Introduction

Oviposition in ladybirds (Coleoptera: Coccinellidae) is influenced by a number of prevailing abiotic (Wang et al. 2013; Papanikolaou et al. 2013; Singh et al. 2018) and biotic (Seagraves 2009; Omkar et al. 2009; Bista and Omkar 2013; Tang et al. 2013) factors. It is essential for the females to select oviposition sites, taking into accord a multitude of factors, since egg being an immobile stage, is vulnerable to numerous dangers, like predation, parasitization, and pathogen infection (Dixon 2000; Seagraves 2009). A strong correlation between oviposition preference and offspring performance is supposed to be a strategy of the mother to enhance her inclusive fitness (Omkar and Mishra 2005).

Ladybirds select oviposition sites based on factors, such as (1) patch quality in terms of prey quantity (Hemptinne et al. 2000a; Fréchette et al. 2006; Oliver et al. 2006; Omkar et al. 2010), (2) age of aphid colony (Hemptinne et al. 2000a), (3) age of host plant (Seagraves 2009), (4) presence of larval tracks (Ruzicka 2006; Mishra et al. 2012, 2013), (5) physical presence of adults (Mishra and Omkar 2006), (6) cost of inter-patch movement in terms of energy and time, and (7) mortality risks to the offspring (Seagraves 2009). A few studies suggest that egg presence also modulates oviposition (Hemptinne and Dixon 1991; Mishra et al. 2012, 2013). In response to poor prey quantity and quality, ladybirds are known to either reduce oviposition (Oliver et al. 2006; Fréchette et al. 2006; Omkar et al. 2010; Barbosa et al. 2014; Santos et al. 2016; Singh et al. 2016), or completely avoid oviposition by resorbing their ovarioles (Osawa 2005; Kajita and Evans 2009; Ferrer et al. 2010) or search for further new patches for oviposition (Sloggett and Majerus 2000). In prey-scarce conditions, instead of avoiding such patches completely, ladybirds tend to lay extra non-viable eggs only for the purpose of nutrition, so as to provide for their viable eggs in the same egg batch (Perry and Roitberg 2005) and also for themselves (Santi and Maini 2007). Laying of trophic eggs as a source of nutrition is found in many animal groups, e.g. fishes (McKaye 1986), amphibians (Kam et al. 2000; Heying 2001; Gibson and Buley 2004), insects (Frechette and Coderre 2000; Kudo and Nakahira 2001), spiders (Kim and Roland 2000; Kim and Horel 2000) and prosobranch snails (Baur 1992).

Poor prey conditions are also responsible for an increase in the incidence of cannibalism and intraguild predation (Takahashi 1989; Schellhorn and Andow 1999b; Dixon 2000; Cottrell 2005), with the former increasing more in ladybirds (Schellhorn and Andow 1999a; Hemptinne et al. 2000b, 2000c). This is probably owing to lesser chances of exposure to toxins in case of cannibalism (Burgio et al. 2002; Agarwala and Dixon 1992; Hemptinne et al. 2000b, 2000c). Sibling cannibalism in particular is suggested to have great adaptive significance in ladybirds by providing essential energy and nutrients, thereby accelerating development, increasing growth rate, and enhancing survival (Banks 1954, 1956; Kawai 1978; Roy et al. 2007). Sibling egg cannibalism is believed to have evolved under prey-scarce conditions due to fitness gains to both cannibals and victims (Osawa 1992).

While egg cannibalism is a more prevalent phenomenon, being an easy source of energy (Takahashi 1987; Agarwala and Dixon 1992), larval cannibalism has been investigated more in ladybirds (Michaud 2003; Pervez et al. 2006). In larval cannibalism, the role of size disparity (Yasuda et al. 2001), species specific context (Yasuda and Ohnuma 1999; Snyder et al. 2000), as well as kinship or relatedness (Michaud 2003; Pervez et al. 2005) have been observed. Results of these studies indicate a proclivity of smaller sized individuals and conspecifics to being cannibalized in prey-scarce conditions (Hodek et al. 2012). The few studies that have been undertaken on egg cannibalism in ladybirds, indicates the preference of conspecific eggs over heterospecific ones (Agarwala et al. 1998; Michaud 2002) with preference changing with the switching of egg surface chemicals (Omkar et al. 2004). This indicates the capability of discriminating between conspecifics and heterospecifics.

In view of the above facts on effect of prey scarcity on oviposition and cannibalism, we decided to assess the oviposition responses of female ladybirds under (a) prey-scarce and abundant conditions, and (b) presence of potential passive competitors in the form of previously laid conspecific eggs. While the prey scarcity condition is commonly found in nature owing to the ephemeral nature of aphids, the latter condition is more likely to occur in conditions of mass rearing or stock culture under laboratory conditions.

To address this we have varied prey density and egg presence to assess whether these factors modulated ladybird oviposition. Egg presence was varied in terms of age as well as relatedness, since the former signals the degree of danger of cannibalism in the future and the latter may have fitness consequences for the females. We hypothesize that in prey-scarce conditions, females will reduce their oviposition and percent egg viability. Secondly, females should lay fewer number of eggs in presence of related eggs. Lastly, we hypothesized that when females were placed in a constrained arena with conspecific eggs and no escape options, under prey-scarce conditions, they would cannibalize the older eggs as a nutrition source or to reduce competition as their hatching would danger the freshly laid eggs via cannibalism, and continue with oviposition. However, cannibalism would rarely happen under abundant prey conditions, while oviposition took place normally, as this prey density would be enough to support the life attributes. This study attempts to define the relationship between oviposition behavior and cannibalistic tendency of females in the presence of conspecific related and unrelated eggs of different ages.

To test these hypotheses, ladybird Menochilus sexmaculatus (Fabricius) (Coleoptera: Coccinellidae) was selected as the experimental model. It is a polymorphic ladybird beetle found abundantly almost throughout the year in area around Lucknow, India and feeds on a wide range of aphids (Omkar and Pervez 2004). Although studies related to reproductive biology have been conducted on this beetle (Omkar et al. 2006; Omkar and Bind 2007; Mishra et al. 2012, 2013), but there are no studies in the context of relatedness of eggs of different ages.

Materials and methods

Stock maintenance

Adults of M. sexmaculatus (30 males and 30 females) were collected from agricultural fields surrounding Lucknow, India (26° 50’N, 80°54′E). They were paired for mating in transparent plastic Petri dishes (9.0 × 2.0 cm2) under laboratory conditions (27 ± 1 °C; 65 ± 5% R.H.; 14 L: 10D). They were provided with an ad libitum supply of pea aphid, Acyrthosiphon pisum (Harris) grown on broad bean, Vicia faba L. in polyhouse cultures (21 ± 1 °C; 65 ± 5% R.H. 14 L: 10D). Females were isolated post mating and provided with prey every 24 h. Eggs laid were separated and incubated under the above abiotic conditions until hatching. Larvae were reared until adult emergence in plastic beakers (14.5 × 10.5 cm2; 5 larvae per beaker) under similar conditions as the parents. The adults for experiments were taken from the stock culture.

Experimental design

Unmated 15–20 day old females were monogamously paired with similar aged males in Petri dishes (9.0 × 1.5 cm2) for 2 days in the presence of excess A. pisum. These multiply-mated females were then removed and placed in fresh Petri dishes for oviposition and shifted to a new Petri dish after every 1 h between 0700 and 1900 h with excess of A. pisum, regardless of whether they had oviposited or not, which helped determine the age of the eggs to the nearest hour for future experiments. From these, twenty eggs were randomly selected and remaining excess eggs, if any, were removed with the help of wet fine camel hair brush. Position of these selected twenty eggs was marked with a permanent marker and they were left for 12, 24 and 48 h, for further use in experiments.

We tested the combined effect of prey density, egg relatedness and egg age as factors affecting oviposition. For this, the Petri dishes containing eggs of different ages (12, 24 and 48 h) were first placed in two groups of either abundant (50 mg) or scarce (5 mg) prey conditions; these densities were decided on previous standardization experiments. Each egg age-prey group was further divided into two sub groups in which either the mother or an unrelated female of similar age was introduced (Fig. 1). After 24 h of these treatments, the females were removed. The number of fresh eggs laid, percent egg viability ([the number of eggs hatched/total number of eggs] ×100) of these eggs and cannibalism by mothers of marked eggs, if any, were recorded. Incidences of cannibalism, if any, were identified by observing for egg traces (remains of eggs) under stereoscopic binocular microscope (Magnus, Olympus India PVT. Ltd. Noida) at 16x magnification. In control treatments, females were provided with abundant or scarce prey conditions in absence of eggs. Similar observations were made. All treatments were conducted with 20 replicates.

Fig. 1
figure 1

Flow chart showing brief methodology of experiment

Full size image

Statistical analysis

All data were subjected to Kolmogorov-Smirnov test of normality and Bartlett’s test of homogeneity of variances. Data were found to be normally distributed with equal variances.

Data obtained in experiment (presence of eggs of different ages at the two prey quantities) were first subjected to two-way ANOVA with prey quantity (abundant and scarce) and egg presence (no eggs [control] and egg presence [pooled data of different egg ages]) as independent factors; number of eggs laid by M. sexmaculatus females, percent egg viability and cannibalism by females were dependent factors followed by Tukey’s post hoc comparison of means test at 5%.

Since significant difference of egg age was seen from control (presented in result section), the data were subjected to another ANOVA after removing the control data. In this, data on number of eggs laid, percent egg viability and number of eggs cannibalized by mothers were subjected to three-way ANOVA with amount of prey (abundant and scarce), egg relatedness (related or unrelated) and egg age (12, 24, 48 h) as independent factors followed by Tukey’s post hoc comparison of means test at 5%. All statistical analyses were performed using MINITAB 15.0.

Results

Effect of prey quantity, egg relatedness and egg age on oviposition

Results of two-way ANOVA revealed that oviposition varied significantly with prey quantity (F1,139 = 6995.34, P < 0.001). Also, oviposition in control was significantly higher than the pooled egg age treatments (F3,319 = 27.20, P < 0.001). The interaction between both independent factors had insignificant effect (Fprey quanity × egg age 3,319 = 1.45, P > 0.05) on oviposition.

Since the effect of egg age was significantly different, the control was discarded and the remaining data (prey quantity, egg age and relatedness) were subjected to three-way ANOVA. Females of M. sexmaculatus oviposit significantly more eggs when provided with abundant prey (F1,228 = 4782.697, P < 0.001). In all egg age treatments, oviposition was significantly reduced in presence of eggs, either related or unrelated, in the presence of scarce or abundant prey (F1,228 = 24.769, P < 0.001) (Fig. 2). There was however no effect of egg age in any treatment (F2,228 = 2.256, P > 0.05). The interaction between prey quantity and egg relatedness was significant (Fprey quantity × egg relatedness 1,228 = 4.071, P < 0.05). However, the interactions, viz. Fprey quantity × egg age 1,228 = 2.280, P > 0.05; Fegg relatedness × egg age 2,228 = 0.147, P > 0.05; Fprey quantity × egg relatedness × egg age 2,228 = 2.081, P > 0.05 were insignificant.

Fig. 2
figure 2

Oviposition by M. sexmaculatus in the presence of different age eggs under abundant and scarce prey conditions. Values are mean ± SE. Large letters represent comparison of means between abundant and scarce prey conditions. Small letters represent comparison of means between egg relatedness treatments. Large letters in parenthesis represent comparison of means between egg age. Similar letters indicate lack of significant difference at P > 0.05

Full size image

Effect of prey quantity, egg relatedness and egg age on percent egg viability

Initial ANOVA of pooled data revealed that percent egg viability was not significantly influenced by prey quantity (F1,319 = 0.20, P > 0.05) and egg age (F3,319 = 0.32, P > 0.05). The interaction between the two independent factors (Fprey quantity × egg age 3,319 = 0.32, P > 0.05) was also insignificant.

The subsequent three-way ANOVA after removing control data, revealed that there was insignificant effect of prey quantity (F1,228 = 0.351, P > 0.05), relatedness of eggs (F1,228 = 0.310, P > 0.05) and egg age (F1,228 = 0.159, P > 0.05) on percent egg viability. All interactions (Fprey quantity × egg relatedness 1, 228 = 0.004, P > 0.05; Fprey quantity × egg age 1,228 = 0.343, P > 0.05; Fegg relatedness × egg age 2,228 = 0.146, P > 0.05; Fprey quantity × egg relatedness × egg age 2,228 = 0.067, P > 0.05) had insignificant influence on percent egg viability (Fig. 3).

Fig. 3
figure 3

Percent egg viability of M. sexmaculatus in the presence of different age eggs under abundant and scarce prey conditions. Values are mean ± SE. Large letters represent comparison of means between abundant and scarce prey conditions. Small letters represent comparison of means between egg relatedness treatments. Large letters in parenthesis represent comparison of means between egg age. Similar letters indicate lack of significant difference at P > 0.05

Full size image

Effect of prey quantity, egg relatedness and egg age on cannibalism by females

Initial analysis of data revealed that cannibalism was significantly affected by prey quantity (F1,319 = 101.82, P < 0.001), egg age (F3,319 = 48.54, P < 0.001) and their interactions (Fprey quantity × egg age 3,319 = 15.05, P < 0.001).

The subsequent three-way ANOVA revealed that cannibalism varied significantly with prey quantity (F1,228 = 102.934, P < 0.001), egg relatedness (F1,228 = 6.179, P < 0.05) and egg age (F2,228 = 4.869, P < 0.05). Cannibalism was significantly higher in prey-scarce conditions, with related eggs and older eggs being cannibalized significantly more in prey-scarce conditions (Fig. 4) but not in prey abundant conditions (Fig. 4). While interaction between prey quantity and egg age was significant (Fprey qauntity × egg age 1,228 = 4.251, P < 0.05), the rest of the interactions were insignificant (Fprey quantity × egg relatedness 1,228 = 2.199, P > 0.05; Fegg relatedness × egg age 2,228 = 0.018, P > 0.05; Fprey quantity × egg relatedness × egg age 2,228 = 0.077, P > 0.05).

Fig. 4
figure 4

Egg cannibalism by M. sexmaculatus in the presence of different age eggs under abundant and scarce prey conditions. Values are mean ± SE. Large letters represent comparison of means between abundant and scarce prey conditions. Small letters represent comparison of means between egg relatedness treatments. Large letters in parenthesis represent comparison of means between egg age. Similar letters indicate lack of significant difference at P > 0.05

Full size image

Discussion

Our results reveal that oviposition as well as cannibalism by females was influenced by combined effect of prey quantity, egg relatedness and egg age. As predicted in our hypothesis, females of M. sexmaculatus oviposited less in prey-scarce conditions. Such an expected decrease in oviposition with decrease in prey quantity has been reported in ladybirds, Coccinella septempunctata L. (Xia et al. 1999) and Anegleis cardoni (Weise) (Omkar et al. 2010). This phenomenon is easily explained through reduced availability of nutrients in general and non-availability of significant dietary proteins in particular, which assist in oogenesis or chorion development, as observed in seed beetle, Stator limbatus (Horn) (Fox et al. 1996), Callosobruchus maculatus (F.) (Cope and Fox 2003) and pentatomid bug, Podisus nigrispinus (Dallas) (Lemos et al. 2001). Resorption of eggs under prey-scarce conditions in anticipation of prey abundant conditions in the future as has been reported in the butterfly Speyeria mormonia (Boisduval) (Boggs and Ross 1993).

However, the percent egg viability was not influenced by prey-scarce conditions. The males in our study had not been kept in prey-scarce conditions thus indicating that percent egg viability is influenced by male nutritional status and not by female nutritional status. Such influence of males on egg viability in ladybirds is supported by earlier studies (Omkar et al. 2004; Pervez et al. 2004; Michaud et al. 2013). In fact we have observed in another study that if eggs of a single clutch of M. sexmaculatus are separated and kept singly to prevent cannibalism, then around hundred percent hatching occurs.

The deterrent effect of related egg on oviposition by the ladybird is in support of our hypothesis and is probably a means of reducing the chances of cannibalism from close relatives. This behavior probably decreases the likelihood of competition between one’s own offspring, thereby indirectly enhancing fitness. Discrimination between conspecifics and heterospecifics (Yasuda et al. 2001; Yasuda and Ohnuma 1999) as well as in kin (Pervez et al. 2005; Joseph et al. 1999) during larval cannibalism has been observed. Mishra et al. (2012) have also reported oviposition avoidance in presence of conspecific eggs, but they did not check for relatedness. The ability to recognize self and non-self in ladybirds has been reported (Pervez et al. 2005) and sibling egg cannibalism is probably more beneficial for females as they are more resource sensitive (Michaud and Grant 2004). Swapping of egg infochemicals has indicated their role in differentiating between con- and heterospecifics (Hemptinne et al. 2000b; Omkar et al. 2004) but not amongst kin and non-kin in ladybirds. It is possible that such infochemicals might allow females to discriminate between related and unrelated eggs.

However, while females refrained from ovipositing in presence of related eggs, they consumed them significantly more than unrelated ones under prey-scarce conditions. It has been reported previously that ladybird females are likely to suffer less loss of inclusive fitness on consumption of own eggs under prey-scarce conditions (Osawa 1992). Previous studies have shown the ability to discriminate and preferably consume conspecific eggs over heterospecific ones in some ladybirds (Hemptinne et al. 2000b; Michaud 2002) and such discrimination has been attributed to (i) similar surface alkanes, and (ii) presence of palatable and nutritious alkaloids in conspecifics. While a couple of studies have shown discrimination ability amongst kin and non-kin (Agarwala and Dixon 1993; Pervez et al. 2005), the mechanism of discrimination is yet to be elucidated.

Not only related but older eggs were cannibalized more. The increased consumption of older eggs by the ovipositing females may be ascribed primarily to removal of an immediate threat to her offspring, since they are more likely to cannibalize the eggs of the female post hatching. The preference for the consumption of older eggs (24 and 48 h) by the ladybird females has also been reported (Timms and Leather 2007). But this discrimination against older eggs during cannibalism did not extend to oviposition, which was similar in the presence of eggs of all ages.

Our study thus establishes that ladybirds are able to modulate their oviposition in response to prey quantity as well as relatedness of eggs in the vicinity, while at the same time not being influenced by the age of eggs. While the females seem to be able to discriminate amongst sites in their search for suitable oviposition sites, they probably additionally make their sites even more suitable by cannibalism of previously laid eggs, especially the older ones. The cannibalism of their own eggs under prey-scarce conditions seems to be a survival strategy. Also of interest in this study, it was shown that percent egg viability is not influenced by female nutritional status, but by male nutritional and ecological status.