately. 2.2.2. Second approach: simulation approach for calculating P-values We have four replicates of each set of experimental conditions. The Index of Dispersion approach described above is a classical approach to the problem of testing for deviations from independent diffusive movement. However, it does not allow combining data over multiple replicates. Thus, we calculated the values separately for each replicate in our data. In order to do a joint test over replicates, we used a test statistic consisting of the product of the Index of Dispersion across replicates and used a simulation approach to calculate P values. We wrote an R program that generated 10,000 simulated null hypothesis data sets for each set of experimental conditions. Under the null hypothesis, the counts in each core follow a Poisson distribution. Each simulated data set consisted of four sets of four counts (one for each core) generated from a Poisson distribution, corresponding to the four replicates with four cores each. The dispersion statistics were calculated for each group of four cores and the test statistic was the product of the dispersion statistics across the four replicate data sets. The simulation P value is thus the proportion of the 10,000 simulated data sets in which the product of D across the four replicates was greater than the product observed in the actual data. A third approach for calculating P values over replicates based on the multinomial distribution is shown in Supplementary Table S2. This analysis was only conducted for the aqueous suspension experiments because the specific number of nematodes in each arena was not known for the cadaver experiments. There was good agreement between these results and the results above. 3. Results Table 1 shows the Index of Dispersion calculated for each arena in the study. This has an expected value of 1 for independent diffusive movement and larger values indicate evidence of aggregative movement. Values greater than 2.6 show statistically significant deviation (a = 0.05) from the null expectation (P-values are given in Supplementary Table S1). The majority of arenas indicated aggregative movement (Fig. 1). In most treatments, aggregative movement was indicated in >60% of the arenas (Fig. 1) and for five treatments aggregative movement was observed in P87.5% of the of the arenas (Fig. 1). In contrast, for the S. glaseri-cadaver treatments none of the arenas showed significant departure from random movement (Fig. 1; Supplementary Table S1). The S. feltiaecadaver treatment and the S. glaseri-filter paper treatments both only indicated aggregative movement in 50% of the arenas. Next, we analysed nematode distribution across replicates for both the filter paper and cadaver applications. We evaluated overall evidence for aggregative movement using the product of the Index of Dispersion over the four replicates for each experimental condition and simulated P values as described in the Section 2.2.2.
51
These results also indicated aggregated movement in most of the analyses (Table 1; Supplementary Table S1). Specifically, 20 of the 24 analyses indicated non-random movement (Table 1). Particularly high levels of aggregation were observed in some treatments such as H. bacteriophora and H. indica when applied in host cadavers. Neither of the S. glaseri-cadaver trials indicated that non-random movement had taken place, and only one of the two trials was significantly aggregated for the S. feltiae-cadaver and H. bacteriophora-filter paper treatments (Table 1). The outcome of the third approach (Supplementary Table S2) shows close agreement between the P values calculated by simulation and those calculated by the method of the first approach. This is evidence that both sets of P values are being calculated correctly. In summary, evidence for aggregative movement was observed in all treatments except in the S. glaseri-cadaver treatment, for which none of the statistical tests suggested non-random movement. In two other treatments, evidence for aggregative movement was detected although the effect was less pronounced than in other treatments. Specifically, aggregative movement was lower in the S. feltiae-cadaver treatment, for which only 50% of individual arenas exhibited a significant effect in the first statistical approach, and only one of the two trials exhibited a significant effect in the second approach. Additionally, evidence for aggregative movement was somewhat lower for the H. bacteriophora-filter paper treatment, as only one of the two trials indicated significance in the second statistical approach, yet aggregative movement was detected in 62.5% of the individual arenas using the first statistical approach.
4. Discussion Our findings provided evidence for aggregated group behavior in entomopathogenic nematode movement (in all treatments except in the S. glaseri-cadaver treatment). Although the study included an array of nematode species, all treatments exhibited aggregated movement patterns regardless of genus (Heterorhabditis or Steinernema) or foraging type (ambusher, cruiser or intermediate). The movement patterns resulted in a patchy nematode distribution in the sand arenas that was not due to the nematodes responding to any extrinsic stimulus. Similar to most soil organisms, entomopathogenic nematodes are in fact generally found in a clumped distribution in nature (Stuart and Gaugler, 1994; Giller, 1996; Campbell et al., 1996, 1998; Convey et al., 2002; Spiridonov et al., 2007). Even when applied to the soil environment in a uniform distribution, soon after application, entomopathogenic nematodes tend to be found in a patchy distribution (Campbell et al., 1998). A number of extrinsic factors have been postulated to contribute to patchiness in entomopathogenic nematode distributions including a patchy distribution of hosts, mass emergence from infected cadavers, variable persistence due to biotic and abiotic factors in the soil (following emergence) and local extinctions (Stuart and Gaugler, 1994; Campbell et al., 1998; Stuart et al., 2006). Based on our findings, we contend that aggregated movement behavior may further contribute to clumping of natural and applied entomopathogenic nematode populations. The possibility of group movement in S. riobrave was suggested by El-Borai et al. (2011) after observing that IJs tended to move toward one side in columns of sand that were supposedly uniform; in the current study we have demonstrated experimentally that entomopathogenic nematodes do indeed move in aggregate patterns, thereby providing an explanation for the previous observation. The group movement behavior observed in this study may have been continuous from the point of origin, or it may have been triggered by a propensity to aggregate after a short period of random movement. Given that we only measured movement from the inoculation site to one location within each quadrant, we cannot
52
D.I. Shapiro-Ilan et al. / International Journal for Parasitology 44 (2014) 49–54
Table 1 Statistical values resulting for entomopathogenic nematode movement in a sand arena. Trial
N
Test
D1
D2
D3
D4
Dprod
P
1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2
Sr Sr Sr Sr Sf Sf Sf Sf Sg Sg Sg Sg Sc Sc Sc Sc Hi Hi Hi Hi Hb Hb Hb Hb
FP FP Cad Cad FP FP Cad Cad FP FP Cad Cad FP FP Cad Cad FP FP Cad Cad FP FP Cad Cad
5.58 16.86 4.99 16.38 14,971.7 12.44 0.53 1.08 2.42 5.33 1.00 0.67 3.92 5.06 1.62 8.96 14.47 2.31 69.78 11,860.3 69.30 0.70 54.70 14,561.2
9.55 3.71 0.89 0.93 32.38 8.75 4.45 3.22 2.27 2.13 1.59 0.20 9.38 2.18 33.32 11.72 7.59 1.79 16.56 13,314.84 15.07 3.44 47.54 25,998.64
25.78 27.93 20.07 17.95 5.43 2.75 0.67 25.81 4.20 0.86 2.00 0.52 4.99 6.50 1.83 2.25 8.71 4.56 37.91 28,410.70 40.86 1.11 23,284.40 13,158.79
3.17 3.99 11.95 14,709.41 2.01 2.84 0.33 3.42 5.02 3.21 0.20
4,347.70 6,983.86 1,062.91 4,038,821.73 5,300,759.70 849.72 0.53 306.49 115.91 31.25 0.64 0.07 642.64 1,712.30 548.14 879.34 708.28 205.22 534,900,612.34 70,454,974,403,547,160.00 1,063,559,253.61 0.90 1,017,372,714.10 73,264,704,285,453.59
0.00 0.00 0.00 0.00 0.00 0.00 0.36 0.00 0.00 0.00 0.32 0.84 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.26 0.00 0.00
3.50 23.88 5.54 3.72 0.74 10.84 12,213.71 15,703.57 24,935.13 0.33 16.80 14.71
Nematodes (N) were applied to sand arenas on filter paper (FP) or in infected Galleria mellonella cadavers (Cad). Soil cores were taken to determine nematode distribution within the arena. Each D product and P value is derived from the combination of four replicates; the experiments were repeated (two trials). When P is listed as zero this means that it was less than 10 324. This table reports the result from the second statistical approach based on simulation of P values (see Section 2.2.2. in the text and Supplementary Table S1 for details). A statistically significant value indicates an aggregated distribution. Hb, Heterorhabditis bacteriophora; Hi, Heterorhabditis indica; Sc, Steinernema carpocapsae; Sf, Steinernema carpocapsae; Sg, Steinernema glaseri; Sr, Steinernema riobrave.
Fig. 1. Percentage of arenas indicating a significant departure from random movement by entomopathogenic nematodes (thus indicating aggregative movement). Nematodes were applied to sand arenas on filter paper (Fp) or in infected Galleria mellonella cadavers (Cad). Soil cores were taken to determine nematode distribution within the arena. Hb, Heterorhabditis bacteriophora; Hi, Heterorhabditis indica; Sc, Steinernema carpocapsae; Sf, Steinernema feltiae; Sg, Steinernema glaseri; Sr, Steinernema riobrave.
rule out the possibility that the nematodes initially moved randomly, but subsequently (within the 3 day interval) moved in group fashion to form an aggregate pattern. Aggregation behavior, such as in response to food sources, has been observed in the nematode Caenorhabditis elegans (Srinivasan et al., 2012; Ludewig and Schroeder, 2013). In the current study, one might argue that continuous group movement was more likely because random movement followed by aggregation would result in unnecessary energy expenditure and would not provide any obvious advantage (such as coming to a food source). Nonetheless, additional research is needed to elucidate whether the group behavior is continuous or temporal in nature. Movement behavior in many animals (including nematodes) is directionally influenced by various stimuli such as physical or chemical gradients and environmental structure (Boender et al., 2011; Bonnell et al., 2013; Perony and Townsend, 2013). Entomopathogenic nematodes respond to a variety of stimuli such as CO2
(Lewis et al., 1993; Lewis, 2002), vibration (Torr et al., 2004), temperature (Burman and Pye, 1980; Byers and Poinar, 1982), chemical compounds (Pye and Burman, 1981; Shapiro et al., 2000), and electromagnetic stimuli (Shapiro-Ilan et al., 2012; Ilan et al., 2013). Aggregative movement behavior has been observed in certain nematode species in response to certain stimuli such as adverse environmental conditions (e.g., to avoid desiccation) (Croll and Matthews, 1977). However, in the current study, arenas were arranged such that all potential internal or external stimuli were held constant under conditions conducive to nematode survival and therefore environmental factors did not cause bias in nematode movement. Accordingly, we conclude that the group movement behavior observed was not due to a response to stimuli, but reflects the nematodes’ natural propensity to aggregate. To our knowledge, this is the first report of group movement behavior in parasitic nematodes in the absence of external stimuli (e.g., without an insect or other apparent biotic or abiotic cue). Furthermore, our observations were made in the absence of hosts. Future studies will investigate the dynamics of entomopathogenic nematode group movement behavior in the presence of varying biotic and abiotic stimuli. Entomopathogenic nematodes exhibit group infection behavior (Hay and Fenlon, 1995; Fushing et al., 2008). Infection behavior was postulated to be consistent with risk sensitive foraging patterns (Fushing et al., 2008). In this ‘‘follow-the-leader’’ or herding paradigm, risk-prone nematodes infect the host, followed by nematodes that are more risk averse (Fushing et al., 2008). Followthe-leader behavior has been observed in group movement patterns of both vertebrates and invertebrates (Kaur et al., 2012; Perony and Townsend, 2013). Conceivably, the aggregative group movement observed in this study may also be due to a propensity toward herding or follow-the-leader behavior. Group movement behavior combined with group infection has adaptive value for individual group members in terms of enhancing potential for foraging success, i.e., resulting in a greater chance of finding a host in sufficient numbers to successfully infect and overcome the host
D.I. Shapiro-Ilan et al. / International Journal for Parasitology 44 (2014) 49–54
immune system. In the case that individual IJs come into contact with a host, even if the host supports reproduction, without other IJs to follow on in infection, these individuals are unlikely to have reproductive success. Thus, natural selection should favor individuals that are part of a group over individuals on their own. The causal mechanisms behind the group dispersal and aggregation behavior observed in this study are unknown. Kaplan et al. (2012) reported that ascarosides caused dispersal in several entomopathogenic nematode species as well as other nematode species. Indeed, the impact of ascarosides on aggregation and other behaviors in C. elegans has been reported in a number of studies (Srinivasan et al., 2012; Ludewig and Schroeder, 2013). Conceivably, similar compounds could signal aggregative group movement behavior in entomopathogenic nematodes. It is interesting that S. glaseri, when applied in an infected host, was the only treatment that did not exhibit aggregative movement. The reasons for this finding are unclear, more so because the nematode did exhibit aggregative movement when applied in aqueous suspension via filter paper. Given that chemical factors associated with the nematode-infected host can elicit nematode behaviors such as infection or dispersal (Shapiro and Lewis, 1999; Kaplan et al., 2012), and these factors and their impact can be diluted when nematodes are placed in aqueous suspension (Shapiro and Lewis, 1999), it is possible that compounds associated with the S. glaseri-infected cadaver differed in quality or quantity from those present in the aqueous treatment. Furthermore, relative to other entomopathogenic nematode species, S. glaseri is highly mobile and competitive (Grewal et al., 1994, 1997; Koppenhöfer et al., 1995), and therefore the need for aggregative behavior in freshly emerging (highly fit) S. glaseri may be reduced. On the other hand, it is possible that the highly mobile nature of S. glaseri IJs results in their aggregations developing on a spatial scale that was not detected in the small arenas used in our experiments; a relatively low average number of IJs were found in S. glaseri-cadaver treatment cores (<10 IJs compared with >35 IJs in all other species), indicating that possibly these nematodes had already moved to the edge of the arena. Additional research is needed to elucidate the basis for group movement behavior in entomopathogenic nematodes.
Acknowledgments We thank Wanda Evans and Kathy Halat for technical assistance.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ijpara.2013.10. 002.
References Boender, A.J., Roubos, E.W., van der Velde, G., 2011. Together or alone? Foraging strategies in Caenorhabditis elegans. Biol. Rev. 86, 853–862. Bonnell, T.R., Dutilleul, P., Chapman, C.A., Reyna-Hurtado, R., Hernández-Sarabia, R.U., Sengupta, R., 2013. Analysing small-scale aggregation in animal visits in space and time: the ST-BBD method. Anim. Behav. 85, 483–492. Burman, M., Pye, A.E., 1980. Neoaplectana carpocapsae: movement of nematode populations on a thermal gradient. Exp. Parasitol. 49, 258–265. Byers, J.A., Poinar Jr., G.O., 1982. Location of insect hosts by the nematode Neoaplectana carpocapsae, in response to temperature. Behaviour 79, 1–10. Campbell, J.F., Kaya, H.K., 1999. How and why a parasitic nematode jumps. Nature 397, 485–486. Campbell, J.F., Lewis, E., Yoder, F., Gaugler, R., 1996. Entomopathogenic nematode (Heterorhabditidae and Steinernematidae) spatial distribution in turfgrass. Parasitology 113, 473–482.
53
Campbell, J.F., Orza, G., Yoder, F., Lewis, E., Gaugler, R., 1998. Spatial and temporal distribution of endemic and released entomopathogenic nematode populations in turfgrass. Entomol. Exp. Appl. 86, 1–11. Clark, C.W., Mangel, M., 1986. The evolutionary advantages of group foraging. Theor. Popul. Biol. 30, 45–75. Convey, P., Pugh, P.J.A., Jackson, C., Murray, W., Ruhland, C.T., Xiong, F.S., Day, T.A., 2002. Response of Antarctic terrestrial microarthropods to long-term climate manipulations. Ecology 83, 3130–3140. Cook, Z., Franks, D.W., Robinson, E.J.H., 2013. Exploration versus exploitation in polydomous ant colonies. J. Theor. Biol. 323, 49–56. Croll, N.A., Matthews, B.E., 1977. Biology of Nematodes. John Wiley and Sons, New York, NY. De Vos, A., O’Riain, M.J., 2013. Movement in a selfish seal herd: do seals follow simple or complex movement rules? Behav. Ecol. 24, 190–197. El-Borai, F.E., Campos-Herrera, R., Stuart, R.J., Duncan, L.W., 2011. Substrate modulation, group effects and the behavioral responses of entomopathogenic nematodes to nematophagous fungi. J. Invertebr. Pathol. 106, 347–356. Fourcassié, V., Dussutour, A., Deneubourg, J.L., 2010. Ant traffic rules. J. Exp. Biol. 213, 2357–2363. Fushing, H., Zhu, L., Shapiro-Ilan, D.I., Campbell, J.F., Lewis, E.E., 2008. State-space based mass event-history model I: many decision-making agents with one target. Ann. Appl. Stat. 2, 1503–1522. Giller, P.S., 1996. The diversity of soil communities, the ‘poor man’s tropical rainforest’. Biodivers. Conserv. 5, 135–168. Grewal, P.S., Lewis, E.E., Gaugler, R., Campbell, J.F., 1994. Host finding behaviour as a predictor of foraging strategy in entomopathogenic nematodes. Parasitology 108, 207–215. Grewal, P.S., Lewis, E.E., Gaugler, R., 1997. Response of infective stage parasites (Nematoda: Steinernematidae) to volatile cues from infected hosts. J. Chem. Ecol. 23, 503–515. Hay, D.B., Fenlon, J.S., 1995. A modified binomial model that describes the infection dynamics of the entomopathogenic nematode Steinernema feltiae (Steinernematidae; Nematoda). Parasitology 111, 627–633. Ilan, T., Kim-Shapiro, D.B., Bock, C.H., Shapiro-Ilan, D.I., 2013. The impact of magnetic fields, electric fields and current on the directional movement of Steinernema carpocapsae. Int. J. Parasitol. 43, 781–784. Jabbour, R., Barbercheck, M.E., 2008. Soil and habitat complexity effects on movement of the entomopathogenic nematode Steinernema carpocapsae in maize. Biol. Control 47, 235–243. Kaplan, F., Alborn, H.T., von Reuss, S.H., Ajredini, R., Ali, J.G., Akyazi, F., Stelinski, L.L., Edison, A.S., Schroeder, F.C., Teal, P.E., 2012. Interspecific nematode signals regulate dispersal behavior. PLoS ONE 7, e38735. http://dx.doi.org/10.1371/ journal.pone.0038735. Kaur, R., Anoop, K., Sumana, A., 2012. Leaders follow leaders to reunite the colony: relocation dynamics of an Indian queenless ant in its natural habitat. Anim. Behav. 83, 1345–1353. Kaya, H.K., Gaugler, R., 1993. Entomopathogenic nematodes. Annu. Rev. Entomol. 38, 181–206. Kaya, H.K., Stock, S.P., 1997. Techniques in insect nematology. In: Lacey, L. (Ed.), Manual of Techniques in Insect Pathology. Academic Press, San Diego, pp. 281– 324. Koppenhöfer, A.M., Kaya, H.K., Shanmugam, S., Wood, G.L., 1995. Interspecific competition between steinernematid nematodes within an insect host. J. Invertebr. Pathol. 66, 99–103. Koppenhöfer, A.M., Grewal, P.S., Fuzy, E.M., 2006. Virulence of the entomopathogenic nematodes Heterorhabditis bacteriophora, Heterorhabditis zealandica, and Steinernema scarabaei against five white grub species (Coleoptera: Scarabaeidae) of economic importance in turfgrass in North America. Biol. Control 38, 397–404. Krebs, C.J., 1999. Ecological Methodology, second ed. Benjamin Cummings, Menlo Park, California. Lacey, L.A., Shapiro-Ilan, D.I., 2008. Microbial control of insect pests in temperate orchard systems: potential for incorporation into IPM. Annu. Rev. Entomol 53, 121–144. Lewis, E.E., 2002. Behavioral ecology. In: Gaugler, R. (Ed.), Entomopathogenic Nematology. CABI, New York, pp. 205–224. Lewis, E.E., Gaugler, R., Harrison, R., 1992. Entomopathogenic nematode host finding: response to host contact cues by cruise and ambush foragers. Parasitology 105, 309–315. Lewis, E.E., Gaugler, R., Harrison, R., 1993. Response of cruiser and ambusher entomopathogenic nematodes (Steinernematidae) to host volatile cues. Can. J. Zool. 71, 765–769. Li, X.-Y., Cowles, R.S., Cowles, E.A., Gaugler, R., Cox-Foster, D.L., 2007. Relationship between the successful infection by entomopathogenic nematodes and the host immune response. Int. J. Parasitol. 37, 365–374. Ludewig, A.H., Schroeder, F.C., 2013. Ascaroside signaling in C. elegans. WormBook, ed. The C. elegans Research Community, WormBook, Available from:
54
D.I. Shapiro-Ilan et al. / International Journal for Parasitology 44 (2014) 49–54
Perony, N., Townsend, S.W., 2013. Why did the meerkat cross the road? Flexible adaptation of phylogenetically-old behavioural strategies to modern-day threats. PLoS ONE 8, e52834. Pye, A.E., Burman, M., 1981. Neoaplectana carpocapsae: Nematode accumulations on chemical and bacterial gradients. Exp. Parasitol. 51, 13–20. Reeve, J.D., Cronin, J.T., 2010. Edge behaviour in a minute parasitic wasp. J. Anim. Ecol. 79, 483–490. Shapiro-Ilan, D.I., Gaugler, R., 2002. Production technology for entomopathogenic nematodes and their bacterial symbionts. J. Ind. Microbiol. Biotech. 28, 137– 146. Shapiro, D.I., Lewis, E.E., 1999. A comparison of entomopathogenic nematode infectivity from infected hosts versus aqueous suspension. Environ. Entomol. 28, 907–911. Shapiro-Ilan, D.I., Stuart, R.J., McCoy, C.W., 2006. A comparison of entomopathogenic nematode longevity in soil under laboratory conditions. J. Nematol. 38, 119–129. Shapiro, D.I., Lewis, E.E., Paramasivam, S., McCoy, C.W., 2000. Nitrogen partitioning in Heterorhabditis bacteriophora-infected hosts and the effects of nitrogen on attraction/repulsion. J. Invertebr. Pathol. 76, 43–48. Shapiro-Ilan, D.I., Campbell, J.F., Lewis, E.E., Kim-Shapiro, D.B., 2012. Directional movement of entomopathogenic nematodes in response to electrical field:
effects of species, magnitude of voltage, and infective juvenile age. J. Invertebr. Pathol. 109, 34–40. Spiridonov, S.E., Moens, M., Wilson, M.J., 2007. Fine scale spatial distributions of two entomopathogenic nematodes in a grassland soil. Appl. Soil Ecol. 37, 192–201. Srinivasan, M., Grant, W.E., Swannack, T.M., Rajan, J., 2010. Behavioral games involving a clever prey avoiding a clever predator: an individual-based model of dusky dolphins and killer whales. Ecol. Model. 221, 2687–2698. Srinivasan, J., von Reuss, S.H., Bose, N., Zaslaver, A., Mahanti, P., Ho, M.C., O’Doherty, O.G., Edison, A.S., Sternberg, P.W., Schroeder, F.C., 2012. A Modular library of small molecule signals regulates social behaviors in Caenorhabditis elegans. PLoS Biol. 10 (1), e1001237. http://dx.doi.org/10.1371/journal.pbio.1001237. Stuart, R.J., Gaugler, R., 1994. Patchiness in populations of entomopathogenic nematodes. J. Invertebr. Pathol. 64, 39–45. Stuart, R.J., Barbercheck, M.E., Grewal, P.S., Taylor, R.A.J., Hoy, C.W., 2006. Population biology of entomopathogenic nematodes: concepts, issues, and models. Biol. Control 38, 80–102. Torr, P., Heritage, S., Wilson, M.J., 2004. Vibrations as a novel signal for host location by parasitic nematodes. Int. J. Parasitol. 34, 997–999.