The effect of 6 host plant species [Mangifera indica L., Cajanus cajan (L.) Millspaugh, Coffea arabica L., Cucurbita moschata Duchesne, Parkinsonia aculeata L., and Ficus benjamina Roxb.], on bionomics of the mango mealybug Rastrococcus iceryoides Green (Hemiptera: Pseudococcidae) was studied in the screenhouse. Biological and life table parameters of the mealybug differed significantly among the host plants. Developmental period (egg to adult) was shortest on M. indica (23.5 days and 25.3 days for females and males, respectively), whereas it was longest on F. benjamina (33.0 days and 37.3 days for females and males, respectively). The egg to adult female survivorship was highest on C. moschata (79.6%) and lowest on C. arabica (30.9%). Fecundity was highest on C. moschata (811.3 egg/female) followed by M. indica (716.8 egg/female). The sex ratio was female-biased on C. moschata, M. indica, C. cajan and P. aculeata. Adult mealybug longevity also varied with host plant for both mated and unmated females. Adult female body length and width were significantly higher on C. moschata, M. indica, C. cajan and P. aculeata than on F. benjamina and C. arabica. The highest intrinsic rate of natural increase (rm), finite rate of increase (λ) and the shortest mean generation time (GT) and doubling time (Td) were recorded on M. indica. The highest and lowest net reproductive rate (Ro) occurred on C. moschata and C. arabica, respectively. The implication of these findings in relation to damage, population growth and management of R. iceryoides on the target crops is discussed.
In Africa, Rastrococcus invadens Williams and Rastrococcus iceryoides Green are regarded as 2 important exotic mealybug species native to Southern Asia that commonly infest mango, Mangifera indica Linnaeus (Anacardiaceae). The former devastated mango production in West and Central Africa, but was brought under biological control through introduction of an exotic parasitoid Gyranusoidea tebygi Noyes from India (Noyes 1988; Bokonon-Ganta & Neuenschwander 1995). Based on its economic importance and the ease with which it colonized major parts of West and Central Africa, R. invadens has been the subject of many studies, both descriptive and experimental (Williams 1986; Agounké et al. 1988; Willink & Moore 1988; Bokonon-Ganta et al. 1995; Tobih et al. 2002). On the other hand, Rastrococcus iceryoides is restricted to East Africa (mainly Tanzania and coastal Kenya) and northern Malawi where it has remained a major pest of mango (Williams 1989; Luhanga & Gwinner 1993; CABI 2000).
In Southern Asia, R. iceryoides is believed to be highly polyphagous and has been reported from over 65 host plants from 35 families (Williams 1989; Ben-Dov 1994). In Kenya and Tanzania, recent observation showed that the insect attacks 29 host plants from 16 families with mango, Mangifera indica L. (Anacardiaceae), as the most preferred cultivated host plant and Parkinsonia aculeata L. (Fabaceae) as the most preferred wild host plant (Tanga 2012). As with other mealybug species, R. iceryoides sucks sap from leaves, young shoots, inflorescences and fruits and their damage can result in shedding of mango fruit-lets. They also excrete sugary honeydew on which sooty mold develops thus reducing fruit marketability. As a result of the development of sooty mould, export opportunities are often impaired due to quarantine regulations (CPC 2002). Sooty mold that fouls the leaves reduces photosynthetic efficiency and can cause leaf drop. In village homesteads, heavy infestations usually render the trees unsuitable for shade. In Kenya, Tanzania and Malawi, fruit losses can range from 30% to complete crop failure in unmanaged orchards (CABI 2000; Tanga 2012). In Tanzania, the pest has become a main target for majority of insecticidal sprays on mango (in addition to pruning and burning of infested plant parts) (Willink & Moore 1988; Tanga 2012). In addition to health concerns attributed to chemical pesticides, resource-limited farmers cannot afford to use them. Chemical pesticides also do not provide adequate control owing to the waxy coating of mealybugs. As a result, some growers have resorted to cutting down mango trees, because of heavy R. iceryoides infestations, while others have abandoned mango cultivation altogether. It has been speculated that the increasing intensity of damage by mealybugs may be due to the expansion of mango production and the introduction of hybrid cultivars, which are highly susceptible to attack by the pest (Boussienguet & Mouloungou 1993).
Like all other herbivorous arthropod pests, host plant range is a key ecological characteristic of the mealybug species, as it defines their resource base, which in turn is an important factor influencing their population dynamics and interactions with other herbivorous species, predators and parasites (Williams 1989; Ben-Dov 1994; Neuenschwander 2003; Calatayud & Le Rü 2006). Understanding insect-host plant interactions and their impact on development and various fitness parameters of an insect pest is a central theme in ecology (Miller et al. 1986; Benrays & Chapman 1994). Also in mealybugs, different host plant species have been shown to affect the insect's life history parameters. For example, the mortality of the citrus mealybug Planococcus citri (Risso) (Hemiptera: Pseudococcidae) was reported to be higher on green than red or yellow variegated Coleus blumei Bellevue (Bentham) (Lamiaceae) plants, and development was faster and fecundity higher on red variegated plants (Yang & Sadof 1995). The developmental time of Planococcus kraunhiae (Kuwana) (Hemiptera: Pseudococcidae) was shorter when reared on germinated faba bean, Vicia faba L. (Fabaceae) seeds than on leaves of a Citrus sp. and on Cucurbita maxima Duchesne, and it survived better when reared on germinated faba bean seeds than on citrus leaves (Narai & Murai 2002). The pink hibiscus mealybug, Maconellicoccus hirsutus (Green) (Hemiptera: Pseudococcidae), was able to develop equally well on Cucurbita pepo L. as on C. maxima (Serrano & Lapointe 2002).
The current R. iceryoides host plant data from Africa (Tanga 2012) are based on field observations of damage by the pest. Compared with R. invadens, there are no comparative data on the biology and demographic parameters of R. iceryoides on different host plants to determine the true value of each plant species as a host of R. iceryoides. Host plants that slow or accelerate the development of the insect are likely to be of considerable relevance to the development of management methods. Studies on the biology and life table parameters of R. iceryoides on different host plants should also provide information in understanding the population dynamics of this pest.
The main objective of this study was to investigate the development and reproduction of R. iceryoides on 6 host plant species, namely mango (M. indica), pigeon pea [Cajanus cajan (L.) Millspaugh, Fabaceae], arabica coffee (Coffea arabica L., Rubiaceae), butternut squash (Cucurbita moschata Duchesne, Cucurbitaceae), Jerusalem thorn (P. aculeata) and weeping fig (Ficus benjamina Roxb., Moraceae), in order to develop life table strictures for the insect and estimate parameters for population increase on the different host plants to guide pest management decisions. The host plants selected represent some of the most economically important plants in terms of horticulture, beverage, ornamental or shade uses in Kenya and Tanzania.
MATERIALS AND METHODS
Host Plant
Twelve-month-old M. indica and C. arabica seedlings were obtained from the commercial nurseries of the Kenya Agricultural Research Institute (KARI) and Coffee Research Foundation (CRF) in Ruiru, Kenya, respectively. Ficus benjamina plants grown from cuttings and P. aculeata of same age as M. indica and C. arabica were obtained from Tropical Nursery, in Nairobi and Malindi, Kenya, respectively. The production polythene bag of each seedling was removed and then the seedling was transplanted into a white plastic container (35 cm height × 29 cm top diam × 20 cm bottom diam) in a soil mix containing sieved sterilized forest soil and sand (1:1 by volume). Cajanus cajan were propagated from seeds (va. ICEAP00040) (Høgh-Jensen et al. 2007) obtained from KARI Seed Unit (KSU), Nairobi, Kenya. All the seedlings of the test plant species were placed on benches in a screenhouse (2 m height by 2.9 m diameter) at the Duduville campus of the International Centre of Insect Physiology and Ecology (icipe), Nairobi, Kenya. Conditions in the screenhouse were: 23 ± 5 °C, 40–80% RH and 12L: 12D photoperiod.
Plants were fertilized with equal volume of farmyard manure as described above, a common agronomic practice by the growers and watered on alternate days. Matured C. moschata fruits were purchased from a local grocery store in Kasarani, Nairobi, Kenya. Since R. iceryoides predominantly infested fruits of C. moschata in the field (C. M. Tanga, unpublished data), all experiments were conducted on fruits of this plant. Prior to commencement of the experiments, mealybugs were reared on mature C. moschata fruits that had been washed in a 0.5% bleach solution to reduce mold growth, triple rinsed with distilled water and air-dried for 24 h. Therefore, C. moschata served in this study as a check against other test plants on which experiments were carried out on seedlings. The fruits of C. moschata were kept in the laboratory maintained at room temperature (25–26 °C), photoperiod of 12 h L: 12 h D, and 40–70% relative humidity (RH).
Insect Culture
Colony of R. iceryoides was initiated from a cohort of 300 adult mealybugs collected from heavily infested mango orchard in Mombasa, Coast Province, Kenya in February 2008. Insects were transported to the laboratory at icipe, Nairobi, Kenya and reared on mature C. moschata fruits (purchased from a local grocery store) in the laboratory maintained at room temperature described above. The colony was maintained on an open table surface (76 cm wide × 245 cm length) in the laboratory for over 20 generations before the start of the experiment. Plywood was fixed firmly on the sides of the table (10.5 cm height × 245 cm length) to prevent the crawling insects from falling off. The colonies were maintained by exposing uninfested C. moschata fruits to adult females with fully developed ovisacs. Eggs hatched within 6-8 days and newly emerged nymphs were allowed to colonize the uninfested C. moschata. This procedure was repeated on a weekly basis. After every 6 months, fresh wild R. iceryoides from mango were injected into the established colonies to ensure broader genetic diversity.
Maintenance of R. iceryoides on the Study Plant Materials
For the bioassay, insects were reared on the different host plants for at least 3 generations in the screenhouse to allow them adapt to the new host and to remove maternal effects (Lacey 1998) before commencement of the experiments. Approximately 40 adult female mealybugs with well-formed ovisacs were obtained from the stock colony to infest each of the different host plants under investigation. The ovisacs were carefully teased open with blunt probes under a stereomicroscope and the number of eggs present in each ovisac counted. The eggs were then refolded into the fine cottony ovisac before inoculation. After the first generation on these host plants, subsequent uninfested plants were similarly infested with ovisacs from their respective cultures. In the screenhouse, the plants were maintained in large cages (30 cm length × 30 cm width × 60 cm height) consisting of a glass top and screened mesh (30 cm length × 30 cm width × 60 cm height) on the sides. Experimental conditions in the screenhouse were as described above.
Assessment of R. iceryoides Development, Survivorship and Sex Ratio on the Different Host Plants
Thirty eggs (collected within 12 h) were obtained from a single female ovisac arising from the different host plant species and transferred to the seedlings of their respective host plants using a camel hair brush. The seedlings of M. indica, C. arabica, P. aculeata and F. benjamina were 12 months old at the commencement of the experiments while C. cajan was 3 months. In the case of C. moschata, the insects were maintained on matured fruits similar to the rearing conditions. After inoculation, each host plant seedling and fruit was housed individually in wooden cages (30 cm length × 30 cm width × 60 cm height).
Host plants were checked twice daily for egg hatch and exuviae to identify emergence of nymphal instars. The sex of each individual mealybug was determined during the latter part of the second instar when the males finally shedoff the white mealy-covering on their body and change their color from orange to pale yellow with light ashy deposit on their body. Development of the males at this point continued with their body completely devoid of lateral processes and the duration of development of each sex could be recorded separately. The following data were collected for each host plant: (1) developmental duration for each stage, (2) the number of insects reaching adult stage for the sexes, (3) sex ratio, measured as proportion of females out of the total number of R. iceryoides [♀/(♀ + ♂)], and (4) percentage survival of each of the immature stages. The experiment was replicated 5 times.
Morphometry
Fifty randomly selected adult females from the plant species under investigation were slidemounted using the methodology of Watson and Kubiriba (2005). The body length (in millimeters) was measured along each insect's dorsal midline from the vertex of the head to the tip of the abdomen. The width (in millimeters) was measured at the widest point across the middle surface of the insect. Images from the slide-mounted specimens were captured using video microscopy — [Leica MZ 125 Microscope (Leica Microsystems Switzerland Limited)], fitted with Toshiba 3CCD camera using the Auto Montage software (Syncroscopy, Synoptics group, Cambridge, UK) at magnification of X25. Measurements were taken using Image-ProÒ Plus version 4.1 for Windows™ (Media Cybernetics, Bethesda, MD, USA) and the data were exported directly to an Excel data sheet. For all parts, measurements were taken in triplicate (to an accuracy of 0.001 mm).
Reproduction, Longevity and Assessment of Demographic Parameters
Forty randomly selected newly moulted virgin adult females (24 h old) derived from nymphs reared on each host plant species were used to determine the effect of host plant on R. iceryoides reproduction and longevity. Within each host plant treatment, half of the mealybugs (i.e. 20 females) were held alone to assess asexual re- production (unmated females) and the other half (i.e. 20 females) were used to assess sexual reproduction (mated females). Each female used for sexual reproduction was transferred individually to plastic Petri dishes (5 cm in diameter and 1 cm height) with a wet cotton ball at the side together with 3 newly emerged males (24 h old) from the same plant species, and allowed to mate for 24 h. After mating, females were transferred to their respective host plants and observed daily until they died. The total number of eggs produced by each female was recorded daily. The eggs were kept separately in transparent polyvinyl chloride (PVC) cylinder (4 cm diameter × 10 cm height × 0.21 mm thick) lined with pieces of moistened black filter paper (3.5 by 1.5 cm) to prevent desiccation, and egg hatch was determined every 12 h for a period of 7 days. Emerging nymphs from each daily cohort of eggs were removed using a camel hair brush (#000) with the help of a magnifier hand lens (size: 100 mm in diameter). Females for asexual reproduction were also observed daily until they died.
The following data were collected for each host plant: (1) pre-oviposition, oviposition, and post-oviposition periods, (2) daily egg production, and (3) adult longevity. Standard life table parameters including age-specific fertility (m x mean number of female progeny per female per day) and female survivorship (lx; the fraction of females surviving to age x) were calculated from daily records of mortality and fecundity of cohorts on each host plant. For each of the 2 reproductive stages (sexual and asexual), each female was considered a replicate.
Statistical Analysis
Data for developmental times, pre-oviposition, oviposition, post-oviposition periods, adult female longevity, egg production and size of R. iceryoides were subjected to one-way analysis of variance (ANOVA) using PROC GLM (SAS Institute 2010). Sex ratio and percentage survival of R. iceryoides were arcsine transformed before analysis of variance (Sokal & Rohlf 1981). Means were separated by Student-Newman-Keuls (SNK) test. T-test (α = 0.05) was used to compare the developmental time (egg-adult) and longevity of different sexes reared on the same host plant. Life table for each host plant species was constructed following the method described by Carey (1993) and the intrinsic rate of increase (rm), net reproductive rate (Ro), mean generation time (GT), doubling time (Td) and finite rate of increase (λ)) were estimated using the Jackknife program (Maia et al. 2000). Differences between life table parameters across the different host plant species based on estimates of variance for each parameter value were separated using SNK (Meyer et al. 1986).
RESULTS
Developmental Time, Percentage Female Progeny and Survival of Immature Stages
The overall developmental duration from egg to adult as well as that of the various developmental stages of both sexes of R. iceryoides varied significantly across the host plants (Table 1). Eggs took comparably longer period to hatch on all host plants, being relatively shorter on M. indica (7.8 days), and significantly different from those reared on F. benjamina (8.7 days).
The development of the first instar nymphs ranged from 5.5 days (on M. indica) to 9.6 days (on F. benjamina). Male second instar development was shortest on M. indica, C. moschata, C. cajan and P. aculeata (4.8–5.0 days) and longest on F. benjamina (11.4 days) (Table 1). The longest developmental duration for female second instars on F. benjamina was 8.1 days. The developmental time of the third instar male ranged from 3.4 days on C. moschata to 9.5 days on C. arabica, while that of the female ranged from 4.8 days on C. cajan to 6.7 days on both F. benjamina and C. arabica (Table 1). The developmental time of fourth instar males was significantly shortest on M. indica (4.7 days) and longest on F. benjamina (8.7 days). On the same host plant, males took significantly longer time to complete development to adult than females on all host plant except on C. moschata where the development of both sexes was similar.
Sex ratio was female biased (0.56–0.64) on M. indica, P. aculeata, C. cajan and C. moschata; and ranged from 0.44–0.49 on F. benjamina and C. arabica (F = 12.13; df = 5, 24; P = 0.0017) (Fig. 1).
The rearing host plant had a strong influence on both overall survival as well as the various developmental stages of R. iceryoides survival (Table 2). Egg survival was highest on M. indica, C. moschata and F. benjamina (85–90%) compared with the other host plants while more of the first instar survived on C. moschata, P. aculeata and C. cajan (80–84%) (Table 2). Survival of second male instar was highest on C. cajan (90%), while it was highest for females on C. moschata, P. aculeata and C. cajan (86, 86 and 87%, respectively) (Table 2). Third instar males survived better on C. moschata, P. aculeata and C. cajan (91, 90, and 91%, respectively) while third female and fourth instar male survival was highest on C. moschata (90 and 94%, respectively) (Table 2).
Morphometry
The body size of female R. iceryoides was significantly influenced by host plant species (Length: F = 328.5; df = 5,54; P = 0.0001; width: F = 218.4; df = 5,54; P = 0.0001) (Fig. 2). Adult female R. iceryoides reared on C. moschata and M. indica were significantly larger (3.93 mm and 3.87 mm, respectively) in body length than females reared on the other host plants while those reared on C. arabica had body length of only 2.24 mm. Adult female body width was significantly higher among R. iceryoides reared on M. indica, P. aculeata, C. cajan and C. moschata (2.63–2.68 mm) compared with those reared on F. benjamina and C. arabica (2.52–1.80 mm) (Fig. 2).
TABLE 1.
MEAN NUMBER OF DAYS (± SE) FOR EACH DEVELOPMENT STAGE OF RASTROCOCCUS ICERYOIDES REARDE ON SIX HOST PLANT SPECIES (CUCURBITA MOSCHATA, MANGIFERA INDICA, PAKINSONIA ACUTEATA, CAJANUS CAJAN, FICUS BENJAMINA, AND COFFEA ARABCA).
Reproduction and Longevity
The mean pre-oviposition, oviposition and post-oviposition periods of R. iceryoides were significantly affected by host plant species (Table 3). The pre-oviposition period was significantly shortest on M. indica (20.4 days) and longest on F. benjamina (29.4 days). Oviposition period was significantly longest on C. moschata (36.8 days) and shortest on C. arabica (15 days) (Table 3). Post-oviposition period of R. iceryoides was longest on C. arabica (15.7 days) and shortest on C. moschata and P. aculeata (6.4 and 7.2 days, respectively) (Table 3).
The females reared on C. moschata laid the highest number of eggs (811.3) compared to females reared on other host plants (Table 3). There was no significant difference in the number of eggs laid by R. iceryoides when reared on P. aculeata and C. cajan. Daily egg production was highest on M. indica (46.6 eggs) although this did not differ significantly from egg production on P. aculeata, C. cajan and C. moschata (37.7–39.3 eggs/female/ day) (Table 3). Unmated adult female mealybugs did not lay eggs on any of the 6 host plant species tested.
TABLE 2.
MEAN (± SE) PERCENTAGE SURVIVAL OF THE DIFFERENT DEVELOPMENTAL STAGES OF RASTROCOCCUS ICERYOIDES REARED ON SIS HOST PLANT SPECIES (CUCURBITA MOSCHATA, MANGIFERA INDICE, PARKINSONIA ACULEATA, CAJANUS, FICUS BBENJAMINA, AND COFFEA ARABICA).
Fig. 2.
Length and width (millimetres ± SE) of adult female Rastrococcus iceryoides reared on 6 host plant species. For each parameter, bars sharing the letter do not differ significantly by Student-Newman-Keuls (SNK) test (P = 0.05).
Longevity of mated as well as unmated R. iceryoides females varied significantly with the host plant (Table 3). Longevity of mated female mealybugs ranged from 56.0 days on C. arabica to 67.4 days on C. moschata. No difference in mated adult female longevities were observed when R. iceryoides was reared on P. aculeata (59.3 days), C. cajan (57.8 days) and F. benjamina (58.9 days) (Table 3). Longevity of unmated adult females ranged from 70.8 days on C. arabica to 90.5 days on C. moschata (Table 3). On the same host plant unmated females lived significantly longer than their corresponding mated ones on all host plants tested (Table 3).
Age-Specific Fecundity and Survivorship
The curves of age-specific fecundity (mx) peaked soon after the onset of reproduction and varied considerably among the different host plant species (Fig. 3). The age-specific fecundity for R. iceryoides reared on M. indica peaked on day 24, P. aculeata on day 25, F. benjamina on day 33 and C. arabica on day 29 (Fig. 3). Agespecific fecundity observed for R. iceryoides reared on C. moschata and C. cajan were remarkably different, each having 2 peaks (Fig. 3). Major peaks for R. iceryoides reared on C. moschata were on day 30 and 33 while on C. cajan, a major peak was recorded on day 27 and a minor peak on day 31. The age-specific survivorship (lx) curves decreased gradually and asymptotically as R. iceryoides aged (Fig. 3). On M. indica, 50% of mortality occurred on day 43, and the entire mealybug cohort died on day 62. On C. moschata, 50% of mortality occurred on day 48, and all mealybugs died on day 68.
TABLE 3.
MEAN DURATION (DAYS ± SE) OF PRE-OVIPOSITION, OVIPOSITION AND POST-OVIPOSITION PERIODS, REPRODUCTIVE RATE AND LONGEVITY OF RASTROCOCCUS ICERYOIDES REARED ON SIX HOST PLANT SPECIES (CUCURBITA MOSCHATA, MANGIFERA INDICA, PARKINSONIA ACULEATA, CAJANUS CAJAN, FICUS BENJAMINA, AND COFFEA ARABICA).
Population Growth Statistics
Host plant had a profound effect on all growth parameters [net reproductive rate (Ro), intrinsic rate of increase (rm), population doubling time (Td), generation time (GT) and infinite rate of increase (λ)] evaluated (Table 4). The net reproductive rate (Ro) on C. moschata was 1.6, 1.4, 1.6, 6.5 and 10.5 times higher than on M. indica, P. aculeata, C. cajan, F. benjamina and C. arabica, respectively. The intrinsic rate of increase was higher on M. indica (0.178) and the population was expected to double in 3.9 days. The lowest rm was recorded on C. arabica (0.102) with a doubling time of 6.8 days. Mangifera indica recorded the lowest generation time of 31 days and the highest duration occurred on F. benjamina. The finite rate of increase was 1.11 on F. benjamina and C. arabica and 1.20 on M. indica and C. cajan (Table 4).
DISCUSSION
Results of this study showed that the 6 host plant species tested support the development of R. iceryoides but the biological parameters measured varied significantly across the host plants tested. Previous field studies have suggested that M. indica, P. aculeata and C. cajan were the most heavily infested host plants by R. iceryoides (Williams 1989; Luhanga & Gwinner 1993; Gado & Neuenschwander 1993; CABI 2000; Tanga 2012). Our results concur with these findings as well as providing evidence that fruit of C. moschata is an equally suitable host for development and survival of R. iceryoides. The duration of development is potentially an important component of fitness in the field, as it will determine how long different developmental stages of the mealybug will be exposed to predators and parasites. For example, the effectiveness of several natural enemies depends on host growth rates: parasitism increases when host growth is slowed (Benrey & Denno 1997; Devine et al. 2000). The prolongation of nymphal stages of R. iceryoides on sub-optimal hosts may provide an important selective advantage under pressure from natural enemies, as demonstrated by several authors (Häggström & Larsson 1995; Parry et al. 1998; Awmack & Leather 2002). The 2 most efficient and widely distributed nymphal parasitoids of R. iceryoides, Anagyrus pseudococci (Girault) and Praleurocerus viridis (Agarwal) attack the second and third instars (Noyes & Hayat 1994; Tanga et al. 2012) and host plant-induced delays in development may increase parasitism rate on mealybugs although this could also be counterproductive in terms of parasitoid fitness if they are unable to make a choice for egg laying based on host quality (Tanga et al. 2012). On the other hand, a reduction in developmental duration on an optimal host could represent an advantage to the mealybug by reducing its vulnerability to parasitism and predation.
TABLE 4.
EFFECTS OF VARIOUS HOST PLANT SPECIES (CUCURBITA MOSCHATA, MANGIFERA INDICA, PARKINSONIA ACULEATA, CAJANUS CAJAN, FICUS BENJAMINA, AND COFFEA ARABICA) ON LIFE TABLE PARAMETERS OF RASTROCOCCUS ICERYOIDES.
Although development and survival were poor on F. benjamina and C. arabica, these host plants supported establishment of R. iceryoides. It is probable that some constituent compounds or physiological barriers inherent in this host plant species significantly reduced feeding, and, consequently, led to a reduction in development and survival of R. iceryoides. Despite these observations, C. arabica especially warrants careful monitoring given the previous history of invasion and impact of Planococcus kenyae Le Pelley on coffee in East Africa and its subsequent classical biological control by Anagyrus kivuensis Compere (Greathead 1971, 2003).
Our results illustrate that host plant affects adult reproductive output and longevity. We observed that in addition to increasing the speed of growth of R. iceryoides, rearing the mealybug on the most suitable host plant (M. indica, P. aculeata, C. cajan and C. moschata) resulted in higher progeny production and adult longevity. Van Lenteren & Noldus (1990) stated that shorter pre-reproductive period and increased reproductive capacity of an insect on a host reflect the suitability of the plant. This is confirmed by the current study and although we did not measure the nutritional content of the tested plant species, it is probable that it might have played a role in enhancing the reproductive success of the mealybug on the suitable host plants. Our find- ings also strongly corroborate the observations of Boavida & Neuenschwander (1995) who reported shorter pre-reproductive period and higher progeny production for Rastrococcus invadens Williams when reared on its most suitable host plant, M. indica. Matokot et al. (1992) showed that the development of R. invadens Williams (Homoptera: Pseudococcidae) varies considerably when reared on M. indica, Ficus sp., Plumeria sp. and Citrus spp. Marohasy (1997) reported no difference in development, survival and fecundity of cohorts of Phenacoccus parvus Morrison, when reared on Lantana camara L., Lycopersicon esculentum Miller and Solanum melongena L, but Gossypium hirsutum L., Ageratum houstonianum Miller and Clerodendrum cunninghamii Benth., were identified as less suitable host plants.
Although R. iceryoides was observed to lay eggs on F. benjamina and C. arabica, egg production was generally low when compared with reproduction on the other host plants. Leather (1995) noted that when an insect pest encounters a poor-quality host plant, it may modify its oviposition behavior, either by reducing the number of eggs laid on each plant or, in some cases, adjusting the size or nutritional content of the eggs. In extreme cases, where the quality of the host plant is too low to support adult survival, female insects may resorb eggs or embryos and use the nutrients gained to increase their longevity and thus their potential to find better-quality host plants for their offspring (Awmack & Leather 2002). Although we did not dissect the insects to observe the ovaries or measure the size and nutritional content of the eggs laid by R. iceryoides, Nelson-Rees (1960) noted egg resorption in the citrus mealybug, Planococcus citri (Rossi) and it is possible that such reproductive strategy occurs in the mango mealybug but this warrants further investigation.
Rastrococcus iceryoides sex ratio was significantly affected by host plant on which the insects were reared. Its progeny were female biased on M. indica, P. aculeata, C. cajan and C. moschata and male biased in the less suitable host plants (F. benjamina and C. arabica). This suggests that maternally influenced sex ratio distortion or mortality of either sex during egg and nymphal development are dependent on the host plant species used. Contrary to our study, developmental studies on R. invadens revealed a significantly male-biased sex ratio with male and female ratio ranging from 2.1:1 to 3.3:1 on its most preferred host plant, M. indica (Sahoo & Ghosh 2000). The reason for the discrepancy in results is unclear but in other insect herbivores, sex ratio on better quality host plants has been reported to be female-biased (Mopper & Whitham 1992; Craig et al. 1992; Barker & Maczka 1996).
The morphometric studies revealed that the body size of R. iceryoides was greatly influenced by the host plant type on which the mealybug was reared. As with the developmental studies, mealybugs reared on M. indica, P. aculeata, C. cajan and C. moschata had significantly larger body size than those reared on F. benjamina and C. arabica, which indicates that no fitness penalty is paid for rapid development and increased body size. Body size is influenced, among other factors, by differential nutritional quality of the host plant species, chemical constituents as well as physical differences in the plant structures that may affect development, reproduction, survival, behaviour and distribution of the phytophagous insect (Slansky & Rodriguez 1987; Bethke et al. 1998). Larger individual mealybugs have the potential to cause more plant damage, as food intake is positively correlated with body weight (Tanga 2012). Positive correlation between body size and fecundity are common in other insects (Haukioja & Neuvonen 1985; Sopow & Quiring 1998; Ekesi et al. 2007) and evidence suggests that similar relationships exist in female R. iceryoides (Tanga 2012). In other mealybug species, it has been reported that mealybugs feeding on host plant species with high nitrogen concentrations have increased egg loads, larger matured females, and shorter developmental time (Klingauf 1987; Bethke et al. 1998; Hogendorp et al. 2006). Conversely, it is likely that adult mealybugs emerging from suboptimal host plants tested may have less potential to inflict damage on the plant if their numbers, size and fecundity are lower.
The life table parameters provide, for the first time, comprehensive information on the survival, development and reproductive capacity of R. iceryoides on the different host plants tested. The intrinsic rate of natural increase (rm) is the most important parameter for describing the growth potential of a population under given climatic and food conditions, as it reflects the overall effect of development, reproduction and survival (Southwood & Handerson 2000). The results from this study indicate that M. indica and C. cajan are the most suitable among the plants tested for R. iceryoides with rm of 0.178 and 0.175, respectively. For R. invadens, Boavida & Neuenschwander (1995) reported rm values of 0.070–0.078 on M. indica. Considering the fact that the intrinsic rate of natural increase of R. iceryoides on mango is at least 2.2-fold that of R. invadens on the same host plant, the former pest could have a higher potential of threatening mango production in Africa compared to R. invadens. In addition to helping predict the population growth potential, and effectively time control strategies within an integrated pest management framework for the pest, life table statistics on the different host plants have practical implications for more efficient and effective production of mealybugs for parasitoid mass rearing and releases. Reproductive values (mx) would be helpful in determining the best host plants for rearing. To judge from the Ro and rm values, mass rearing would be suitable on the 4 most optimal host plants.
Overall, the results of this investigation provide strong indication that host plant species of R. iceryoides will have a significant impact on development and reproduction of the pest, thereby affecting the population growth parameters, and the timing and extent of mealybug damage to the plant species. Rastrococcus iceryoides development on highly suitable plants such as M. indica and C. cajan, as well as on the fruits of C. moschata may result in rapid development and greater numbers of mealybugs surviving to adulthood, and hence more damage on these host plants; and this observation has significant implications for management of the pest on the suitable host plants. Demographic parameters on the most suitable host also showed that these plant species as well as C. moschata fruits should be excellent targets for mass rearing of R. iceryoides parasitoids for field releases. The life table parameters also provide important new information on the biotic potential of the pest on its host plants that should be useful in developing simulation models that include other factors for field use and management of R. iceryoides (Gutierrez 1996; Sporleder et al. 2012). From the standpoint of conservation biological control, P. aculeata is an important ornamental shade plant used by growers in the vicinity of M. indica and C. cajan crops. An important management strategy would be to conserve the indigenous parasitoids of R. iceryoides: Anagyrus pseudococci Girault (Hymenoptera: Encyrtidae), Leptomastrix dactylopii Howard (Hymenoptera: Encyrtidae), Leptomastidea tecta Prinsloo (Hymenoptera: Encyrtidae), Agarwalencyrtus citri Agarwal (Hymenoptera: Encyrtidae), Aenasius longiscapus Compere (Hymenoptera: Encyrtidae) and Anagyrus aegyptiacus Moursi (Hymenoptera: Encyrtidae) (Tanga 2012) on these plants, and or augmentative releases of appropriate parasitoids on P. aculeata for parasitoid population buildup and subsequent suppression of R. iceryoides population before their spread into the cultivated crop. The information provided in this investigation should be essential in understanding the dynamics of R. iceryoides and should contribute to an integrated management action plan that allows for targeted suppression of the pest in East Africa.
ACKNOWLEDGEMENTS
This research was supported by a grant from the German Federal Ministry for Economic Cooperation and Development (BMZ) to the icipe-African Fruit Fly Programme (AFFP) and a student fellowship to the senior author by the German Academic Exchange Service (DAAD). We are thankful to the late Dr. A. Chabi-Olaye and to Dr. D. Salifu for helping with the statistical analyses, and Dr. S. Subramanian for his comments on an earlier draft. The assistance of P. Nderitu in acquisition of the plant materials and rearing activities is gratefully acknowledged.
