Aphis gossypii ( Hemiptera : Aphididae ) as a factor inhibiting the survival and population increase of the predator Macrolophuspygmaeus ( Hemiptera : Miridae ) on cucumber

The influence of cucumber offered as a host plant either alone or with Aphis gossypii Glover (Hemiptera: Aphididae) was studied on the various life table and biological characteristics of the predatory bug Macrolophus pygmaeus Rambur (Hemiptera: Miridae). The nymphal development was studied at 15, 20, 25, 30 and 35°C while adult performance was assessed at 15, 20, 25 and 30°C, using a 16L : 8D photoperiod and 65 ± 5% r.h. Nymphs completed their development at all temperatures except at 35°C. Nymphal development took significantly longer time in the absence than in the presence of prey at 20 and 25°C, but the reverse was true at 15°C. Nymphal mortality was highest at 15°C in the presence of prey and it was mainly recorded at the first and second stages. Females oviposited a small number of eggs at all temperatures but not at 30°C in the absence of prey. The average number of eggs per female was almost similar with or without prey, being highest at 20°C, and adult longevity was highest at 15°C. The results concerning population parameters clearly showed that cucumber with or without prey can not support a population increase of M. pygmaeus. However, it seems that A. gossypii on cucumber inhibits development of M. pygmaeus more than when this aphid species is not present. This adverse effect on this host plant-prey system possibly results from the particular aphid genotype on cucumber, leading to high nymphal mortality, reduced fecundity and short adult life-span of M.pygmaeus.

Aphis gossypii Glover (Hemiptera: Aphididae) is one of the most important insect pests of vegetables and particu larly of cucumber.Considering both its resistance to insecticides (Wang et al., 2002), and the lack of effective natural enemies (Alvarado et al., 1997), the evaluation of the potential of the polyphagous predator Macrolophus pygmaeus Rambur as an effective biological control agent of this pest becomes worthwhile.
In an effort to obtain a more complete knowledge of the factors involved in the predator's behaviour and effec tiveness as a natural enemy of A. gossypii on cucumber, its ability to utilize both prey and plant material indicates that its performance should be studied not only with but also without the presence of prey.Moreover, the thorough investigation of this host plant -prey -predator system could highlight a strategy for a more appropriate use of this predator in future biological control programmes.
Under the above considerations, the objective of the current work was to investigate M. pygmaeus perform ance on cucumber as affected by A. gossypii presence.This was conducted by the study on (1) the biological characteristics and (2) the population parameters of the predator M. pygmaeus in the presence and in the absence of A. gossypii on cucumber, at an ecologically relevant range of temperatures.

MATERIALS AND METHODS
A colony of the predator M. pygmaeus was initiated from insects collected from a tomato field in the area of Boeotia, cen tral Greece.The cultures were maintained on potted eggplants (cv.Bonica) infested with Myzus persicae (Sulzer) (Hemiptera: Aphididae) in woodframed (80 x 80 x 70 cm) cages.Cultures were kept in a glasshouse at 22.5 ± 2.5°C (mean ± S.D.) under natural day light.

Egg incubation period and hatching percentage
Due to the very low fecundity of M. pygmaeus on cucumber, it was not possible to estimate the egg incubation period and hatching percentage.For the calculation of the population parameters of M. pygmaeus, we used the values for egg incuba tion period and hatching percentage recorded on eggplant with the prey M. persicae, at the respective temperatures.In previous studies (Perdikis, 2000;Perdikis & Lykouressis, 2002), it was Bars at each temperature followed by different letters differ significantly (P < 0.05).
shown that the egg incubation period and hatching percentage were not significantly affected by either the presence/absence of prey or plant type (eggplant or tomato).

Nymphal development
The development and mortality rate of the nymphs were studied by placing newly emerged nymphs (less than 24 h old) from the stock cultures individually into Petri dishes.Each Petri dish (9 cm in diameter and 1cm deep), had a 3cm diameter hole in its top, covered with fine muslin, to allow adequate ventila tion.A water moistened layer of cotton was placed on the bottom of each dish, with a leaf of cucumber (cv.Brunex) placed upside down.The length and width of the leaf was about 2 and 1cm, respectively, smaller than the diameter of the dish.The leaves came from young plants with 8-12 leaves that were grown free of chemicals.The leaf contained A. gossypii or not, depending on the experiment.About 100 to 150 individuals including all stages and adults were present on a single leaf.They were coming from a population which was at the phase of increase.
The experiments were conducted in growth cabinets at 15, 20, 25, 30 and 35°C with 65 ± 5% r.h. and a 16L: 8D photoperiod.In each experiment 20 nymphs (replicates) were used.Develop ment and mortality of the nymphs were recorded at 24-h inter vals, but at 35°C at intervals of 12 h.
The sex ratio [2 / (2 + k)] of the emerged adults was recorded in all treatments.

Preoviposition period, longevity and fecundity
Preoviposition period, longevity and fecundity of M. pyg maeus, were studied in the presence and in the absence of A. gossypii.For that purpose, newly emerged adults, less than 1-day old were used.One pair of adults was released into a cyl inder (6.5cm in diameter and 30cm high), made from a trans parent 0.4mm thick PVC sheet.Each cylinder had two rectangular (1 3 x 5 cm) sides and one top opening covered with muslin.
In each cylinder two cucumber leaves bearing prey or not, depending on the treatment, were placed by inserting the base of their petioles into a layer of foam material firmly attached to the cylinder base.Also, a piece of cucumber stem, 3-5 mm in diameter and 6-7 cm long, was placed into each cylinder as an ovipositional substrate.This method is described in detail by Perdikis (2002).
Stem and leaves were replenished daily after having been examined under a stereomicroscope for oviposited eggs.Simul taneously, mortality of the adults was recorded and each dead male was replaced by a newly-emerged one.In each treatment about 20 replicates were used.The experiments were conducted in growth cabinets at 15, 20, 25 and 30°C, 65 ± 5% r.h. and a 16L:8D photoperiod.

Data analysis
Nymphal development, fecundity and longevity were ana lyzed using analysis of variance (ANOVA) after being logtransformed, and the means were compared using the Tukey-Kramer HSD test (p = 0.05).Survival curves of females were compared using the Wilcoxon test (p = 0.05).The differences in mortality rates and sex ratios were compared using a x^-test after data had been arcsine transformed.The analyses were con ducted using the statistical package JMP (version 4.0.2,SAS Institute 2001).

Life tables
The data obtained were used to construct life tables for M. pygmaeus.The age-specific survival rate (lx) and age-specific fecundity (mx) were calculated per day.The net reproductive rate (Ro), mean generation time (T) and intrinsic rate of natural increase (rm) were estimated (Birch, 1948;Andrewartha & Birch, 1954;Southwood, 1978).

Nymphal development
The nymphs of M. pygmaeus completed their develop ment at all temperatures tested with and without prey, except at 35°C where no nymphs completed development (Fig. 1a).
The presence/absence of prey significantly affected the period of development (f i ,86 = 7.89,p < 0.007) and a sig nificantly shorter duration of development was recorded in the presence, than in the absence of prey at 20 and 25°C.Interestingly, a significant interaction between tem perature and presence/absence of prey was recorded (^3,86 = 15.65,p< 0.001).This was ^mainly attributable to the significantly longer period that nymphs needed to com plete development at 15°C, when prey was present than when it was absent (Fig. 1a).The period of development was also significantly affected by temperature (F 3 86 = 36.1,p< 0.001), and was shortest at 25°C in the presence of prey, whereas in prey absence it was shortest at 30°C.
Nymphal mortality was always higher in the presence than in the absence of prey, although significant differ ences were not found.At 15°C mortality was higher with prey available than with prey absent (71 and 55%, respec tively).However, at 30°C the reverse was true (36 and 67%, in the presence and absence of prey, respectively) (Fig. 1b).Nevertheless, the effect of temperature was sig nificant (x2= 8.55, df = 3,p < 0.05 andx2 = 20.52,df= 3, The period that each nymph survive was expressed as a percentage of the mean total developmental period.Based on these data, nymphal survival rate curves were con structed at each temperature, either with or without prey (Fig. 2).This figure shows that when prey was available, nymphal mortality at 15°C mainly occurred at young stages, whereas in the absence of prey mortality was mainly recorded at older nymphal stages.At 20 and 25°C mortality occurred in a more or less similar way either with or without prey.Contrary, at 30°C mortality was similar between presence and absence of prey at younger stages, although at the older stages it highly increased when prey was absent.
It was observed that some of the nymphs found dead, when they developed with A. gossypii, had their tarsi cov ered with a sticky material.Similarly, during develop mental period, nymphs with exuviae not entirely shed or with third and fourth antennal segments black in colour, were sporadically recorded.

Sex ratio
Temperature had no significant effect on the sex ratio of M. pygmaeus in the presence (x2 = 1.67, df = 3, p > 0.5) or in the absence of prey (x2 = 1.74, df = 3 ,p > 0.5) (Table 1).However, the availability of prey led to a sig-  nificant increase in sex ratio at 25°C (x2 -6.78, df = l ,p < O.Ol).Most females were produced at 25 and 20°C, whereas the lowest number of females was recorded at 15 and 30°C, in the absence of prey.

Preoviposition period
The recorded percentages of ovipositing females are shown in Table 1.Females of M. pygmaeus oviposited at all treatments, except at 30°C in the absence of prey.The percentage of females that oviposited was significantly affected by temperature, when prey was present (x2 = 17.27, df = 3,p < 0.01), but not when prey was not avail able (x2 = 0.83, df = 2 ,p > 0.50).It is worth mentioning that the percentage of females that oviposited at 15°C was significantly lower in the presence than in the absence of prey (Table 1).
The length of the preoviposition period was signifi cantly affected by presence of prey and temperature, how ever, no significant interaction between the two factors was recorded (Tiy6 = 206.26,p< 0.01; F256 = 43.83,p< 0.01 and F256 = 2.42,p > 0.1, respectively).At each tem perature, a significantly shorter period was required by females in the presence than in the absence of prey to start ovipositing (Table 1).The shortest preoviposition period was recorded at 25°C (averaged 10.3 days and 15.2 days) and the longest at 15°C (averaged 14.5 days and 19.7 days) in presence and absence of prey, respec tively.

Fecundity
A significant effect of temperature was found on fecun dity (F2,56 = 8.72,p < 0.01).However, the availability of prey did not significantly affect the fecundity ( F ^^ 2.09, p > 0.15), whereas the interaction between the two factors was insignificant (F2,56 = 2.43, p > 0.1).Females were most fecund at 20°C (5.6 and 6.5 eggs/female in the pres ence and in the absence of prey, respectively) (Table 2).In addition, age-specific fecundity rates of M. pygmaeus were relatively higher at 20°C than at the other tempera tures (Fig. 3).

Female longevity
Longevity of females was significantly affected by presence/absence of prey (FU50 = 9.07,p < 0.01) and tem perature (F3150 -59.97,p < 0.01), whereas the interaction of those factors was not significant (F3 150 = 1.59, p > 0.19).Longevity was longest at 15°C (25.9 and 34.1 days) and shortest at 30°C (11.6 and 11.2 days), in the presence and absence of prey, respectively (Table 2).At 15 and 20°C, female longevity was longer in the absence than in the presence of prey; whereas at 25 and 30°C female longevity was not affected by presence/absence.
The age-specific survival rate of M. pygmaeus females, both in the presence and absence of prey, is shown in Figure 4. Generally, survival rate of females remained high for a relatively long period and then declined sharply (Fig. 4).Both in the presence and absence of prey, female survival rate significantly differed between 20-25 and 25-30°C (x2 = 4.78, df = l ,p < 0.03; x2 = 6.53, df = l ,p < 0.01 in the presence andx2 = 14.02, df = l ,p < 0.0l; % 2 = 18.32, df = 1, p < 0.01, in the absence of prey, respec tively).However, survival rate was not significantly dif ferent between 15 and 20°C in the presence as well as in the absence of prey (x2 = 1.47, df = 1,p > 0.23 and x2 = 0.97, df = 1, p > 0.32, respectively).Dead females bearing a quantity of sticky material on their tarsi, were observed in some cases.

Male longevity
Longevity of males was significantly affected by pres ence of prey (FU50 = 9.64,p < 0.01) as well as by tem perature (^3,150 = 92.12,p < 0.01).A significant interaction was recorded between those factors (F3450 = 3.7,p < 0.02), mainly due to the significantly higher lon gevity in the absence than in the presence of prey at 20°C.Longevity was highest at 15°C (31.1 and 36.7 days) and shortest at 30°C (10.8 and 10.60 days), with the presence and absence of prey, respectively (Table 2).
Generally, although males lived longer than females, statistical analyses with factors "sex", "temperature" and "presence/absence of prey" showed no significant differ ence between female and male longevity (F3300 = 1.19,p > 0.27).

Population parameters
The value of net reproductive rate (Ro) was highest at 20°C (0.653 and 0.735 females/generation, in the pres ence and absence ofprey, respectively) (Table 3).
The mean generation time (T) was much shorter at 25°C (40.573 and 48.767 days in the presence and in the absence of prey, respectively) than at the other tempera tures (Table 3).Generally the intrinsic rate of natural increase (rm) -values were found to be negative in all cases, with higher values at 20°C (Table 3).

DISCUSSION
Biological and life table parameters of M.pygmaeus fed on cucumber with A. gossypii, in this study, were inferior to those recorded when the predator was fed with M. persicae on eggplant or T. vaporariorum on tomato (Perdikis & Lykouressis, 2002).This adverse effect was reflected by a high percentage of nymphal mortality, highly reduced fecundity and much shorter adult longevity.
A very severe nymphal mortality occurred at 15, 20 and 25°C compared with that recorded in previous studies on tomato or two varieties of eggplant, ranging from 5-25%, in presence of prey (Perdikis & Lykouressis, 2000;Lyk ouressis et al., 2001).Younger nymphs suffered higher mortality than the older ones, which could be attributed to the accumulation of honeydew on their tarsi.Honeydew inhibits either walking and/or feeding of the predator, leading to high nymphal mortality.As the younger nymphs have shorter legs, the possibility to be trapped could be greater and therefore higher mortality would appear.This finding is in accordance with that from a pre vious study (Perdikis & Lykouressis, 1997), when a reduced survival of young nymphs was observed on melon with A. gossypii at 25°C.Moreover, Alvarado et al. (1997), working with Macrolophus caliginosus Wagner on cucumber with A. gossypii, recorded also high mortality of younger nymphs at 25°C, although the number of prey used was considerably smaller than that in our experiments.According to Tedeschi et al. (1999) providing early nymphs of A. gossypii did not permit M. caliginosus to complete development.Adverse effect of honeydew has been also observed on movement of the larvae of Propylea quatuordecimpunctata (L.) (Coleoptera: Coccinellidae) (Banks, 1957) whereas entrapment in whitefly honeydew, reduced the effectiveness of Encarsia formosa Gahan (Hymenoptera: Aphelinidae) (van Lenteren, 1990).In addition, Bush et al. (1993) reported a high nymphal mortality rate of Orius insidiosus (Say) (Hemiptera: Anthocoridae) due to the entrapment in a "sticky substance" produced by A. gossypii on beans and which excreted from the cornicles.
The presence of A. gossypii caused highly adverse effects on the performance of M. pygmaeus females, permitting oviposition of only half of them (Table 1), whereas when M. persicae or T. vaporariorum had been used as prey on eggplant and tomato, respectively, all the tested females oviposited (Perdikis & Lykouressis, 2002).In addition, females were much less fecund and showed a considerably reduced lifespan (Table 2) than those feeding on M. persicae and T. vaporariorum (213.90 and 228.25 eggs/female at 20°C, and female longevity averaged 122.40 and 129.35 days at 15°C, on each host plant respectively) (Perdikis & Lykouressis, 2002).Similar results have been found by Fauvel et al. (1987), where fecundity of M. caliginosus was lower when A. gossypii was offered as prey, compared to that observed in the presence of T. vaporariorum or M. persicae.Moreover, the intrinsic rate of natural increase of M. pygmaeus showed negative values as compared to those estimated on eggplant with M. persicae or tomato with T. vaporariorum (0.0975 and 0.0969 d a y 1 at 25°C, on each host plant, respectively) (Perdikis & Lykouressis, 2002).In general, the sticky material found to cover the tarsi of dead M. pygmaeus females, indicates that honeydew could be involved in both short adult longevity and highly reduced fecundity.
In tri-trophic systems adverse effects on predator per formance could be due to factors related with prey and/or host plant.As far as the role of prey is concerned, A. gossypii appeared in the present work to be an acceptable prey for M. pygmaeus, because the predator can complete its development feeding on it, at a significantly higher rate than in the absence of prey, in most cases.The acceptability of A. gossypii by M. pygmaeus is further supported by the work of Perdikis & Lykouressis (2000) in which the length of developmental period of this predator on eggplant with A. gossypii, was similar to that when M. persicae or M. euphorbiae were used as prey, under identical experimental conditions.The much better performance of M. pygmaeus nymphs when fed with A. gossypii, on eggplant than on cucumber could be, at least partially, explained taking into consideration that the hon eydew production rate differs among host-plant specific genotypes of A. gossypii as was found by Guldemond et al. (1995).It should be mentioned here that in spite of repeated efforts A. gossypii colonizing cucumber could not develop on eggplant and vice versa.Honeydew pro duction rate in A. gossypii on cotton is affected by tern-perature and aphid age whereas temperature influences also the ratio of sugars contained in the honeydew (Hen neberry et al., 2000).
As far as the role of host plant is concerned, it seems that in prey absence cucumber supports nymphal develop ment at a similar rate to that of eggplant and tomato, whereas reproduction, albeit lower than that on eggplant, was similar to that on tomato (21.55 and 8.28 eggs/female, on eggplant and tomato, respectively, at 20°C) (Perdikis, 2000).In accordance, cucumber infested with T. vaporariorum has been found to be a favorable diet for nymphs of M. pygmaeus to develop (Perdikis & Lykouressis, 2000).
Therefore, the reasons why development and reproduc tion of M. pygmaeus are adversely affected by the pres ence of A. gossypii, should be looked for in the specific host plant -prey combination, i.e. cucumber with A. gossypii.In fact the presence and feeding of this aphid on cucumber very likely results in the production of high amounts of honeydew deposited on the leaves.Addition ally, the dense hairs of cucumber leaves may retain the droplets of honeydew facilitating the entrapment of M. pygmaeus, as it was the case in E.formosa (van Lenteren, 1990).
Nymphal mortality appears thus to be caused by mainly the high amounts of honeydew and in a much lesser extent the sticky material from the cornicles.This is also supported by the low nymphal mortality of M. pygmaeus when fed with A. gossypii on eggplant, on which very little quantity ofhoneydew is produced.
In conclusion, the performance of an omnivorous predator in a host plant -prey system is affected not only by factors related to the plant or prey species, but rather the interactions of these trophic levels.Our results show that A. gossypii on cucumber inhibits M. pygmaeus popu lation increase.

Fig. 1 .
Fig. 1.Developmental period (a) and mortality rate (b) of Macrolophus pygmaeus nymphs on cucumber at various tem peratures with or without prey (Aphis gossypii).Bars at each temperature followed by different letters differ significantly (P < 0.05).

Fig. 2 .
Fig. 2. Survival rate of Macrolophus pygmaeus nymphs on cucumber at various temperatures with (a) or without prey (b) (Aphis gossypii).Numbers above x-axis represent the approximate starting points of each nymphal stage.

Fig. 4 .
Fig. 4. Age-specific survival rate of Macrolophuspygmaeus females fed on cucumber at various temperatures with (a) or without prey (b) (Aphis gossypii).
Means followed by different capital letters in a row for each parameter studied and by different small letters within a column are significantly different (x2-test, Tukey-Kramer HSD test, P < 0.05).Numbers in parentheses indicate the number of replicates in each experiment.

Table 2 .
Fecundity (eggs/female ± SEM) and longevity (days ± SEM) of Macrolophus pygmaeus when fed on cucumber leaves with or without the prey Aphis gossypii, at various temperatures.Means followed by different capital letters in a row for each parameter studied and by different small letters within a column are significantly different (x2-test, Tukey-Kramer HSD test, P < 0.05).Numbers in parentheses indicate the number of replicates in each experiment.

Table 3 .
Life table parameters of Macrolophus pygmaeus when fed with Aphis gossypii or without prey on cucumber, at various temperatures: R0,net reproductive rate; T, mean genera tion time (days); rm, intrinsic rate of increase ( /day); DT, dou bling time (days).