Predation by Macrolophus pygmaeus ( Hemiptera : Miridae ) on Acyrthosiphon pisum ( Hemiptera : Aphididae ) : In fl uence of prey age / size and predator ’ s intraspeci fi c interactions

Macrolophus pygmaeus (Hemiptera: Miridae) is an important predator of pests of horticultural crops and here its ability as a predator of Acyrthosiphon pisum (Hemiptera: Aphididae) is addressed for the fi rst time. The percentage predation of the different aphid instars and the number partially consumed were studied. Our results, obtained using choice and no-choice tests, revealed that M. pygmaeus caught and consumed more young than later instars of A. pisum, which confi rms results of previous studies using other species of aphids. We also studied the interactions between predators (male/female) foraging in the same patch. When the prey/predator ratio is kept constant at 10 : 1 the average percentage of aphids completely consumed by individual females or males does not change with increase in the number of foraging predators. However, the number of partially consumed aphids decreased when females shared the same patch. In contrast, there was an increase in the number of aphids partially consumed when two males shared the same patch. The results were discussed in terms of potential predator foraging strategies since intraspecifi c competition is a key factor modulating the dynamics of prey-predator systems.


Insect cultures
M. pygmaeus came from a colony established in a laboratory in Potenza, Italy on potted tomato plants kept at 21°C (18L : 6D; 80-90% RH).This colony was established and renewed with individuals purchased from Koppert Italia srl., which were reared under similar conditions to the original culture.Sterilized Ephestia kuehniella Zeller (Lepidoptera: Pyralidae) eggs, purchased from the same company, were used to supplement their diet.For this study we chose the adults of M. pygmaeus because they are more mobile and therefore more likely to move to a non-host plant attracted by the presence of prey.
Acyrthosiphon pisum was reared and maintained on broad bean plants (Vicia faba c.v. Agua dulce) in a climatic chamber at 21°C, 80-90% RH and a 18L : 6D photoperiod.The colony was started in 1985 with a few hundred specimens collected during spring from Medicago sativa L. (Fabales: Fabaceae) near Salerno, Italy (40°37´01˝N, 15°03´23˝E) and reared in the laboratory on broad bean plants.Broad bean plants were grown in pots (10 cm diameter) containing commercial soil (COMPO SANA® Universal Potting Soil) in a greenhouse.
For experiments that required same-aged aphids of each nymphal instar, approximately 100 adult virginoparae females were isolated from the aphid culture and put on a fresh potted broad bean plant kept in a plastic box (22 × 15 × 40 cm high) for 24 h at 21°C (18L : 6D; 80-90% RH).Females were then removed and lay eggs on it and its offspring successfully complete their development feeding on A. pisum infesting this plant (pers.observ.).
Our research aimed to assess the consumption and feeding preferences of M. pygmaeus for each nymphal instar of the pea aphid in no-choice and choice tests.According to Optimal Foraging Theory, prey size/age can potentially affect the number of each prey item consumed, depending on the intrinsic value of the prey in terms of its nutrient and energy content and handling time (the time a predator requires to extract the food from a prey item) (Pulliam, 1974;Pyke et al., 1977;Pyke, 1984).The handling time is infl uenced by the size, mobility and defensive ability of the prey (Sih & Christensen, 2001;Jeschke et al., 2002).However, the number of prey consumed is also determined by competition between individuals of the same species (Andersen, 1960;Klomp, 1964).In this sense, intraspecifi c competition is a key factor, which also modulates the dynamics of prey-predator systems (Andersen, 1960;Klomp, 1964;Lucas & Alomar, 2002;DeLong & Vasseur, 2011).
Diverse types of potential interactions may occur between two or more individuals of the same species when competing for food, space and/or oviposition sites (Klomp, 1964;Alley, 1982;Symondson et al., 2002;DeLong & Vasseur, 2011).When predators compete for prey, one often observes non-interactive as well as interactive individuals, which by mutual interference or synergic interactions (density dependent predation), infl uence in different ways prey population dynamics (Klomp, 1964;Alley, 1982;Losey & Denno, 1998;DeLong & Vasseur, 2011).Until now, the effect of intraspecifi c competition between individuals of M. pygmaeus has only been recorded in two studies in which whitefl y was the prey (Lucas & Alomar, 2002;Moreno-Ripoll et al., 2012).In the present study we report the effect of intraspecifi c interactions between individuals of M. pygmaeus on its consumption of A. pisum as these interactions could infl uence prey population dynamics.
Studying how prey size/stage and intraspecifi c interactions between predators can infl uence the numbers of A. pisum consumed by M. pygmaeus is important in determining whether it is likely to be an effective biological control agent of aphids and relevant to the current tendency to integrate M. pygmaeus into biological control programmes against aphids (Perdikis et al., 2008;Pérez-Hedo & Urbaneja, 2014).discarded.The neonate nymphs were maintained as a synchronous colony on a broad bean plant at 21°C for specifi c periods of time: 24, 48, 72 and 120 h; at the end of which the nymphs were, respectively, in L1, L2, L3 and L4 nymphal instars.Nevertheless, before their use in the experiments, aphids were examined under a stereo-microscope and all nymphs that were not in the appropriate stage, based on the morphological features described in Digilio (1995), were discarded.

Number of aphids killed by Macrolophus pygmaeus in no-choice tests
The predatory activity of females of M. pygmaeus based on the number of each of the four nymphal stages of A. pisum consumed was fi rst assessed in a no-choice experiment.A cut broad bean leaf, with the stalk inside an Eppendorf tube fi lled with water and sealed with parafi lm to prevent desiccation, was placed in a plastic cylinder (diameter: 5.5 cm, height: 7.5 cm, volume: 150 ml) with a mesh covered ventilation hole in the screw top (4.5 cm in diameter).Aphids of the same instar were gently placed on a leaf and allowed to settle.Preliminary experiments showed that careful manipulation caused negligible aphid mortality.Two prey densities, ten and twenty aphids per leaf, were tested.One M. pygmaeus adult female, taken from the rearing culture, which had emerged one week before the start of the experiment, was introduced into the cylinder and allowed to forage for 24 h.These females were not starved before the experiment, indeed they were taken directly from the rearing cage, where E. kuehniella eggs were available ad libitum.The predator was allowed to forage for 24 h and subsequently removed.The number of dead aphids were recorded, making a distinction between those completely consumed, those partially consumed and those unconsumed.An aphid was regarded as completely consumed when the predator had extracted its body fl uids and only its exoskeleton remained.Killed but unconsumed aphids were those found dead with a droplet on their siphunculi, as described in Fantinou et al. (2009).In addition, a small melanized area on the abdomen was generally evident.Partially consumed aphids were those in which the body contents were not completely removed by sucking.Different levels of partially consumed aphids were observed: with a third, half or more than half of their body contents removed.For each aphid instar, there were 25 and 17 replicates of the experiments with 10 and 20 aphids, respectively.

Feeding preferences of Macrolophus pygmaeus in choice tests
This predator's preference for different aphid instars was assessed in choice tests where two aphid instars were provided, which involved a total of six sets of replicated choice tests: 1 st vs 2 nd , 1st vs 3 rd , 1 st vs 4 th , 2 nd vs 3 rd , 2 nd vs 4 th and 3 rd vs 4 th .There were thirty replicates of each aphid size/stage combination with the exception of the 2 nd vs 4 th test, for which there were forty replicates.All trials were carried out as described above for the no-choice experiment, but the predator was provided with a total of twenty aphids made up of ten of each instar/choice.

Intraspecifi c interactions
Predatory activity of both M. pygmaeus males and females was assessed at different predator densities, keeping the A. pisum nymphs/predator ratio constant, at 10 : 1.To assess the density and/or effect of sex on the numbers of aphids killed by the interacting predators, the effect of prey age/size was kept constant by always providing them with only 2 nd instar aphids.Prey consumption in each of six different male and/or female M. pygmaeus combinations, was recorded after 24 h (Table 1).Partially consumed prey was also recorded.

Statistical analyses
Numbers of completely and partially consumed prey were analyzed using two-way ANOVAs with the "prey instar" (four levels) and "prey density" (two levels, ten and twenty aphids) as fi xed effects.Tukey post-hoc tests for multiple comparisons of means were also performed.
Predator preferences in terms of the number of aphids killed in the different aphid instar pairings (dichotomous choice tests) were analyzed using pairwise t-tests.Data on predatory activity were always calculated in terms of the number of aphids consumed per predator, and were analyzed using one-way ANOVA with female/ male combination as the main factor, followed by a Tukey posthoc comparison of means.Statistical analyses were performed using R 3.0.1 for windows (www.r-project.org).The numbers of completely consumed aphids, those killed but not consumed and those partially consumed were recorded.Given the overall low incidence of aphids killed but not consumed during this assay, we grouped them together with the number of partially consumed prey in the statistical analysis.In the text and graphs the aphids killed but not consumed are recorded as partially consumed.

Number of aphids killed by Macrolophus pygmaeus in no-choice tests
In the no-choice assay, there were signifi cant differences in the number of aphids of the different aphid instars completely and partially consumed (F 3,155 = 88.2,P < 0.001 and F 3,145 = 14.1, P < 0.001, respectively, Fig. 1A and B) and at the different aphid densities (F 1,155 = 33.4,P < 0.001 and F 1,145 = 3.95, P < 0.05, respectively, Fig. 1A and B).In general, at both prey densities, more of the young than of the older aphids were completely consumed (Tukey post-hoc tests: P < 0.001), although more aphids were consumed when twenty aphids were provided rather than ten.More interestingly, there were signifi cant interactions between "aphid instar" and "aphid density" in terms of both completely and partially consumed aphids (F 3,155 = 13.2,P < 0.001 and F 3,145 = 5.2, P < 0.001, respectively).When twenty aphids were provided M. pygmaeus completely consumed more fi rst and second instar aphids than they did when only 10 aphids were provided, whereas for the other two aphid stages the number completely consumed was similar at both densities.When provided with ten aphids, signifi cantly more second instar than the other instars of aphids were consumed (Tukey post-hoc test: P < 0.05, Fig. 1A), whereas when twenty aphids were provided the number of partially consumed aphids of the third instar was signifi cantly higher than for the other three aphid stages (Tukey post-hoc test: P < 0.05, Fig. 1B).

Feeding preferences of Macrolophus pygmaeus in choice tests
The numbers of aphids killed by each female in the pairwise tests (Choice assay) is shown in Fig. 2. For each instar combination, the number of younger aphids killed was greater than the number of older aphids (1 st vs 2 nd : t 29 = 5.76; 1 st vs 3 rd : t 29 = 9.09; 1 st vs 4 th : t 29 = 14.79; 2 nd vs 3 rd : t 39 = 7.6; 2 nd vs 4 th : t 27 = 9.09; 3 rd vs 4 th : t 29 = 7.73; P < 0.001 in all cases).

Intraspecifi c interactions
Fig. 3 shows the number of aphids partially and completely consumed by a single M. pygmaeus when on its own and in the presence of other conspecifi cs.There were signifi cant differences in the total number of aphids killed per predator in the different female/male M. pygmaeus combinations (F 5,248 = 50.9,P < 0.001).These differences are mainly due to the low number of aphids killed by males.The average number of aphids killed by each female was the same in the treatments with different numbers of females (F1, F2 and F5 treatments).The number of partially consumed aphids differed statistically in the different female and/or male combinations (F 5,208 = 12.2, P < 0.001).For females, the lowest number of partially consumed aphids was recorded at the highest female density (Tukey post-hoc test: P < 0.001 for F5 vs F1 and P < 0.05 for F5 vs F2, Fig. 3).In contrast, when two males of M. pygmaeus were put together, the number of aphids consumed per individual did not change, but the number of aphids partially consumed increased (Tukey post-hoc test: P < 0.01, Fig. 3).Also, the number of aphids partially consumed per predator increased when a male was present together with a female in comparison to that recorded when only one female (Tukey post-hoc test: P < 0.05, Fig. 3) or one male was present (Tukey post-hoc test: P < 0.01, Fig. 3).

DISCUSSION
In the present study, although the ability of M. pygmaeus to complete its development feeding only on A. pisum was not checked, we showed that at least in the short term, A. pisum is a suitable prey for M. pygmaeus females and males.The predatory activity of adults was evaluated because they colonize and make use of new hosts and prey resources.In this case, Vicia faba infested with A. pisum was the new host-plant system for M. pygmaeus.
The expectation was that adult M. pygmaeus would be more voracious than the fi fth instar nymph.For this reason we cannot compare the number of A. pisum consumed by M. pygmaeus with previously reported numbers of M. persicae and M. euphorbiae consumed, as fi fth instar nymphs were used in those studies (Lykouressis et al., 2007;Fantinou et al., 2008Fantinou et al., , 2009)).The difference in voracity of these two stages could explain why, for example, when ten fi rst instar aphid nymphs of A. pisum were provided, M. pygmaeus in this study consumed seven, whereas it consumes 3.54 M. persicae and 3.24 M. euphorbiae (Lykouressis et al., 2007), despite the fact that fi rst instar nymph of A. pisum weighs twice as much as M. euphorbiae and four times as much as M. persicae.
On the other hand, M. pygmaeus consumed a greater number of the younger A. pisum instars in both choice and no-choice tests, which is in accordance with the reports in the literature for M. euphorbiae and M. persicae (Lykouressis et al., 2007;Fantinou et al., 2008Fantinou et al., , 2009;;Pérez-Hedo & Urbaneja, 2014).Although older A. pisum instars are larger than young ones and potentially contain more nutritional resources, they are more diffi cult to kill since they can defend themselves better than young ones by kicking, dropping, running or walking away from predators (Dixon, 1958;Evans, 1978;Roitberg & Myers, 1978, 1979;Gerling et al., 1990;Wyckhuys et al., 2008).This defensive behaviour probably leads to an increase in the average handling time (and energy) spent by M. pygmaeus when attacking an older aphid, which reduces the time available to catch and consume further prey.Optimal Foraging Theory states that the fi tness of a foraging organism is a function of the amount of energy gained while foraging and that natural selection favours those organisms that, while foraging, are more effi cient in acquiring energy (Krebs, 1977;Pyke et al., 1977).M. pygmaeus preference for feeding on young pea aphids could be because the net gain in energy when young aphids are the prey is greater than when older aphids are the prey.
A greater number of pea aphids were consumed (completely plus partially) when more were provided, which is similar to the results previously reported for M. pygmaeus feeding on M. persicae (Fantinou et al., 2008(Fantinou et al., , 2009)), as is the behaviour of killing aphids but not consuming them (Alvarado et al., 1997;Fantinou et al., 2008Fantinou et al., , 2009)).However, in studies on M. persicae, M. euphorbiae and A. gossypii (Alvarado et al., 1997;Fantinou et al., 2008Fantinou et al., , 2009)), the presence of partially consumed aphids in addition to unconsumed ones was not recorded.
In the present study the presence of partially consumed aphids might be related to the greater size of A. pisum.Large prey usually take longer to kill and completely consume (Flinn et al., 1985;Baily, 1986;Foglar et al., 1990;Milonas et al., 2011).A disadvantage of preying on large prey item is they take longer to kill and consume, which increases the risk of the predator being attacked (Sih, 1980b;Abrams, 1982).Attacking and feeding on large prey also reduces the time available to the predator for other fi tnessenhancing activities (e.g.ovipositing) (Abrams, 1982).Thus, in the case of large prey, to only partially consume it could be adaptive, because Optimal Foraging Theory (Krebs, 1977;Pyke et al., 1977) predicts that predators should optimize the food intake per unit of feeding time by selectively feeding on the most easily digestible or most nutritionally valuable parts of its prey (Sih et al., 1980a).Consequently, this might account for why pea aphids were abandoned before they were completely consumed.
Partial consumption of prey is an important factor to be considered when studying the predatory behavior of a biocontrol agent because it infl uences the number of prey kill when attacking different prey population densities of p, which is referred to as the functional response and includes attack rate, time taken to handle prey and predator/ prey size ratios (Cohen, 1995).
In the last part of this study we addressed for the fi rst time how intraspecifi c interactions among different numbers of mirid predators infl uences the number of prey killed when M. pygmaeus feeds on aphid prey.In this biossay, we did not record cannibalism between conspecifi cs when feeding on aphids, which confi rms results of previous studies on different species of prey (Lucas & Alomar, 2002;Moreno-Ripoll et al., 2012).Cannibalism is only recorded for adult females of M. pygmaeus kept at high densities and only fed leafl ets of tomato (Moreno-Ripoll et al., 2012).Here, we maintained a constant predator-aphid ratio of 1 : 10 when the female/male density was increased.Under these conditions, in the treatments without males, the average number of aphids killed by each female did not change with increase in the number of females.Foraging theory predicts that the number killed per predator should increase with increase in prey abundance because the prey encounter rate increases, reducing the time spent by a predator searching for prey (Sih, 1980b;Abrams, 1982).This is only true if predators foraging in the same patch do not disturb each other.In the absence of interactions between predators, a higher individual consumption of aphids is expected in patches with higher numbers of predators at higher prey densities.For both males and females the number of aphids killed did not change when the number provided was increased.This could be a consequence of the females interacting with one another and in the case of males by their low voracity.
The number of aphids eaten by males is clearly less than by females.The greater consumption of prey by females is associated with the need to support reproduction, which is a major factor determining their fi tness (Pyke et al., 1977;Reznick, 1985;Zera & Harshman, 2001;Harshman & Zera, 2007).Since M. pygmaeus females have telotrophic ovaries, egg production requires a continuous supply of energy throughout adult life with egg development independent of mating (Castañé et al., 2007;Franco et al., 2011).Arnó et al. (2003) and Urbaneja et al. (2009) report that M. pygmaeus females consume more prey than males when fed on Lyriomyza trifolii (Burgess) (Diptera: Agromyzidae) larvae and Tuta absoluta (Meyrick) (Lepidoptera: Gelechiidae) eggs, which is similar to our results.
Regarding partial prey consumption, it is likely that a predator leaves already killed but unconsumed or partially consumed prey if disturbed by a conspecifi c while feeding.This should occur more frequently as the number of predators in the same patch increases.Therefore, at the highest predator density a higher level of partial prey consumption was expected.However, we found that partial aphid consumption decreased or increased with increase in predator density, depending on the sex of the interacting individuals.In the case of females, the number of partially consumed aphids decreased when more than one female shared the same patch; the opposite pattern was recorded for males.
The interaction between M. pygmaeus females seems to result in them maximizing their exploitation of individual prey items when there is competition for a limited resource.In this way, it is possible that reproductive females optimize their food intake by avoiding prey waste as female density increases.
In contrast, the interaction between males results in an increase in the number of partially consumed aphids probably due to mutual interference when foraging, or alternatively, to an increase in the aphid encounter rate due to their higher relative abundance.Studies on other predators have also documented an increase in partially consumed prey with increase in prey abundance (Sih et al., 1980a;Samu & Bíró, 1993;Lang & Gsodl, 2003).Given that males consumed relatively few aphids, then when a male and female interacted, the presence of the male may not have significantly affected the number of aphids entirely consumed by a single female, but the number of partially consumed aphids increased.
In conclusion, females and males infl uence in different ways the number of pea aphids consumed depending on the way in which they interacted.It is important, however, to highlight that plant architecture and plant surface may affect these interactions.Moreover, we cannot totally exclude that aphid behaviour could have played a role in the observed interactions between predators.Therefore, these results cannot be generalized without further study.
Commercially available M. pygmaeus are primarily sold to control whitefl ies and Tuta absoluta in tomato crops, although prospects for its use as a biocontrol agent of aphids in other crops was recently considered (Perez-Hedo & Urbaneja, 2014).In this context, our results are relevant and highlight the importance of considering aphid size/age in developing models that describe the effect of mirid predation on the age structure of aphid colonies.We also demonstrate that intraspecifi c competition is a key component in predator-aphid dynamics and predicting the effi ciency of M. pygmaeus for controlling pest aphids, specially when it is used in augmentative biological control.
M. pygmaeus predation on A. pisum is described here for the fi rst time.This novel system can be used to test predictions of theoretical models of inter-and intraspecifi c interactions, as done here and in another recently published work (Trotta et al., 2015).In the fi eld M. pygmaeus has never been reported feeding on the pea aphid probably because this predator mainly forages on plants belonging to families other than legumes (Alomar et al., 1994).However, the observation that M. pygmaeus feeds on A. pisum in the laboratory indicates some interesting lines of research.For example, it would be interesting to study whether the intercropping of legumes and Solanaceae plants results in an increase in the number of M. pygmaeus on legumes.If this were the case then this mirid bug could positively contribute to the biological control of pea aphid.Moreover, A. pisum and M. pygmaeus might encounter each other in tomato greenhouses when the aphid parasitoid Aphidius ervi is introduced on banker plants (Huang et al., 2011).In this case, interference by M. pygmaeus is possible, so it would be interesting to study the extent of this phenomenon and the conditions affecting it.

Fig. 1 .
Fig. 1.Number of A. pisum nymphs (mean ± SE) completely or partially consumed by one M. pygmaeus female in 24 h when provided with 10 (A) and 20 (B) aphids.Signifi cant differences among means based on Tukey post-hoc tests for multiple comparisons are indicated by different letters.Upper-case letters on bars indicate highly signifi cant differences (P < 0.001, Tukey's test) for the number of entirely consumed aphids and lower-case letters indicate signifi cant differences for the number of partially consumed aphids (P < 0.05, Tukey's test).

Fig. 2 .
Fig. 2. Number of aphids eaten (mean ± SE) by a M. pygmaeus female in 24 h when provided with different combinations of instars of A. pisum (*** P < 0.001, pairwise t tests).

Fig. 3 .
Fig.3.Numbers of completely and partially consumed aphids (mean ± SE) per predator recorded in the different combination of females and/or males of M. pygmaeus over two days.Upper-case letters on bars indicate highly signifi cant differences (at least P < 0.05, Tukey's test) for the number of completely consumed aphids and lower-case letters indicate signifi cant differences in the number of partially consumed aphids (at least P < 0.05, Tukey's test).1F -one female; 2F -two females; 5F -fi ve females; FM -one female and one male; 1M -one male; 2M -2 males.

Table 1 .
Combinations of M. pygmaeus females and/or males used in experiment 3 (Intraspecifi c interactions).