Fire Ants Protect Mealybugs against Their Natural Enemies by Utilizing the Leaf Shelters Constructed by the Leaf Roller Sylepta derogata

The importance of mutualism is receiving more attention in community ecology. In this study, the fire ant Solenopsis invicta was found to take advantage of the shelters constructed by the leaf roller Sylepta derogata to protect mealybugs (Phenacoccus solenopsis) against their natural enemies. This protective effect of fire ant tending on the survival of mealybugs in shelters was observed when enemies and leaf rollers were simultaneously present. Specifically, fire ants moved the mealybugs inside the shelters produced by S. derogata on enemy-infested plants. Compared with that in plants without ants, the survival of mealybugs in shelters in the presence of natural enemies in plants with ants markedly improved. Both the protection of ants and the shelters provided by leaf rollers did not affect the survival of mealybugs in the absence of enemies in plants. Ants and leaf rollers significantly improved the survival of mealybugs in predator-infested plants, whereas no such improvement was observed in parasitoid-infested ones.


Introduction
Mutual interactions between hemipterans and invasive ants, which are known for their aggressiveness and high colony density, occur extensively in ecosystems [1][2][3]. The relationship between ants and hemipterans is characterized by the protection that ants provide hemipterans against predators and parasitoids [4,5]. In return, the ants receive large amounts of honeydew, which is essential to their colony growth and survival because it contains sugars mixed with various amino acids [2,[6][7][8][9]. Previous studies have shown that the tending capability of ants improves the population growth of hemipterans by reducing not only predation and parasitism from natural enemies but also the risk of fungal infection [7,[10][11][12][13][14]. Fire ants are able to protect honeydewproducing hemipterans with their ability to deter natural enemies. The red imported fire ant Solenopsis invicta from the southeastern region of the United States has been reported to construct shelters for honeydew-producing hemipterans [2]. However, whether S. invicta protects hemipteran insects using shelters constructed by other species remains to be further explored.
S. invicta is a dangerous pest with a worldwide distribution, such as in the United States, Australia, New Zealand, and China [15,16]. The negative effects of S. invicta on agriculture and forestry production, human health, and poultry production in South China have already been reported [17,18]. Phenacoccus solenopsis Tinsley is native to the United States and has spread throughout the world, including Central America, South America, and Africa. It was reported to have invaded China and has shown potential damage to cotton [19,20]. Mutual interactions mediated by honeydew occur commonly between hemipterans and invasive ants [1][2][3]. P. solenopsis produces large amounts of honeydew being consumed by fire ants [10], indicating a conditional mutualism between these two invasive species mediated by the sugar source [21]. Sylepta derogata (Fabricius) is a phytophagous lepidopteran that feeds on many host plants, such as Hibiscus rosa-sinensis, Hibiscus syriacus, Gossypium hirsutum, Solanum melongena, and Vigna unguiculata [22]. These pests are greatly populated and widely distributed in China, with H. rosa-sinensis and G. hirsutum being their favorite host plants [23]. We found that H. rosa-sinensis leaves were curled up by leaf roller larvae and served as pupation sites as well as that the curled leaves were probably utilized as shelters for mealybugs. Therefore, we hypothesized that fire ants protect mealybugs against their enemies using the leaf shelters constructed by S. derogata. We also verified whether the tending capability of ants and the shelters constructed by leaf rollers could enhance the survival of mealybugs on enemy-infected and uninfected plants. Our results provide new insights about the mutualism between invasive ants and honeydew-producing hemipterans and may contribute to improving this interaction for the successful invasion of these species.

Host Plants
H. rosa-sinensis was purchased from a commercial horticultural farm. All plants measured approximately 25-30 cm in height and were cultivated in plastic flower pots.

Insects
All insects were collected from the suburbs of Guangzhou. P. solenopsis were fed on H. rosa-sinensis and maintained in the laboratory at 2762uC and a relative humidity of 60%-70%. Each colony of S. invicta was divided into several subcolonies (,1,000 workers and one queen) measured using a microbalance (Sartorius, BS, 224S). The ants were placed in a 9-cm plastic Petri dish with moistened plaster, which served as their artificial nest. They were given fresh, live Tenebrio molitor larvae and a 10% honey solution mixed with water (50 ml) weekly.
The lady beetle Menochilus sexmaculata and parasitoid Aenasius bambawalei are common enemy species of P. solenopsis in South China [24]. For this study, they were collected from H. rosa-sinensis plants in the field. M. sexmaculata were fed with mealybug nymphs in the laboratory, whereas A. bambawalei were initially collected as mummies (parasitoid pupae encased in dead aphids) and subsequently separated into gel capsules (,10 mm in length) until adults emerged. Afterward, the wasps were classified according to sex and randomly paired. Copulation was observed in all pairings, and female wasps were used in experiments 24 h after the initial pairing.

Field Investigation
This investigation was conducted in the experimental fields of South China Agricultural University. After ant colonies (,50,000 workers and five queens) were transferred to the plot (2 m66 m) edge and established, we planted four H. rosa-sinensis plants in a single file at distances of 0.8, 1.6, 2.4, and 3.2 m, respectively. Two leaf roller larvae were introduced to each H. rosa-sinensis plant, and they subsequently formed shelters. Mealybugs (120 per tube) were placed in a small plastic tube (1 cm64 cm) and fixed on the epicormic branch of each H. rosa-sinensis plant. The mealybug larvae would climb out of the tube onto the leaves. These experiments were conducted using (1) mealybugs alone, (2) mealybugs with leaf rollers, (3) mealybugs with tending ants and leaf rollers, and (4) mealybugs with tending ants alone. A single H. rosa-sinensis was planted on an ant-free plot. [Plots without fire ant infestation were chosen to exclude colonies of S. invicta.] After 2 weeks, we counted and recorded the number of foraging ants on the plants, the numbers of mealybugs inside and outside the shelters, and the total number of mealybugs present on the plants. Each treatment was replicated eight times.
We conducted our field studies in areas where fire ants and mealybugs occur and for which no specific permits were required. The plots used were covered sparsely with weeds and seldom disturbed artificially before the experiments. The land used as the study area is neither privately owned nor protected in any way, and the field studies did not involve endangered or protected species.

Laboratory Experiments
Two leaf roller larvae were transferred to H. rosa-sinensis leaves without mealybugs. When the leaf roller pupated, second instar mealybugs were transferred to the H. rosa-sinensis leaves (100 per plant). After 24 h, two lady beetle larvae and two parasitoid adults were separately placed on the plants. All plants were caged in wooden frames (60 cm660 cm640 cm) covered with nylon mesh. After another 24 h, the ant subcolonies were connected to their assigned plants through a plastic tube (1.5 cm in diameter). These experiments were conducted using (1) mealybugs alone, (2) mealybugs with leaf rollers, (3) mealybugs with predators and leaf rollers, (4) mealybugs with predators, (5) mealybugs with parasitoids and leaf rollers, and (6) mealybugs with parasitoids. We counted and recorded the numbers of surviving mealybugs on enemy-removed plants, on predator-infected plants, and in shelters after 1 week, whereas those of surviving mealybugs on parasitoidpresent plants and in shelters were counted after 2 weeks. Each treatment without ant tending served as the control. All treatments were replicated 10 times.

Statistical Analysis
After arcsine transformation, all data were tested for normal distribution using the Shapiro-Wilk test to compare differences in the survival of P. solenopsis between leaf roller-infected and uninfected plants, between predator-or parasitoid-infected and uninfected plants, and between plants with and those without ant tending. If the data were normally distributed, the t test was performed to compare means among all measured variables. The non-parametric Kruskal-Wallis test was used to compare the numbers of foraging ants present on plants placed at different distances from ant colonies. Linear regression was used to analyze the correlation between the number of surviving mealybugs in shelters and the number of foraging ants on plants. All statistical methods were conducted using SPSS 14.0.

Effects of Ant Tending and the Shelters on the Survival of Mealybugs in the Field
We found that ant workers occurred in the shelters and gradually transported or drove mealybugs toward them when their natural enemies were present. Fire ants increased the survival of mealybugs, although predators and parasitoids could enter the shelters. In the field, no significant difference between surviving mealybugs inside the shelters and those surviving outside was detected when the plants were close to an ant colony (0.8 m: t = 22.294, df = 7, p = 0.056; 1.6 m: t = 22.141, df = 7, p = 0.069) (Fig. 1A). However, a significant difference was observed when the plants were distant from an ant colony (2.4 m: t = 23.330, df = 7, p = 0.013; 3.2 m: t = 22.620, df = 7, p = 0.034; .5 m: t = 23.582, df = 7, p = 0.009) (Fig. 1A). When the plants were only 0.8 and 1.6 m away from the ants' nest, the total number of surviving mealybugs in sheltered plants was significantly higher than that in unsheltered ones (t = 2.964, df = 14, p = 0.010; t = 2.169, df = 14, p = 0.048). However, no significant difference was observed when the plants were 2.4 and 3.2 m away from the ants' nest (t = 1.218, df = 14, p = 0.243; t = 1.268, df = 14, p = 0.226) (Fig. 1B). In addition, fewer ants were found when plants were placed at a greater distance from ant colonies (Kruskal-Wallis test: x 2 = 16.796, df = 3, p = 0.001) (Fig. 2A). The number of surviving mealybugs in shelters was significantly linearly (y = 0.089x+5.847; F = 4.647, p = 0.039) correlated with the number of foraging ants on a given plant (Fig. 2B).

Effects of Ant Tending and the Shelters on the Survival of Mealybugs on Enemy-infested Plants
When lady beetle larvae were present and the ant colony had access to the plants, the survival of mealybugs on plants with leaf rollers was remarkably greater than that of mealybugs on plants without leaf rollers (t = 2.765, df = 9, p = 0.022) (Fig. 3A). However, no significant difference was found between plants with and those without leaf rollers when the ant colonies were separated from them (t = 1.090, df = 9, p = 0.304) (Fig. 3A).
When parasitic wasps were present, no apparent difference was observed in the survival of mealybugs between plants with and those without leaf rollers, regardless of whether or not the ant colonies had access to the plants (t = 1.443, df = 9, p = 0.183; t = 1.620, df = 9, p = 0.140) (Fig. 3B).

Effects of Ant Tending and the Shelters on the Survival of Mealybugs on Enemy-excluded Plants
When natural enemies were removed from the plants, no significant difference was found in the survival of mealybugs between plants with and those without leaf rollers when the ant colonies had access to the plants (t = 20.663, df = 9, p = 0.524) (Fig. 3C). The situation in plants without an ant colony was similar to that in plants with one (t = 21.042, df = 9, p = 0.325) (Fig. 3C).

Effects of Ant Tending on the Survival of Mealybugs in Shelters on Enemy-infested Plants
When lady beetle larvae were present on the plants, the survival of mealybugs in shelters with an ant colony was significantly greater than that of mealybugs on plants without an ant colony (t = 2.782, df = 9, p = 0.021) (Fig. 3D). When parasitic wasps were present on the plants, a significant difference in the survival of mealybugs was found between plants with and those without an ant colony (t = 2.295, df = 9, p = 0.047) (Fig. 3D).

Discussion
Previous studies have indicated that S. invicta constructs shelters for hemipterans to avoid parasitic enemies [2]. Our study is the first to report that fire ants used the shelters constructed by leaf rollers to protect their mealybug mutualists. We verified this interesting interaction in the field and under laboratory conditions. We also found that mealybugs had difficulty entering the shelters successfully without the assistance of fire ants, demonstrating that the co-occurrence of ant tending and the shelters constructed by leaf rollers improved their survival in shelters on predator-infested plants. Shelters also enabled ants to monopolize the honeydew produced by the hemipterans. Our results indicate that the mutualism between fire ants and mealybugs facilitates the fitness of both species through their interaction [7,13].
We have confirmed that mealybugs benefit from the abovedescribed interaction because their survival in shelters was significantly higher than that of mealybugs on plants with an ant colony (Fig. 3D). The results demonstrated that the ants transferred mealybugs into the shelters produced by leaf rollers (Fig. 1), in agreement with research showing that S. invicta actively moves homopterans between plants to avoid predators and parasitoids [25]. In addition, our results indicate that ant tending and the shelters constructed by leaf rollers could not significantly improve the survival of mealybugs on parasitoid-infested plants (Fig. 3B).
We therefore concluded that ants still move mealybugs toward the shelters on plants with parasitoids but that parasitic wasps find them and lay eggs on their body. This conclusion was supported by the number of mealybug mummies determined in the shelters at the end of the experiments.
In summary, mealybugs benefit from their interaction with fire ants and S. invicta ants recognize the situation of mealybugs and respond accordingly. Although S. invicta is considered an omnivorous species and colonies of fire ants can survive without honeydew-producing insects, this study confirms that mealybug husbandry is beneficial in that ants may evolve and develop defenses against their natural enemies and could be an important mechanism for establishing and spreading invasions.