Parasitoid survival and fecundity is generally enhanced with access to carbohydrate food sources. In many agricultural ecosystems, there is often a scarcity of suitable carbohydrates for parasitoids. This study compared the suitability of aphid honeydew and buckwheat nectar as diet for the aphid parasitoid Lysiphlebus testaceipes. Wasp lifespan and egg load were both increased with access to carbohydrates, but no differences were detected between the various carbohydrates diets tested. As aphid honeydew is a sufficient source of nutrition and L. testaceipes is a short-lived species, adding additional sources of carbohydrates like floral nectar strips to the agricultural landscape is unlikely to significantly increase the biological control exerted by L. testaceipes.
While the benefit for parasitoids from feeding on carbohydrate-rich food sources is well documented (Wäckers 2001; Williams & Roane 2007; Wade et al. 2008; Wyckhuys et al. 2008; Vollhardt et al. 2010a), availability of suitable sources of carbohydrates in agricultural systems is often limited (Landis et al. 2000; Vollhardt et al. 2010b). Within agricultural systems, honeydew from their hosts or other sucking insects is often the most prevalent source of carbohydrates available to parasitoids (Wäckers 2005). Although honeydew might be available in large quantities, its accessibility and quality potentially reduces its value as a food source for some parasitoid species (Wäckers 2000, 2001; Lee et al. 2004; Hogervorst et al. 2007). Due to uncertainty over the quality, accessibility and availability of honeydew, growing nectar-producing plants adjacent to crops to provide sugar, pollen and shelter for a range of beneficial arthropods may be a useful tactic as part of an IPM program (Landis et al. 2000).
Buckwheat, Fagopyrum esculentum (Moench) is one such nectar-producing plant that is a candidate for use as a companion plant. It has been demonstrated to increase the lifespan of the aphid parasitoid Aphidius ervi Haliday (Araj et al. 2006; Wade & Wratten 2007) and the reproductive capacity of Diadegma insulare (Cresson) (Lee & Heimpel 2008a). Before implementing a companion planting program to improve parasitism of a pest species, it is important to understand adult parasitoid food requirements and feeding behaviour (Jervis & Kidd 1996). This is because despite the availability of nearby beneficial flowering plants, some parasitoids may still use naturally occurring honeydew if it is encountered while searching for hosts (Heimpel & Jervis 2005).
The aphid parasitoid, Lysiphlebus testaceipes (Cresson) is a potential biological control agent of cotton aphid, Aphis gossypii (Glover). Cotton aphid is a temporally variable and spatially patchy pest of cotton and other crops in Australia (Wilson et al. 2008). Populations tend to build up during the peak growing season of the crop (Room & Wardhaugh 1977; Wilson et al. 2007), but in some years, crops can be heavily infested early in the growing season. Early season infestation can result in problems with cotton bunchy top, an aphid-transmitted virus of cotton (Reddall et al. 2004). As insecticide resistance among A. gossypii populations makes control difficult (Herron & Wilson 2011), implementing IPM strategies for A. gossypii including the use of biological control is a challenge for the Australian cotton industry. Lysiphlebus testaceipes has the potential to contribute to the control of A. gossypii in cotton, but some aspects of its ecology are poorly understood. While its lifespan has been studied (Sekhar 1957; van Steenis 1994; Persad & Hoy 2003), only one of these studies has looked at the effect of carbohydrate diet on lifespan (Persad & Hoy 2003). Lysiphlebus testaceipes is a synovigenic species that does not host feed (Dieckhoff & Heimpel 2010), so carbohydrates consumed as adults are likely to be an important source of nutrition. Here, we examine the suitability of A. gossypii honeydew and buckwheat nectar as food sources for the aphid parasitoid, L. testaceipes.
Effect of sugar diet on egg load
Wasps were obtained by collecting mummies from a glasshouse colony of A. gossypii reared on hibiscus, Hibiscus rosa-sinensis (L.). Mummies were removed off leaves with a moistened paint brush and placed into a Petri dish. Mummies were then moved individually into gelatin capsules (size 0) and held at 25°C, 50% RH until emergence. Wasp emergence was monitored daily and newly emerged virgin wasps (<16 h old) were used in all experiments.
This experiment investigated the effect of a dietary carbohydrate on wasp egg load. The diets used were a carbohydrate-rich diet comprised of honey (Capilano brand, blend of eucalypt and ground floral honey) diluted in water (15% concentration) and a control diet, which was water alone. The effect of these diets on egg load was measured after four durations of exposure: 0, 24, 48 and 72 h. The experimental design was two diets each with four feeding durations. Each diet-duration combination was replicated 10 times. Replicates were kept in plastic Petri dishes (80 mm diameter, 12 mm depth) in a constant temperature room at 25°C. One female wasp was released into each dish. A disc (20 mm diameter) of absorbent paper towelling saturated with one of the diets was added to each dish, and, every 24 h, the diet was reapplied to the absorbent paper.
To estimate the effect of diet on potential fecundity, egg load was assessed. Egg load was measured by first killing the wasps in 70% alcohol and then dissecting them individually in cavity slides filled with saline solution (1.5 g NaCl/200 mL) and 1% detergent (Jones et al. 2003). Ovaries were carefully removed with a fine probe and forceps under a stereomicroscope at ×50 magnification. The ovaries were burst to disperse the eggs. The separated eggs contained on each slide were then counted under a compound microscope (×63 magnification).
The interaction effect of diet and wasp age on egg load was tested in a two-way analysis of variance (ANOVA). Where significant, means were compared using Fisher's least significant difference (LSD) test. All statistical analyses were carried out using GenStat (11th Edition).
Effect of various sugar diets on lifespan and egg load
Insects and plants
Wasps used in the experiment were collected as mummies from a glasshouse colony of A. gossypii reared on hibiscus. Parasitoids were individually stored in gelatin capsules and checked daily for emergence. Wasps were sexed using a stereomicroscope (×50 magnification). Newly emerged (<10 h old) virgin wasps were used in this experiment.
Cotton (variety Sicot 71) and buckwheat (F. esculentum) plants were grown in a glasshouse in 6 L pots containing soil and commercial potting mix (50:50). Cotton plants at approximately 10th node stage were infested with A. gossypii to produce honeydew contaminated leaves for experimental use (see later). Buckwheat and cotton plants were fertilised weekly with all-purpose liquid fertiliser.
Four treatments were tested to examine the effect of diet on wasp lifespan and potential fecundity: aphid honeydew, buckwheat nectar, honey solution and distilled water. To observe lifespan, male and female pairs of wasps were placed into individual Petri dishes (80 mm diameter, 12 mm depth) with one of the diet treatments. Ten replicates (dishes) for each diet treatment were used. To determine potential fecundity, single female wasps were placed in Petri dishes (80 mm diameter, 12 mm depth) and fed one of the four diets (honeydew, buckwheat nectar, honey solution or water) for durations of 24, 48, 72 or 96 h. For each diet-duration combination, 10 replicates were used.
With the exception of the buckwheat treatment, treatments were presented as 18 mm diameter discs. The honey and water treatments were offered as saturated paper towelling discs. For the honeydew treatment, a disc was cut from a cotton leaf containing droplets of fresh honeydew. The buckwheat was provided as four fresh flowers. In addition, each treatment had a paper towel disc (18 mm diameter) saturated with distilled water. For the duration of the experiment, the Petri dishes were kept in a constant temperature room at 25°C, 14:10 L:D photoperiod. All diets were provided in volumes in excess of daily requirements and were replaced daily, at the same time, wasp survival was recorded.
Potential fecundity of wasps was determined after they were allowed to feed on the treatments for 24, 48, 72 or 96 h. Egg load was determined by killing wasps in 70% ethanol and then dissecting them in a cavity slide under a stereomicroscope as detailed earlier. Hind tibia length (HTL) of wasps was measured using a stereomicroscope (×50 magnification) fitted with a micrometer to estimate wasp body size.
Data on lifespan and egg load were compared using ANOVA. Lifespan data were analysed in a two-way ANOVA with diet and wasp sex as treatments. Data from each of the feeding durations (24, 48, 72 and 96 h) were analysed by one-way analysis of covariance to test the effect of diet on wasp fecundity. Wasp body size is known to affect egg load (Chau & Mackauer 2001; He et al. 2005), so, in this study, HTL was used as a covariate to correct the egg load of wasps for size. A regression analysis was used to test the relationship between HTL and egg load: data from the three carbohydrate diets from 48 to 96 h feeding durations were used in the analysis. Where significant, means were compared using Fisher's LSD test. All statistical analyses were carried out using GenStat (11th Edition).
Effect of sugar diet on egg load
Wasps were aged between 1 and 16 h at the start of the study. Egg load of these wasps ranged from 7 to 210 mature eggs. At 24 h, egg load ranged from 14 to 354 eggs, confirming that the wasps have synovigenic egg maturation, i.e. they mature eggs over their lifetime (Fig. 1). The interaction effect of diet and duration on egg load was not significant (F3,72 = 1.58, P = 0.202). However, the main effects of diet and duration on egg load were both significant (P <0.05). Wasps fed the honey diet had an average egg load of 169.0 ± 16.0 eggs, while wasps fed water had 114.5 ± 12.4 eggs (F1,72 = 9.42, P = 0.003). Egg load increased over time in wasps fed either diet. Wasps at 48 and 72 h had 176.4 ± 19.0 and 187.3 ± 21.8 eggs, respectively, which was significantly (P <0.05) more eggs than wasps at 0 and 24 h which had 78.2 ± 10.8 and 125.0 ± 21.9 eggs, respectively.
Effect of various sugar diets on lifespan and egg load
The two-way ANOVA testing the interaction of diet and wasp sex on lifespan was not significant. The main effect of diet was significant (P <0.05); providing a carbohydrate-rich diet increased wasp lifespan (Fig. 2). Wasps that fed on any of the carbohydrate diets lived 2–3 days longer than wasps with access to water only. There was no significant difference (P > 0.05) in wasp lifespan between any of the carbohydrate diets. The main effect of wasp sex was not significant (P> 0.05); male wasps lived an average of 4.3 ± 0.4 days, while females lived an average of 3.8 ± 0.3 days.
Wasps readily fed on all the diets, including the buckwheat nectar. Using a linear regression, there was a significant (P <0.05) positive correlation between HTL and egg load, so HTL was included as a covariate when comparing potential fecundity of wasps on the different diets. At the start of the experiment, wasps were <10 h old and had a mean potential fecundity of 91.1 ± 8.8 eggs per female. At 24 h, there was no significant difference (P > 0.05) in the egg load of wasps fed any of the diets (Table 1). At 48 h, wasps fed on the water diet had matured significantly fewer (P <0.05) eggs than the wasps feeding on the carbohydrate diets. There was no significant difference (P > 0.05) between any of the carbohydrate diets at any of the time intervals (24, 48, 72 and 96 h).
Table 1. Mean egg load ± SE of Lysiphlebus testaceipes fed on diets of water, honey solution, buckwheat nectar or aphid honeydew for durations of 24, 48, 72 and 96 h
Means in a column followed by the same letter are not significantly different (P > 0.05).
Tibia corrected values used in the analysis are shown in brackets.
185.8 ± 16.8
210.0 ± 21.0
235.6 ± 24.5
266.8 ± 25.2
294.7 ± 39.2
270.0 ± 17.2
187.6 ± 27.4
254.3 ± 22.7
317.8 ± 23.9
285.3 ± 20.3
216.6 ± 13.4
252.0 ± 17.6
272.2 ± 25.0
283.2 ± 11.7
Wasps fed on honey, which contains several carbohydrates including fructose and glucose (Doner 1977), had higher potential fecundities than wasps fed the control diet of water. This result is a common finding for synovigenic parasitoids like L. testaceipes (Thompson 1999; Hogervorst et al. 2007). Wasps fed water alone matured similar numbers of eggs for up to 48 h. The ability to mature eggs on carbohydrate-poor diet is likely due to a combination of using stored energy (teneral reserves of sugars, glycogen and lipids) obtained during immature development to mature eggs (Rivero & Casas 1999; Olson et al. 2000; Jervis et al. 2008), having nutrient-poor eggs (Mackauer & Völkl 1993) and limited use of energy for other activities such as locomotion and oviposition (Hein & Dorn 2008; Jervis et al. 2008), as would occur in Petri dishes. Under the conditions experienced in the water treatment (limited food resources and no hosts present), some parasitoid species resorb eggs (Heimpel et al. 1997; Lee & Heimpel 2008b; Richard & Casas 2009). In these species, recovered resources from resorbed eggs can be used for somatic maintenance. However, aphidiid wasps do not appear to be capable of egg resorption (Le Ralec 1991; Mackauer & Völkl 1993). This may be because eggs are small and nutrient poor (Mackauer & Völkl 1993). If egg production costs are low and are produced mostly from teneral reserves, the tendency of aphid parasitoids like L. testaceipes to continue maturing eggs prior to locating hosts and food may be a viable strategy to maximise lifetime fecundity in patchy environments (Olson et al. 2000).
Feeding on any of the carbohydrate diets (honey, honeydew or buckwheat nectar) increased the lifespan of wasps compared with the water control. It is likely that to maximise lifespan, L. testaceipes requires an intake of carbohydrate every 1–2 days to supplement its teneral reserves (Azzouz et al. 2004; Jervis et al. 2008; Wu et al. 2008). There was no difference in the lifespan of wasps feeding on any of the carbohydrate diets, suggesting that the wasps were obtaining sufficient nutrition, regardless of the carbohydrate composition of each of the diets.
At 48 h, dietary carbohydrates increased the egg load of wasps compared with the water control. No comparisons between carbohydrate diets and water could be made past 48 h as all the wasps in the water treatment had died. Potential fecundity of wasps did not differ between any of the carbohydrate diets. Availability of carbohydrate appears to be very important for wasp survival, but of less importance in egg maturation (Jervis et al. 2008). Egg load appears to be greatly influenced by the volume of energy reserves stored by the parasitoid during immature development and is strongly correlated with adult body size.
Data from this experiment indicate that honeydew from A. gossypii is a suitable source of carbohydrates for L. testaceipes. In other studies, honeydew has been reported as a poor source of carbohydrates for parasitoids (Wäckers 2000; Lee et al. 2004; Wyckhuys et al. 2008). A possible reason for inferiority of honeydew compared with nectar is the high concentration of the oligosaccharide melezitose. Gustatory response and rapid crystallisation of melezitose may explain the poor performance of some parasitoids feeding on honeydew (Wäckers 2000). Other potential problems for parasitoids when feeding on honeydew include the presence of plant-secondary metabolites and the high viscosity of honeydew (Wäckers 2000; Faria et al. 2008; Hogervorst et al. 2009). Honeydew carbohydrate composition varies with aphid species so parasitoid lifespan can be affected by the source of the honeydew (Hogervorst et al. 2007). In a recent review of honeydew as a food source, Wäckers et al. (2008) found limited evidence that honeydew is inferior to other sugar sources.
Aphis gossypii honeydew composition on cotton has been reported as principally a mix of glucose, fructose, sucrose and either melezitose (Hendrix et al. 1992; Henneberry et al. 2000; Byrne et al. 2003) or erlose (Lawo et al. 2009). Why melezitose was not detected in the second study is not clear, but honeydew composition has been found to vary between different cotton varieties, developmental stage, age, ant attendance, presence of symbionts as well as rate and duration of infestation on the plant (Lawo et al. 2009). We did not analyse the composition of the honeydew used in our experiment, but the absence or a low concentration of melezitose may explain its suitability as a food for L. testaceipes. Further, offering honeydew with supplementary water may have increased the humidity in the Petri dish slowing the crystallisation process of the honeydew, increasing the duration of its availability.
Utilising aphid honeydew offers several advantages for natural enemies. Firstly, for some species including L. testaceipes, honeydew acts as a kairomone assisting the parasitoid in host location (Grasswitz & Paine 1993; Leroy et al. 2009), so time spent searching for hosts also locates food. This provides more time for host searching and for time-limited species, this may increase realised fecundity. Secondly, proximity of honeydew to the host reduces energy and time spent searching for nectar or other food sources which allows parasitoids to focus more on activities such as mating, host searching and oviposition. From a biological control point of view, obtaining sufficient nutrition from honeydew is an advantage in agricultural systems where flowering plants with accessible nectar are often scarce (Jervis et al. 1993; Jervis 1998; Landis et al. 2000). Although adding flowering plants on field margins may not significantly increase aphid parasitoid lifespan or parasitism especially if aphid honeydew is present in the crop and is palatable to the parasitoid (Vollhardt et al. 2010b). For example, a study of parasitoids collected in cabbage fields with floral borders showed that 80% of Cotesia glomerata (L.) and 55% of Microplitis mediator (Haliday) had fed on aphid honeydew instead of nearby flowers (Wäckers & Steppuhn 2003).
For aphid management in cotton with L. testaceipes, it appears that no increase in parasitoid fitness or aphid parasitism would be gained by adding floral resources to cotton fields. Cotton contains extra-floral nectaries that were not investigated in this study, but have been shown to be useful for other parasitoids (Röse et al. 2006). At times and/or places of the crop where aphids are absent, this source of carbohydrates could be important in sustaining parasitoids until they locate hosts and honeydew.
We thank Bernie Franzmann for initial postgraduate supervision, Susan Fletcher for statistical advice and Richard Lloyd for comments on an earlier version of the manuscript. Funding was provided by the Australian Cotton Research and Development Corporation (postgraduate research project DAQ 134).