Introduction

The performance of sap-sucking insects such as mealybugs can be influenced by the content of the phloem sap of the host plants (Larsson 1989). Among other nutrients in the phloem sap, free amino acids are the most important source of protein for phloem feeders (Panizzi and Parra 1991). Thus, the amount of nitrogen in a plant influences the insect performance (Miles et al. 1982; Jansson et al. 1991; Jiao et al. 2012). The N content in the plant can vary as a function of N availability in the soil, where the plants are grown and augmented via addition by fertilizations, while its absorption is influenced by the balance of water availability in the plant and in the soil.

The use of fertilization, especially with N, is a common agronomic practice to maximize yield and quality of products such as the cotton fiber (Jadoskil et al. 2010). Nevertheless, incorrect application of fertilizers, including excessive dosages of N, may result in variable levels of free amino acids in the plant, which may in turn favor the occurrence of herbivores. Research data show that fertilization resulting in high concentration of N in the plant promote better performance of chewing insects such as the beet armyworm Spodoptera exigua (Hübner) (Lepidoptera: Noctuidae) on cotton plants (Chen et al. 2008) and the red postman butterfly Heliconius etaro phyllis (Fabr.) (Lepdoptera: Nymphalidae) on passion flower (Kerpel et al. 2006). Furthermore, N fertilization positively enhances the performance of sap-sucking insects such as the mealybugs Planococcus citri (Risso) (Hemiptera: Pseudococcidae) on cocoa and coleus plants (Fennah 1959; Hogendorp et al. 2006), Saccharicoccus sacchari (Cockerell) on sugarcane (Rae and Jones 1992), planthoppers Prokelisia dolus Wilson and P. marginata (Van Duzee) (Hemiptera: Delphacidae) on saltmarsh cordgrass (Huberty and Denno 2006), and aphids such as Aphis gossypii Glover (Hemiptera: Aphididae) on cotton (Barros et al. 2007), Rhopalosiphum padi (L.), and Sitobion avenae (Fabr.) on wheat (Aqueel and Leather 2011).

Besides correct fertilization, appropriate water supply plays a key role in plant nutrient uptake and in subsequent distribution of nutrients throughout the plant. Under dryland cotton farming, the plants can be subjected to water stress because of temporary periods of insufficient rain, especially in the semiarid areas. Further, according to Miles et al. (1982), plants under water stress can reduce moisture in the plant tissues, starch, and carbohydrates in the leaves, and mainly, soluble nitrogen and sugars. Furthermore, increased densities for some insect species have been reported on water-stressed plants. According to White (1969), outbreaks of the psyllid Cardiaspina densitexta Taylor on eucalyptus in Australia were due to a previous deficit of water, supporting the plant-stress hypothesis (PSH) proposed by the same author. Moreover, greater densities of the aphid Diuraphis noxia (Mordvilko) on wheat (Archer et al. 1995) and of the leafminer fly Phytomyza conyzae Hendel (Diptera: Agromyzidae) on cadowydd (Staley et al. 2006) have been reported for water-stressed plants. In addition, results with plants subjected to water stress have shown enhanced performance of the aphids Myzus persicae (Sulz.) and Brevicoryne brassicaae (L.) on oilseed rape (Miles et al. 1982) and the mealybug Phenacoccus herreni Cox & Williams on cassava (Calatayud et al. 2002).

Enhanced performance of some insects on water-stressed plants is related to improved quality of the plant to the herbivores, corroborating with the PSH proposed by White (1969). The enhanced performance of the herbivorous insects on water-stressed plants is related to improved nutritional quality of the host plant. However, it is overcome by other morphological and physiological changes, which also occur on plants under water stress. For instance, the reduction in turgor pressure and the amount of water content in the plant (Hsiao 1973; Inbar et al. 2001; Huberty and Denno 2004) increase in allelochemicals (Gershenzon 1984; Mattson and Haack 1987; Inbar et al. 2001) and reduction in plant development (Price and Clancy 1986; Price 1991). Nevertheless, according to Huberty and Denno (2004), plants submitted to water stress negatively affect the performance of sap-sucking insects, while the chewing insects are not affected. Therefore, this subject is an open avenue for research to better understand the response of insect pests to cultivated plants, and the outcome can be species specific.

Infestations of mealybugs including the cotton mealybug Phenacoccus solenopsis Tinsley and the striped mealybug Ferrisia virgata Cockerell (Hemiptera: Psdeudococcidae) on cotton plants have been reported for different areas of Brazil, which is the fourth largest cotton producer in the world (Bastos et al. 2007; Ferreira et al. 2009; Miranda et al. 2011). The striped mealybug was reported recently infesting cotton fields in the semiarid area of Pernambuco state (Torres et al. 2011) and exhibited high performance with nearly 412-fold population growth on cotton (Oliveira et al. 2014). This new arrival has worried growers and integrated pest management practitioners due to lack of information about the pest and certified practices to restrain the pest population growth either using chemical or biorational methods.

Most area of cotton cultivation in Brazil is rain fed, where areas utilized for seed production are irrigated during the dry season. In the semiarid region the rainfall is irregular, with plants often been submitted to water stress during their development, what also can take place in the Cerrado, the region with largest area of cotton in Brazil, for short periods of time during the summer rainy season. Further, cotton cultivated in these regions by small and large scale producers occasionally occurs under N fertilization for increased yield. Therefore, knowledge of how the striped mealybug respond to cotton plants subjected to water tress and nitrogen fertilization will provide understanding of the potential of this species as a cotton pest in Brazil, especially in the fields located in the semiarid areas. Thus, the objective of this work was to evaluate the development and reproduction of F. virgata fed cotton plants fertilized with nitrogen and subjected or not subjected to water stress.

Materials and methods

Experimental conditions

The studies were conducted in the Laboratory of Insect Ecology and in the Experimental Area of the Crop Protection Unit of the “Universidade Federal Rural de Pernambuco” (UFRPE), Recife, Pernambuco State. Cotton plants were cultivated in a greenhouse (6 m long × 2.5 m high × 2 m wide) under environmental conditions of temperature (average: 26.6 °C; range 19.7–37.2 °C), relative humidity (average: 69.0 %; range 34.6–93.8 %), and photophase (12:12 h) monitored by the DataLogger Hobo® (Onset Computer, Bourne, MA, USA), which was regulated to register the conditions at 30-min intervals.

Experimental design

Cotton plants (Gossypium hirsutum L.) of cultivar CNPA 7H were cultivated in pots of 10-L volume filled with a standard mixture of clay soil, sand, and humus (2:1:1) plus 10 g of phosphorus and 10 g of potassium fertilizers (Yara Brasil Fertilizantes S.A., Porto Alegre, RS, Brazil), while the N was subsequently applied as supplementary fertilization to vary its amount in the plant. Then, 56 plants aged 20 days (~5 expanded leaves) and exhibiting roughly the same height were selected. These plants were divided into two groups: 28 plants were maintained only with the fertilization at planting date, whereas the other 28 plants received fertilization at planting date and subsequent N fertilizations. Nitrogen fertilization was applied 20, 30, 40, 50, 60, and 70 days after plant emergence. The subsequent N fertilizations after plant emergence were performed by applying 40 mL/pot of the dilution made with 10 g/L of ammonium sulfate [(NH4)2SO4; 20 % N]. Thus, we considered the 28 cotton plants cultivated without supplementary nitrogen fertilization (−N) and the 28 cotton plants cultivated with six supplementary nitrogen fertilizations (+N).

At 30 days after plant emergence, the plants belonging to the +N and −N treatments were subjected to two factors regarding the irrigation regimes. Two water regime supplies were determined based on previous measurements taken under the plant cultivation condition. The two water regimes were chosen to obtain plants that were visually with permanent turgid leaves, as well as plants that were with partial wilting leaves. In plants cultivated under regular irrigation, the soil water potential averaged 0.5 PSI, while, in under water stress, had soil water potential of 0.01 PSI; these soil water potentials were determined using analogical tensiometers (SondaTerra® Equipamentos Agronômicos, Piracicaba, SP). Water was supplied via a drench over the soil near the plant base, daily between 8 a.m. and 10 a.m. Therefore, after setting the irrigation regimes, four treatments were set up, combining N fertilization (+N and −N) under two regimes of irrigation [plants with (−W) and without water (+W) stress]. Thus, 14 plants of the +N treatment and 14 plants of the −N treatment were irrigated daily with 600–800 mL of water per plant (hereafter, “plants without water stress”), and the other 14 plants of the same treatments were irrigated with 100–300 mL/plant (hereafter, “water-stressed plants”).

Performance measures of F. virgata-fed cotton plants under variable nitrogen fertilization and water supply

Cotton plants (aged 40 days) maintained under the four conditions [+N and under water stress (−W), +N and without water stress (+W), −N and −W plants, and −N and +W plants] were infested with newly hatched F. virgata nymphs (aged <12 h). These nymphs were collected from the stock colony of F. virgata reared on pumpkin Cucurbita moschata (Duch.) Duch. Ex Poir cultivar Jacarezinho. To collect the nymphs from infested pumpkins, leaf discs of 4.8-cm diameter were obtained from cotton leaves harvested from the median section of plants of each treatment. These leaf discs were laid over a pumpkin infested with adult females of F. virgata, allowing the neonate nymphs (i.e., crawlers) to move to the discs. After 15 min, the leaf discs were removed from the pumpkins and placed on Petri dishes (9-cm diameter) and closed with polyvinyl chloride (PVC) plastic film (Alpfilm, São Paulo, SP, Brazil). The number of neonate nymphs was calculated as 150 nymphs/leaf disc by adding or removing neonate nymphs, if necessary, using a fine paint brush (No. 0) under a stereomicroscope (Opton™–NTB 3A). The transfer of infested-leaf discs to the cotton plant in the greenhouse was performed by attaching the leaf discs with an iron clip to the upper expanded leaves, being one leaf disc per leaf and two leaf discs per plant. The petiole of the respective infested leaf received entomological glue (BioStop Cola®, Biocontrole S.A., São Paulo, SP, Brazil) to avoid nymphs escaping from the releasing site. In addition, the main stem of the plant was smeared with entomological glue to prevent other insects from climbing on cotton plants.

The experimental design consisted of 2 × 2 factorial with two treatments (+N and −N) under two irrigation regimes (+W and −W) with 14 replications each. The data for survival of mealybugs, duration of development + pre-oviposition periods, and the number of offspring produced were averaged per plant considering the results obtained from two leaves as subsamples per plant. The percentage of survival consisted of the number of mealybugs found per leaf 25 days after infestation as a function of the initial release density of 150 neonate nymphs/leaf. After this evaluation, two adult females were transferred to upper leaves on the same plant. The petiole of the newly infested leaves with adult females was smeared with entomological glue to prevent females from escaping. For these females, we recorded the age of initial offspring production and the number of offspring produced during 16 consecutive days of evaluation. According to Oliveira et al. (2013), the striped mealybug under the studied conditions molted to adult within approximately 25 days after infestation and exhibited reproductive period of about 16 days. Furthermore, a period longer than 16 days could also affect the results due to cotton leaf senescence. Thus, the average of development and pre-oviposition periods, adult sex ratio, and the number of offspring produced during 16 days were recorded for females confined to the plants.

The offspring sex ratio of F. virgata female raised feeding cotton plants under N fertilization and irrigation regimes was investigated. To do so, 128 s-instar nymphs were randomly collected from eight plants (16 nymphs per female) from each treatment. These nymphs were transferred to cotton leaf discs harvested from the same plant and placed into Petri dishes at rate of 25 nymphs per disc and considered as one replications. The nymphs were kept under 26 °C in the laboratory, which was the average temperature in the greenhouse where they were previously raised. The sex ratio of nymphs was determined when male nymphs molted to the third-instar nymph. Male nymphs exhibit defined head, slender body, and denser wax layer over the body as compared to the females in the third-instar.

The role of the N fertilization and irrigation regimes on the development of cotton plants and on the nutrient contents in the leaves of these plant was also evaluated. Five days after applying the fourth N fertilization (plant age: 55 days) leaves from the median section of the plants (determined by the number of nodes in the main stem) were harvested from four plants of each treatment (three leaves/plant). The collected leaves (n = 12 leaves per treatment) were placed in plastic bags, identified, and conducted to the laboratory to run foliar analysis regarding the nutrient contents (LabFert—Laboratório de Análises Agrícolas Ltda, Recife, PE, Brazil). The analysis was performed using the methodology of Orlando Filho and Zambello Jr (1983). In addition, the development of the plants 123 days old and subjected to the treatments (+N and −N) and treatment factors (+W and −W) was evaluated using ten plants (replication) per N fertilization and irrigation regime. The evaluation step consisted of determining the fresh weight of the whole plant, dry weight, and the number of bolls produced per plant. To obtain the dry weight, the plants were sliced into small parts, placed into paper bags, and then dried using a forced-air oven regulated at 54 °C for 72 h.

Data analyses

The results on survival, developmental times, offspring sex ratio, reproduction of mealybugs, and development response of plants submitted to the treatments were tested for normality (Kolmogorov–Smirnov’s test) and homogeneity (Bartlett’s test) of variance using the Proc Univariate and Proc analysis of variance (ANOVA) of SAS, respectively (SAS Institute 2001) and transformed, if necessary. Untransformed results are presented in tables and figures. Furthermore, the percentage of mealybug survival determined at 25 days after infestation, the development + pre-oviposition duration, the number of offspring produced, sex ratio, the fresh and dry weight of plants, and the number of bolls produced per plants were subjected to two-way ANOVA (N fertilization and irrigation regimes) using PROC ANOVA of SAS (SAS Institute 2001). The mean comparisons were performed using Fisher’s test (df = 1) of ANOVA with 0.05 significance levels. The data of leaf nutrients were transformed into percentage for each nutrient evaluated as a function of nutrient total and compared using pair-wise chi-square test after the null hypothesis of equality was rejected at 0.05 significance levels by PROC FREQ of SAS (SAS Institute 2001).

Results

The leaves of cotton plants submitted to N fertilization and irrigation regimes showed variation in the nutrient content (Table 1). A significant increase of the N content was observed in the leaves of plants receiving successive N fertilizations and water stress over those not receiving N supplements regardless of water levels. On the other hand, +N and +W plants and −N and −W plants exhibited similar amounts of nitrogen content in the leaves (χ 2 = 0.11, P = 0.7360). Leaves of +N and −W plants exhibited twofold greater N content than leaves of −N and −W plants (χ 2 = 7.85, P = 0.0051).

Table 1 Leaf tissue nutrients from cotton leaves harvested from median third of the cotton plant submitted to nitrogen fertilization and water stress conditions
Full size table

Plant development and boll production varied as function of N fertilization and irrigation regimes (Table 2). −W plants, irrespective of N fertilization (both +N and −N), exhibited greater weights (fresh and dry) and number of bolls (Table 2). The overall results show that plant weight and boll production were twofold and fourfold greater, respectively, for plants under N treatment without water stress as compared to plants with water stress (Table 2).

Table 2 Fresh and dry weight, and number of bolls of cotton plants in response to N fertilization levels and irrigation regimes in the greenhouse (cotton plants were at 123 days after emergence) with N fertilization initiated for 20-day-old plants (six N fertilizations were made, spaced 10 days apart), without water (+W) and with (−W) water stress for 30-day-old plants
Full size table

The survival of mealybug nymphs evaluated 25 days after infestation, the development + pre-oviposition duration, and the sex ratio of the offspring produced were similar across the treatments studied (+N and −N, +W and −W) (Table 3). The overall mean percentage (±SE) across all treatments for nymphal survival evaluated 25 days after infestation was 38.0 ± 1.60 %. Also, the developmental period from infestation to offspring production was, on an average, 47.3 ± 0.65 days, and the mean of offspring sex ratio was 84 ± 3.0 % females. On the other hand, the offspring production of F. virgata was variable as a function of N fertilization and water-stress condition (Table 3). Despite the significance for both treatment effects (N fertilization and irrigation regimes), there was no interaction effect between these factors because the same pattern of greater offspring production was observed in water-stressed plants irrespective of N fertilization (both +N and −N) (Fig. 1).

Table 3 Results of two-way analysis of variance (ANOVA) for mealybug survival 25 days after infestation, development, and adult traits regarding N fertilization and irrigation regimes of cotton plants
Full size table
Fig. 1
figure 1

The mean number of offspring produced per female of F. virgata reared on cotton plants submitted to fertilization levels (+N and −N) and irrigation regimes, characterizing plants under regular water supply and water stress. Note: Means (±SEM) under the same capital letters compare irrigation regimes within the same fertilization level (vertical); while means (±SEM) under the same small letters compare N fertilizations within the same irrigation regime (horizontal) (Fisher’s test of ANOVA at 0.05 levels)

Full size image

The average offspring production was 359.5 ± 25.41 and 225.3 ± 11.62 nymphs/female on −W and +W plants, respectively, resulting in 37 % more offspring produced by females reared on −W cotton plants as compared to +W plants irrespective of N fertilization. Regarding the condition of N fertilization, irrespective of the irrigation regime, mealybug females raised on cotton plants receiving supplementary N fertilization produced 320.3 ± 25.41 nymphs as compared to 264.5 ± 19.90 nymphs for females raised on plant without N supply. Thus, 18 % more offspring was produced by females fed cotton plants with increased N content.

Discussion

Leaf nutrient contents and plant developmental performance revealed significant effects of N fertilization and irrigation regimes studied. Cotton plants receiving successive N fertilizations and cultivated under water stress exhibited greater N content in leaves as compared to plants with fertilizations but not under water stress. These findings agree with other studies in that water stress (White 1969; Miles et al. 1982; Brodbeck and Strong 1987) and application of N fertilizers increase nitrogen content in plants and thereby increase the amount of N available to the herbivores (Mattson 1980; Bentz et al. 1995; Nevo and Coll 2001; Jiao et al. 2012). Furthermore, cotton plants without water stress showed better development than water-stressed plants, and plants receiving N fertilization developed more as compared to plants without N fertilizations. However, the water-stress condition was the major factor that contribute to the differences in the development of study plants (Table 2).

Nitrogen fertilization and water stress positively affected the offspring production of F. virgata with an additive effect. Plants subjected to water stress produced higher number of offspring and was even superior when N was applied. However, the studied condition of N fertilization and water stress did not interfere with the survival of the neonate nymphs and developmental times of F. virgata. Thus, in the light of our studied conditions, the results obtained indicate that successive N fertilization and water stress had only detectable effects on the reproductive output of F. virgata. This evidence suggests a straightforward conclusion that N-enriched cotton plants under water stress are more suitable for the mealybug offspring production; this is incompliance with previous data on better host plant quality to sap-sucking insects under both N fertilizations (Mattson 1980; White 1993; Jiao et al. 2012) and water stress (White 1969; Mattson and Haack 1987; Brodbeck and Strong 1987).

The results help us to explain the frequent outbreaks of mealybugs in cotton fields of the semiarid region of Brazil (Bastos et al. 2007; Torres et al. 2011; Silva-Torres et al. 2013), which is a common event at the end of cropping season when the rain becomes scarce. Also, these findings help us explicate the heavy infestations of mealybugs observed on cotton plants under water stress in the corners of cotton fields in the west of Bahia State, which is cultivated in the dry season and irrigated with pivot (J. B. Torres, personal observations).

Despite much information on the effects of water stress and N fertilization on sap-sucking insects, as discussed in the Introduction section, only few studies have evaluated the effect of these factors (N fertilizations: Lema and Mahungu 1983; Hogendorp et al. 2006; and plant water stress: Calatayud et al. 2002) on mealybug performance. The developmental time of the citrus mealybug P. citri was shorter when fed on coleus plants enriched with different concentrations of N (Hogendorp et al. 2006), while the mealybug P. herreni responded with faster development when fed on cassava plants under water stress (Calatayud et al. 2002). Furthermore, these two mealybug species (P. citri and P. herreni) exhibited high offspring production in the same conditions that partially corroborate with our data on F. virgata. It is interesting to note that the data are not standardized among the mealybugs, as the cassava mealybug Phenacoccus manihoti (Malite-Ferrero) (Hemiptera: Pseudococcidae) did not respond positively to cassava plants enriched with N fertilization (Lema and Mahungu 1983). Therefore, it seems that there are variations in response to N fertilizations and water stress among the mealybug species. The response can be of specific developmental or of reproductive trait, as for F. virgata in the present study.

In general, results of N fertilization show an improvement in herbivores performance (Mattson 1980; Rae and Jones 1992; White 1993; Hogendorp et al. 2006; Kerpel et al. 2006; Huberty and Denno 2006; Barros et al. 2007; Chen et al. 2008; Aqueel and Leather 2011). However, there are some divergences regarding the effect of plant water stress on herbivores. Huberty and Denno (2004) analyzed 82 studies and concluded that plants under water stress caused negative effect on sap-sucking insects, which contradicts the PSH supported by other authors (White 1969; Mattson and Haack 1987; Brodbeck and Strong 1987). According to Huberty and Denno (2004), under continuous water stress, the turgor pressure of the plant is negative and the insects are impeded to obtain nutrients from the plant sap. On the other hand, when the plants are subjected to discontinuous water stress, they become more suitable with enriched sap available to sap-sucking insects as the plant turgor pressure turn positive again under regular water supply. Thus, sap-sucking insects can exhibit better performance on plants subjected to intermittent water stress such as those short droughts during the summer rainy season. In our study, the cotton plants were continuously exposed to water stress during the study period duration of 30–123 days, and F. virgata performed reproductively better. The striped mealybug is a highly polyphagous and cosmopolitan species (Normark and Johnson 2011) and thus potentially adapted to various species of host plants and climatic conditions.

The survival success and the sex ratio of F. virgata offspring produced of F. virgata were not affected by the condition of cotton plants studied. About 38 % of the released neonate nymphs reached adult stage across all treatments. The loss of nymphs is especially during the crawler stage, when dispersing over the plant, and could account for the remaining nymphs not reaching adult stage. According to Strickland (1950), Grasswitz and James (2008), and Silva-Torres et al. (2013), several factors can determine the success of F. virgata survival on cotton plants including wind and rain and plant morphology. Regarding the sex ratio, in most mealybugs, it is determined by the parental female (Ross et al. 2010). According to Ross et al. (2011), the sex ratio of P. citri is variable as a function of food shortage. This result indicates that variations in the amount of food intake by the female influences the offspring sex ratio. Thus, we hypothesized that the variable conditions of the cotton plant (N fertilizations and irrigation regimes) may have altered the proportion of females and males in the offspring produced by females of F. virgata raised on improved host characteristics through fertilization and water stress. However, our results do not support this hypothesis and require more investigation to ascertain regarding the sex ratio of the offspring produced by F. virgata, because, beyond host quality and host species, other factors such as variation in temperature, population densities, and the age of the female at first mating are important, as observed for other mealybug species (Nelson-Rees 1960; Varndell and Godfray 1996; Ross et al. 2010, 2011). The knowledge of offspring sex ratio adjustment may help adopt integrated pest management strategies such as survey and mass capture of males using male-attracting semiochemicals; meanwhile, this practice become useless for conditions favoring only female offspring production.

The present study aimed to obtain information about the performance of striped mealybug F. virgata infesting cotton plants with respect to N fertilization and water stress, considering that the latter is often observed in arid areas and may also occur during rainy summer season due to short periods of drought. Among our findings, we can highlight that the offspring production of F. virgata was nearly twofold higher on water-stressed plants with successive N fertilizations. The offspring production was increased by 37 % as a function of water stress and 18 % as a function of N fertilization, while developmental times and survival were similar across the studied conditions. Therefore, cotton plants either enriched with N fertilizations or water-stressed are suitable host for F. virgata infestation, especially when these conditions occur simultaneously. To counteract these favorable conditions to the pest, attention should be paid in risk areas caused by prolonged drought periods. Therefore, to avoid generalized F. virgata infestation, fertilization of cotton plants should be judiciously managed, including managing of irrigation as well, whenever possible, to avoid successful offspring production and survival of the pest.