Response of the reduviid bug, Rhynocoris marginatus (Heteroptera: Reduviidae) to six different species of cotton pests

In Indian agro-ecosystems Rhynocoris marginatus (F.) is one of the most abundant predatory arthropods and feeds on a wide range of insect pests. We investigated the responses of R. marginatus to six species of cotton pests: Spodoptera litura (F.), Sylepta derogata (F.), Pericallia ricini (F.), Mylabris indica (Thunberg), Mylabris pustulata (Thunberg) and Dysdercus cingulatus (F.), in terms of its predatory behaviour (approach and handling times), weight gain, macromolecular profi le (content of carbohydrates, proteins, free aminoacids and lipids) and haemocytic profi le. We also determined the predator’s reliance on kairomones from different species of prey. Larvae of the species of Lepidoptera studied were approached and captured more quickly than adults of the two meloid coleo pteran and one heteropteran pest and were more benefi cial to the predator in terms of weight gain. Predators had a higher total protein content when reared on larvae of the three lepidopteran species, higher lipid content when reared on adults of the two meloid coleo pteran species and a higher carbohydrate content when reared on adults of one heteropteran species. The number of haemocytes was greater in predators reared on larvae of the Lepidoptera studied, followed by those reared on adults of the heteropteran and lowest in those reared on adults of the two meloid coleopterans. Response to kairomones was strongest for S. litura followed by S. derogata and M. pustulata. We conclude that the Lepidoptera studied tended to be, for this predator, superior prey, with S. litura being especially benefi cial and the prey species for which R. marginatus has the highest kairomonal preference. Moreover, we propose that R. marginatus may be useful as a biocontrol agent against lepidopteran cotton pests.

Rhynocoris marginatus (F.) (Heteroptera: Reduviidae), a polyphagous reduviid predator, is present in most Indian agro-ecosystems including groundnut, cotton, soybean (Sahayaraj, 1995(Sahayaraj, , 2000(Sahayaraj, , 2007(Sahayaraj, , 2014)), sugarcane (Easwaramoorthy et al., 1994) and pigeon pea (Ambrose & Claver, 2001).Based on laboratory and fi eld studies of pest suppression effi ciency, we know that R. marginatus consumes a broad range of prey (Sahayaraj, 2006) and that it specifi -predator adapts its rate of movement to that of the prey, closes in on the prey and captures and holds it tightly by its forelegs, (4) rostral probing -the predator probes various parts of the body of the prey with its rostrum, (5) paralysing the prey -the predator withdraws its rostrum; the prey becomes immobile indicating that paralysis has occurred, (6) feeding -the predator transports the prey to a secluded place and sucks out the body content by inserting its rostrum at one or more places and (7) post-predatory behaviour -the predator drops the carcass and cleans its rostrum, antennae and forelegs.The behavioural elements 1-2 imply that R. marginatus is able to detect and identify its prey from a distance as stated by Haridass et al. (1988) and Claver & Ambrose (2003).The time that the predators spend in carrying out each of the fi rst three events (1-3) was summed and used as a collective measure of the time spent locating and capturing the prey (termed approach time, AT).Similarly, the time that the predators spent in each of the subsequent three events (4-6) was summed and used as a collective measure of time spent handling and eating the prey (termed handling time, HT).These observations were recorded in the laboratory at 27-29°C, 60-80% RH.After 24 h, the weight of the prey was recorded and the difference in weight of the prey was used as a measure of the weight gained by the predator.Weight gains were corrected by predator weight loss due to desiccation as determined by weighing 6 predators kept individually in separate plastic containers without prey for 24 h.For each prey, the behaviour was recorded separately for ten male and female reduviids.

Prey extracts and predator choice
To test for possible kairomonal responses of R. marginatus to the different species of prey, compounds from each of them were extracted using hexane, following the method used by Singh & Paul (2002) and modifi ed by Sahayaraj et al. (2003) and Sahayaraj (2008).Two pieces (1 cm 2 ) of fi lter paper (Whatmann No. 1) was each impregnated with extracts from either M. pustulata or M. indica (treatment "TC-1").A third piece of fi lter paper was impregnated with hexane and served as a control.The three pieces of fi lter paper were arranged in a triangle in a Petri dish (9 cm in diameter).Subsequently one predator (starved for 24 h as described above) from the mixed-pest rearing was randomly selected and introduced into the centre of the Petri dish and the lid gently replaced.The piece of fi lter paper fi rst approached by the predator was recorded.Observations took place in the laboratory at 27-29°C, 60-80% RH.Ten predators were tested and the preferred prey extracts determined were those fi rst approached by the majority of the individuals tested.The preferred extracts were subsequently tested in another 10 replicates in a similar way with a piece of fi lter paper with extracts of D. cingulatus (treatment "TC-2") and the fi ve other prey species in "TC-3", "TC-4", "TC-5" and "TC-6" (refer to Table 3 for the prey extracts tested in each test).No predator was tested more than once in these experiments.From the results, the percentage of predators preferring the extracts of each pest species was calculated (Sahayaraj, 2008).

Sample preparation for macromolecule estimation
Ten predators from each of the single-pest rearings were randomly selected, pre-weighed separately, killed and dried in a hotair oven at 50°C for 6 h.The weight of the dried insects was recorded after which the insects were powdered using a mortar and pestle and the total content of carbohydrates (Sadasivam & Manickam, 1997), proteins (Lowry et al., 1951), lipids (Bragdon, 1951) and total free amino acids (Palanivelu, 2001) R. marginatus prefers particular species of prey, we also investigated whether it responds to the kairomones of the different species of prey and whether the response is similar for all these prey species.As a reference species for the studies on the responses of R. marginatus to the cotton pests mentioned above, we included the rice moth, Corcyra cephalonica Stainton (Lepidoptera: Pyralidae) in our studies.The suitability of this species as prey and for rearing R. marginatus is recorded (Sahayaraj, 2002;Sahayaraj & Sathiamoorthi, 2002).

Collection and maintenance of insects
A mixture of the different life stages of R. marginatus was collected from cotton fi elds in the Tirunelveli district in the state of Tamil Nadu and maintained individually in plastic vials (5.5 by 4.4 cm with perforated lids) under laboratory conditions at 27-29°C, 60-80% RH and 11L : 13D h on a mixture of larvae of S. litura, S. derogata, P. ricini and adults of M. indica and M. pustulata, and fi fth stadium nymph of D. cingulatus and larvae of C. cephalonica in equal amounts.The prey was replenished every second day.After emergence, 30 adult predators (139 ± 2 mg and 181 ± 3 mg for male and female, respectively) were maintained individually in similar plastic vials as above for one month to standardize their physiological state.The adults were kept in the same conditions as above on either just one of the mentioned species of pests ("single-pest rearing") or on a mixture of the six pest species in equal amounts ("mixed-pest rearing") with the prey being replenished every second day.After one month, the predators from the single-pest rearing were killed and used to estimate their content of free amino acids and total protein, carbohydrate and lipid Except for C. cephalonica all pest species were collected from the same cotton fi eld from which the reduviids were collected.Larvae of the three noctuid pests S. litura, S. derogata and P. ricini were maintained in groups in plastic troughs (300 ml) on young cotton leaves, whereas the adults of two meloid coleopterans, M. indica and M. pustulata, and the adults of the heteropteran, D. cingulatus, were reared on fl owers and cotton bolls in plastic vials (500 ml).C. cephalonica eggs were purchased from the Agriculture Offi ce, Palayamkottai, and maintained in plastic troughs using crushed wheat and groundnuts as food for the emerging larvae.All species were kept at the same conditions as above.Laboratory reared fi fth instar larvae of S. litura (374 ± 10 mg), S. derogata (393 ± 2 mg), P. ricini (415 ± 14 mg) and C. cephalonica (42 ± 1 mg) and adults of D. cingulatus (106 ± 5 mg), M. indica (323 ± 2 mg) and M. pustulata (344 ± 3 mg) were used in the experiments.

Predatory behaviour
Three pre-weighed individuals of the same species were introduced into a plastic container (300 ml) and allowed to move undisturbed for fi ve minutes.Predators were randomly selected from the mixed-pest rearing and starved for 24 h by confi ning them individually in an empty Petri dish (9-cm in diameter) at 27-29°C, 60-80% RH and 11L : 13D.Predators were subsequently introduced singly into the Petri dishes, and the sequence of feeding events were recorded and timed visually until feeding ceased.The sequential pattern in predatory behaviour is (see Sahayaraj, 2007, for details) : (1) arousal -the predator closely watches the movement of the prey, is restless and has straightened legs, (2) approach -the predator moves towards the prey, pointing its rostrum and antennae forward, (3) capture -the

Haemocytes
Ten predators from each of the single-pest rearings were randomly selected and used to determine the total number of haemocytes per mm 3 (total haemocyte count, THC) and the percentages of the different types of haemocytes (differential haemocyte count, DHC).The foreleg of a predator was amputated with a pair of fi ne sterilized scissors and up to 0.5 μl of the haemolymph was directly drawn in to a WBC pipette (Naubauer's haemocytometer).The haemolymph was then diluted with an acidifi ed physiological saline (NaCl -4.65 g, KCl -0.5 g, CaCl 2 -0.11 g, Gentian violet -0.005 g, acetic acid -1.25 ml diluted in 100 ml distilled water).The diluting fl uid was drawn into the pipette up to mark 1.1 giving 20 times dilution.The inner surface of the WBC pipette was rinsed several times with diluting fl uid before the haemolymph was drawn.After at least one gentle stirring to ensure complete dispersion of the cells and prevent agglutination and plasma clot formation, the haemocyte population (cells/mm³) (Jones, 1962) was counted under microscope.For differential haemocyte counts, the haemolymph that oozed-out of the foreleg was spread out on a clean slide and a neat thin smear was prepared with a glass rod or a cover slip.It was then fi xed with 1 ml of 100% methanol or acetic acid, stained with 4% Giemsa stain and allowed to dry out in the air.The preparations were observed under a microscope.The haemocytes were categorised based on the identifi cation keys of Gupta (1985) and Ambrose (1999) as prohaemocytes, plasmatocytes, granulocytes, cystocytes and oencytoids.

Statistical analysis
Individual data on approach time, handling time, weight gain, total carbohydrate, protein and lipid level and various haemocyte levels in male and female predators fed on different species of prey were subjected to two-way ANOVA and Tukey's Test with signifi cance recorded at the 5% level.Various relationships between the parameters measured were analyzed using linear regression.The percentage preference of the predator for extracts of the different species of prey was compared among treatments using chi-square tests of independence.All analyses were performed using SPSS Version 13.0 (SPSS, 2004).

Approach time (AT) and handling time (HT)
Generally, R. marginatus became aroused, approached and captured its prey quickly with AT values varying between 2.5 and 6.7 min (Table 1).The shortest ATs were recorded for both male and female reduviids attacking S. litura and the longest for females attacking the re ference prey C. cephalonica and for males attacking D. cingulatus and M. pustulata (Table 1).Generally both males and females had shorter ATs for larvae of the three le pi do pteran cotton pests, with shorter times recorded for larvae of the noctuid S. litura than for the adults of one heteropteran and two coleopteran pests (Table 1).There was not a signifi cant relationship between AT and the weight of the prey species (df 6, 68; F = 4.15; P > 0.05; r 2 = -0.544).
The time (HT) required for the reduviids to probe, paralyse and eat the different species of prey varied greatly from about half an hour for females attacking larvae of the noctuid S. derogata to more than 200 min for those attacking C. cephalonica and M. indica (Table 1).There were no consistent differences in HTs of females and males, with the latter having the shortest HT when attacking P. ricini and the longest when attacking C. cephalonica (Table 1).There were no signifi cant relationships between HT and AT for either male or females reduviids (males: df 6, 62; F = 4.00; P > 0.05; r 2 = -0.070;females: df 6, 60; F = 4.01; P > 0.05; r 2 = 0.466).Similarly, there were no signifi cant relationships between HT and the weight of the prey (df 6, 65; F = 4.08; P > 0.05; r 2 = -0.401).

Weight gain
The weight gained by R. marginatus after feeding on the different species of prey differed greatly with values ranging between 10.6 and 90.5 mg, with the highest and lowest weight gain recorded when the reduviids (both males Different lower case (for male) and upper case (for female) letters within a column indicate signifi cant differences at 5% level.Signifi cant differences between male and female predators feeding on the same species are indicated by *.
and females) ate larvae of the noctuid S. derogata and adults of the meloid coleopteran, M. pustulata, respectively (Table 1).Although the weight gain of males and females that ate the same species of prey often differed signifi cantly, there was no consistent difference (Table 1).
The reduviid effi ciency in obtaining food from prey in relation to the time invested in capturing and feeding varied between prey species from 0.07 to 2.4 mg per minute (Table 1).The highest effi ciencies were recorded for both males and females attacking larvae of S. derogata followed by P. ricini and S. litura, and the lowest when attacking M. pustulata (Table 1).Except for M. indica and D. cingulatus the order in the effi ciencies was the same for both males and females (Table 1).

Prey extracts and predator choice
When the reduviids were exposed to prey extracts, they showed the same initial behaviour as recorded in the sequence of behaviour normally leading to capture of prey (Sahayaraj, 2008), i.e. they moved towards the paper with their rostrum and antennae pointing forward, protruded the rostrum and tried to obtain food from the fi lter paper.In the fi rst three tests ("TC-1" to "TC-3"), R. marginatus invariably chose extracts of M. pustulata over those of M. indica, D. cingulatus and P. ricini (Table 2).In contrast, in the fourth test, R. marginatus chose extracts of larvae of the lepidopteran, S. derogata, over those of the meloid coleopteran, M. pustulata, and in the remaining two choices the reduviid chose extracts of S. litura over those of S. derogata and M. pustulata.

Macromolecular profi le
The prey species used as food in the single-species rearing of R. marginatus had a signifi cant effect on their macromolecular composition (Table 3).Compared with predators reared on the reference prey, predators reared on larvae of either of the two lepidopteran species, S. litura and S. derogata, had a signifi cantly higher content of free amino acids followed by those reared on P. ricini or M. pustulata.In contrast the amino acid content of predators reared on M. indica or D. cingulatus was not different from that of predators reared on the reference prey.These differences were partly refl ected in the protein content, which was signifi cantly higher for predators reared on larvae of either of the three lepidopteran species, S. litura, S. derogata and P. ricini, compared with predators in the reference group.However, the protein content of predators reared on M. pustulata, M. indica or D. cingulatus was signifi cantly lower than that of the predators in the reference group.Regarding the carbohydrate content, this was signifi cantly higher for predators reared on D. cingulatus compared with the predators in the reference group, whereas it was signifi cantly reduced compared to the reference group for predators reared on the remaining prey species (Table 3).Compared with predators reared on the reference prey, total lipid content was higher for predators reared on the two meloid coleopterans (M.pustulata, M. indica) or adults of the heteropteran, D. cingulatus, being highest for coleopteran-reared R. marginatus and lowest for those reared on larvae of either of the two lepidopteran species, S. derogata and S. litura (Table 3).

THC and DHC of R. marginatus
Total haemocyte count for predators from the singlespecies rearing is shown in Table 4. THC of the predators reared on S. litura, S. derogata, P. ricini or D. cingulatus was signifi cantly higher than that of the predators in the reference group.In contrast, THC of predators reared on M. indica was not signifi cantly different from that recorded for the reference group, and predators reared on M. pustulata had a signifi cantly lower THC than those reared on the reference prey group.Generally, the THC level was higher in predators reared on larvae of the three lepidopteran species and adults of the heteropteran pest compared with those reared on the two meloid Coleoptera.Regression analyses between total protein content and THC (r 2 = 0.621; df 1, 5; t = 2.230; P < 0.05) and also between total lipid content and THC (r 2 = 0.597; df 1, 5; t = 2.230; P < 0.05) were signifi cant.
Results of the differential haemocyte count (DHC) (Table 4) revealed that prohaemocyte and plasmatocyte levels were higher in predators reared on S. litura or S. derogata than for those reared on the reference prey.For predators reared on M. pustulata, M. indica or D. cingulatus only the prohaemocyte level was higher in comparison with predators reared on the reference prey, whereas plasmatocyte levels were the same or lower.Compared with predators reared on the reference prey, the level of cystocytes was higher for predators reared on all pest species except S. litura, whereas the oenocytoid level was higher compared to the reference prey group for predators reared on P. ricini, M. pustulata, M. indica or D. cingulatus but lower for predators reared on S. derogata or S. litura.

DISCUSSION
There were signifi cant diff erences in how fast R. marginatus became aroused, approached and captured (approach time, AT) the different species of prey studied, which presumably refl ect combinations of differences in prey size, mobility and speed of movement, prey defence and/or prey chemical cues (Ambrose, 1999;Sahayaraj, 1991Sahayaraj, , 2006)).
Arousal, approach and capture were generally faster for larvae of the three species of lepidopteran pests (except the reference prey C. cephalonica) than the small heteropteran (D. cingulatus) and the two somewhat smaller meloid coleopterans (M.indica, M. pustulata).This is in accordance with previous fi ndings for other reduviids presented with the same species of prey (Sahayaraj, 1991).The longer arousal, approach and capture time for the coleopteran and heteropteran prey is most likely due to the fact that they were offered as adults, whereas the lepidopterans were offered as fi fth instar larvae, which are presumably less mobile and thus easier to capture.However, it is possible that R. marginatus was more motivated and/ or more able to capture prey that provide high nutritional rewards more quickly, based on the negative relationship between weight gain and AT.An additional explanation for the long arousal, approach and capture time for the two meloid coleopterans is evidently their release of a defensive viscous fl uid, which could delay reduviid approach and/or impede their attack (Sahayaraj, 1991).Cantharidin, a common defensive terpenoid present in the defensive viscous fl uid of Meloidae, is toxic to the enemies of these beetles (Hashimoto & Hayashi, 2014).Although, S. litura is known to produce a defensive fl uid (Kumaraswami, 1991), this was not observed in the present study.
Compared to the three Lepidoptera studied the AT recorded for the much smaller lepidopteran reference prey was relatively longer, although it might be assumed that it was more easily captured.It may be speculated that this refl ects a lower interest of the reduviid in this species of prey, perhaps due to a perception of a smaller reward or of a longer handling time, which was the longest recorded, longer even than that recorded for the meloid coleopteran, M. indica.Apart from this possible link between AT and HT, no overall correlation was found between AT and HT, which indicates that quickly approached and captured species of prey were not also probed, paralyzed and eaten more quickly than prey approached and captured more slowly.
There were no consistent differences between male and female reduviids in the approach and capture time of the different species of prey, although females in three cases approached their prey faster than males.Similarly, although the time spent probing, paralyzing and eating differed signifi cantly between the sexes for all the different species of prey, there was no consistency in one sex being faster in handling prey than the other.This is in contrast to the results of other studies on reduviids in which males both approach and handle prey faster than females (Sahayaraj & Ambrose, 1992, 1993, 1994;Rocha & Redaelli, 2004).Although other authors have shown that female reduviids consume more food than males (Ambrose, 1980;Kumaraswami, 1991;Sahayaraj, 1991Sahayaraj, , 2008)), presumably to meet their reproductive needs, as demonstrated for Pentatomids, in which fecundity increases with body mass (Evans, 1962;Phoofolo et al., 1995;Grundy & Maelzer, 2000), we did not fi nd similar differences in the weight gains of males and females; only when they fed on two of the species of prey (P.ricini and S. litura) did females gain more weight than males.The reason for this inconsistency is unknown.
There was no consistent pattern in the HTs associated with the different species of prey; relatively long HTs were recorded for both meloid coleopterans with hard cuticles (Sahayaraj, 2008) and the soft-bodied larvae of the lepidopteran reference prey.The reason for these differences in HT for the different prey species is not known but may be due to a longer time needed for paralyzing or liquefying the body contents, the viscosity of the liquefi ed prey or the amount ingested (Cohen, 1990), although the latter is not supported by the present result of no correlation between reduviid weight gain and HT.
Compared with the weight gained when they fed on the two adult meloid Coleoptera and the heteropteran, R. marginatus gained more weight when reared on the larvae of three Lepidoptera, P. ricini, S. derogata and S. litura.This most likely refl ects a nutritionally superiority of this type of prey, resulting in a higher content of amino acids and proteins in the bodies of predators reared solely on either of these three species, and that larval prey are more easily digested than adult prey.However, the weight gained from feeding on the larvae of the lepidopteran reference prey, C. cephalonica, was, despite the longer handling time, considerably lower than that gained from the other Lepidoptera.This indicates that C. cephalonica is an inferior prey for R. marginatus, presumably due to its nutritional quality, since predators reared solely on this prey had a lower content of amino acids and proteins compared with those that fed on the other Lepidoptera.The differences in weights gained illustrate that long handling times do not necessarily result in high amounts of food being ingested but also point to the fact that considerable weight gains can be achieved from rather short handling times as seen when the reduviids were reared on S. derogata.
Our study has demonstrated that compounds or mixtures of compounds extracted from the different cotton pests examined here function as kairomonal stimulants in eliciting probing behaviour in R. marginatus.This supports previous fi ndings for the same and other reduviid species (Cohen, 1990;Sahayaraj & Paulraj, 2001;Tebayashi et al., 2003;Sahayaraj & Delma, 2004;Sahayaraj, 2008;Sujatha et al., 2012).R. marginatus responded differently to the extracts from the different species of prey, which presumably refl ect differences in volatile components and composition of the extracts (Ananthakrishnan, 1996), but showed a clear preference for S. litura.The seemingly general notion that reduviids prefer lepidopteran larvae to other types of prey (Ambrose, 1980;Kumaraswami, 1991;Sahayaraj, 1991Sahayaraj, , 2008) ) was, however, not fully supported by the present study in which the predator preferred the extract from the meloid coleopteran pest, M. pustulata, over that from the lepidopteran P. ricini.This may be due to the hairy appearance of P. ricini larvae, a noxious smell or unpleasant taste (McMahan, 1983), or to the presence of toxic allelochemicals as reported by Lamdin et al. (2000) for Lagoacris pata Packard (Lepidoptera: Megalopygidae).
The prey species had a signifi cant infl uence on the macromolecular profi le of R. marginatus, which had higher total protein content when reared on the three lepidopteran larvae, higher lipid content when reared on two adult meloid coleopterans and higher carbohydrate content when reared on the adult heteropteran prey, relative to that recorded when reared on the reference prey.This suggests that different metabolic pathways are activated by ingestion of these prey, with the increased synthesis of amino acids and proteins being of particular interest, as the biological control potential of reduviids increases logarithmically in response to the protein content of their prey (O'Neill et al., 2008).Also the number of haemocytes, taken as an index of the physiological state of the predator (Ambrose & George, 1996), was higher in predators reared on the lepidopteran larvae than on the reference prey, also a lepidopteran, followed by predators reared on one species of heteropteran prey, with a positive correlation between the number of haemocytes and content of amino acids and total lipids.The differences recorded here in the number of haemocytes in predators reared on different species of prey support the notion that total haemocytes counts, among other things, are infl uenced by nutrition (Jones, 1962;Wheeler, 1963).The regression analysis between total protein and lipids with THC was highly signifi cant supporting the views of Wheeler (1963) and Jones (1962).
Haemocytes are involved in cellular immunity (Lavine & Strand, 2002) and protein storage (Geiger et al., 1977), transport of nutrients and hormones (Patton, 1983) and histo-morphological changes during development (Hazarika & Gupta, 1987).The especially high content of amino acids and protein in predators reared on larvae of the noctuid S. litura together with the high number of haemocytes (with elevated proportions of prohaemocytes and plasmatocytes) indicate that this species of prey is important for R. marginatus not only in terms of direct nutrition but also in terms of increased immunity.This presumed essential character of S. litura was refl ected in the higher kairomone-based preference for this pest over the other three lepidopteran species.
The present study has demonstrated signifi cant differences in the response of R. marginatus to several species of pests.Among the six species of prey offered, the larvae of lepidopteran cotton pests were generally of superior quality in terms of arousing this predator's interest (approach time), weight gain, content of amino acids and protein and increasing haemocyte numbers.This fi nding indicates that R. marginatus might be used as a good biocontrol agent against lepidopteran cotton pests, especially the larvae of the noctuid S. litura.This study also indicates that the body extracts of the prey species tested contain kairomones, which elicit probing behaviour in R. marginatus.Further studies should be undertaken to investigate the prey preferences of R. marginatus in plants infested with one or several herbivores (Moayeri et al., 2007).Further studies will also be needed to investigate suitable strategies for the use of R. marginatus for controlling insect pests of cotton.

TABLE 1 .
Mean (± SE) approach time (AT), handling time (HT), weight gain (WG) and weight gain effi ciency (WGE) of Rhynocoris marginatus males and females when fed different cotton pests.

TABLE 2 .
Percentages of Rhynocoris marginatus that approached extracts of the six cotton pests used in this study.

TABLE 3 .
Mean (± SE) macromolecular content of Rhynocoris marginatus fed on different pest species.

Table 4 .
Differential (DHC in %) and total haemocyte (THC) counts of Rhynocoris marginatus fed on different species of cotton pests.
gDifferent letters within a column indicate signifi cant differences at 5% level.