Neotrop Entomol (2015) 44:374–384 DOI 10.1007/s13744-015-0286-5
BIOLOGICAL CONTROL
Intraguild Predation in Heteroptera: Effects of Density and Predator Identity on Dipteran Prey S BRAHMA1, D SHARMA1, M KUNDU2, N SAHA1, GK SAHA1, G ADITYA1,2 1
Dept of Zoology, Univ of Calcutta, Kolkata, India Dept of Zoology, The Univ of Burdwan, Kolkata, India
2
Keywords Biological control, generalist predator, intraguild predation, predator substitutability Correspondence G Aditya, Dept of Zoology, Univ of Calcutta, 35 Ballygunge Circular Road, Kolkata 700019, India;
[email protected] Edited by Fernando L Cônsoli – ESALQ/USP Received 2 September 2014 and accepted 7 March 2015 Published online: 15 April 2015 * Sociedade Entomológica do Brasil 2015
Abstract In tropical freshwaters, different species of water bugs (Heteroptera) constitute a guild sharing similar prey resources including chironomid and mosquito larvae. Assuming possibilities of intraguild predation (IGP) among the constituent members, an attempt was made to evaluate the effects of prey and predator density on the mortality of mosquito and chironomid larvae (shared prey), using Laccotrephes griseus GuérinMéneville (Hemiptera: Nepidae) and Ranatra filiformis Fabricius (Hemiptera: Nepidae) as IG predators and Anisops bouvieri Kirkaldy (Hemiptera: Notonectidae) as IG prey. The predation on mosquito and chironomid larvae varied with the density and combinations of the predators. When present as conspecific IG predators, L. griseus exhibited greater effect on the prey mortality than R. filiformis. The effects on shared prey suggest that the two predators are not substitutable in terms of the effect on the shared prey mortality. The mortality of A. bouvieri (IG prey) at low shared prey density was significantly different (p20 mm in length, 3.3–5.1 mg in wet weight) sorted and placed within smaller trays with little sediment from where they were used for the experimental purposes.
Single predator system (experiment 1) The prey consumption of the water bugs A. bouvieri, R. filiformis, and L. griseus on mosquito larvae and chironomid larvae were noted separately using two levels of prey density: low (50 individuals) and high (200 individuals). The density of the predators was10 individuals for A. bouvieri and 2 individuals for R. filiformis and L. griseus. Eighteen trials were considered for each prey type and densities and predators. In each trial, the required number of prey and predators were placed in the arena and the prey consumption was recorded at the end of a 24-h period. The conspecific IGP system (experiment 2) In this experiment, R. filiformis and L. griseus were used in a density of 2 individuals against 10 individuals of A. bouvieri constituting two separate combinations of IGP system. Mosquito larvae and chironomid larvae were considered as shared prey separately. Thus, for each combination of IGP system, two different shared prey were used in two density levels—low (50 individuals) and high (200 individuals). Eighteen trials were considered for each prey type and densities against the two combinations of IGP system. The predator and prey individuals were placed in the arena and the shared prey and IG prey consumption were observed for each trial for a period of 24 h. The heterospecific IGP system with predator substitutability (experiment 3) In this experiment, one individual each of R. filiformis and L. griseus was used as IG predators and 10 individuals of A. bouvieri as IG prey. Using two density levels—low (50 individuals) and high (200 individuals) of shared prey (mosquito larvae and chironomid larvae), 18 trials were carried out for each shared prey type and density levels to observe the shared prey and IG prey consumptions for a period of 24 h. Thus, in IGP systems, the IG predator density was two individuals, while the IG prey density was 10 individuals at
Intraguild Predation in Heteroptera
(a)
377
(b)
(c)
Fig 1 Outline of IGP system showing the trophic relation between heteropteran predators and shared dipteran prey. In a, single predator system is portrayed; in b, IGP system with conspecific predators are shown, and in c, heterospecific predator combinations is shown. Arrows are directed towards the predator; figures inclusive of dashed arrow represent IGP system. Experiments were carried out using single predator and both predators (IGP system) against two prey types. Two species—Ranatra filiformis and Laccotrephes griseus were considered as IG predator separately; density of two individuals per replicate: two IGP system: conspecific and heterospecific IG predator individuals. Constant density of Anisops bouvieri; 10 individuals per replicate, in both single predator and IGP system. Two density levels, high (200) and low (50) in both single predator and IGP system using mosquito and chironomid larvae separately as shared prey.
each prey density in any trial. A predator individual was used in a single trial only. The trials were carried out at different time interval (interspersion) with random selection of insect individuals (randomization) such that trials represent true replicates (Hurlbert 1984). Assuming generalized linear model (GLM), the data on the prey consumption was subjected to a regression following binomial GLM using a logit link with predator species, prey species, and prey density as predictors for experiments 1 and 2 and prey density for experiment 3. In the binomial GLM, the response variable ‘prey consumption’ is assumed to follow binomial (n, p) distribution with n replicates for each combination of explanatory variables. The probability parameter p is here a linear combination of explanatory variables. A logit link was used and parameters were estimated through maximum likelihood using the software XLSTAT (Addinsoft 2010). A Chi square value was used to deduce the significance of the estimated parameters of the model that includes predator combination (pred-comb), prey type (prey-type), and prey density (prey-den). To highlight the differences in the prey consumption among the different combination of prey and predators, for each level
of density and prey identity, the data obtained on prey mortality was separately subjected to one-way ANOVA (Zar 1999) using predator combinations as explanatory factor. Similarly, the data on the shared prey and IG prey consumption in IGP system was subjected to one-way ANOVA, to compare the effects of predator substitutability, highlighting conspecific and heterospecific IG predator identity. Prey risk analysis The risk of the shared prey (mosquito larvae) to predation by the IG predator and IG prey in IGP system was assessed using the models of Crumrine & Crowley (2003) and Crumrine (2005). The observed shared prey mortality rate under the conditions of single predator (A. bouvieri, L. griseus, and R. filiformis as single predators) was used as input to construct the expected shared prey mortality rate under IGP system. The expected mortality rate was then compared with the observed mortality rate of shared prey under IGP system (with R. filiformis and L. griseus as IG predators and A. bouvieri as IG prey). In this model (Crumrine & Crowley
Brahma et al
378 Table 1
Summary of the experimental design adopted in the present study.
Parameters
Details
Remarks
Single predator experiment (experiment 1) Predators Rantra filiformis, Laccotrephes griseus, Anisops bouvieri Prey
4th instar of Culex quiquefasciatus and Chironomid midge
Replicates 18×2 prey densities per prey per predator IGP system experiment (experiment 2; conspecific IG predators) IG predator Ranatra filiformis, Laccotrephes griseus IG prey Anisops bouvieri Shared prey 4th instar larvae of Culex quiquefasciatus or Chironomid midge IG predator: IG prey ratio 2:10 Replicates 18×2 prey densities per shared prey type and IG predator species IGP system experiment (experiment 3; heterospecific IG predators) IG predator One each of Ranatra filiformis, Laccotrephes griseus Replicates for predator 18×2 prey densities per shared prey substitutability Prey consumption Observed for 24 h Analysis Factorial ANOVA on the predation data followed by post hoc multiple comparison tests (Crumrine & Crowley 2003, Crumrine 2005) Students’ t test for comment on predator substitutability and testing the predation risk analysis
2003, Crumrine 2005, 2010), the shared prey mortality rate (ki) for ith replicate of treatments over the period of trial (t= 24 h) could be presented as: . ki ¼ 1nð1−pi Þ t where, proportion of prey killed was assumed as pi and thus the proportion surviving is (1− pi). Thus, for A. bouvieri, R. filiformis, and L. griseus predation on mosquito and chironomid larvae, if kA is assumed as shared prey mortality in the presence of IG prey A. bouvieri alone and kTP is the shared prey mortality in the presence of R. filiformis or L. griseus alone, then the null hypothesis for predation when both are present is: kA þ kTP ¼ kAþTP This is based on the assumption of additive effects of two predators in multiple predator system. Based on this equation, if observed kA+TP expected kA+TP, then it indicates higher number of prey mortality, i.e., risk enhancement occurs. In this model, risk reduction and risk enhancement is derived from the null additive models for multiple predators (Soluk & Collins 1988, Sih et al 1998). The interactions between the IG predator (R. filiformis and L. griseus), IG prey (A. bouvieri), and the shared prey (mosquito larvae and
Density of 2, 2, and 10 conspecific individuals, respectively Density of 50 (low) and 200 (high) for each prey species, respectively 216 replicates Density 2 conspecific individuals Density 10 individuals Density of 50 (low) and 200 (high) for each prey species, respectively 144 replicates
Density 2 heterospecific individuals 72 replicates
For both IG prey and shared prey; Predation risk analysis for mosquito and chironomid prey.
chironomid larvae) is assumed as multiple predator system with possibilities of IGP system, and the model (Crumrine & Crowley 2003, Crumrine 2005, 2010) was applied to analyze the risk reduction and risk enhancement on the shared prey. Interpretation of the model can be extended to judge the effect of IGP on shared prey mortality. Thus risk reduction of shared prey would mean possibility of IG prey mortality and risk enhancement of shared prey would mean additive effect of both predators and absence of IGP. A two-tailed t test was applied to find out significant deviation of the ratio of observed kA+TP/ expected kA+TP from unity (Zar 1999). The effect of density of the shared prey and IG predator on the risk reduction and risk enhancement for the shared prey and the IG prey was assessed. In addition to conspecific top predators, the analysis included combinations with heterospecific top predators (one each of L. griseus and R. filiformis).
Results The vulnerability of mosquito and chironomid prey varied with the predator identity and composition. The prey mortality due to consumption by the IG predator R. filiformis and L. griseus and IG prey A. bouvieri when present as individual predator species varied with the relative densities of the prey available. Prey consumption as a function of the predator combinations, prey density, and the prey-type, following
Intraguild Predation in Heteroptera
379
consumption of mosquito prey at low (F 4, 89 = 53.49; p
