The Asian citrus psyllid, Diaphorina citri Kuwayama (Hemiptera: Psyllidae), has a widespread geographical distribution, and is a cosmopolitan insect with worldwide economic importance as a crop pest. First reported in Brazil in the 1940s (Costa Lima 1942), it was considered of secondary importance until the occurrence of huanglongbing (HLB), or citrus greening, a disease transmitted by this insect, which was confirmed in São Paulo in 2004 (Coletta-Filho et al. 2004; Teixeira et al. 2005).

Although D. citri causes direct damage such as deformations of growing leaves , shoot drying, defoliation, and the presence of sooty mold (Hodkinson 1974; Aubert 1987; Michaud 2004), it is as a vector of the plant pathogens causing HLB that this psyllid has acquired major importance for citrus, becoming the main insect pest of this crop. In Brazil, 2 species of bacteria, “Candidatus Liberibacter americanus” and “Candidatus Liberibacter asiaticus”, which inhabit the plant phloem, have been identified in association with plants showing HLB symptoms (Coletta-Filho et al. 2004; Teixeira et al. 2005). HLB is considered the most destructive and serious disease of citrus worldwide (Da Graça 1991; Halbert & Manjunath 2004; Bové 2006), and no effective control method is known, notably for the Asian HLB caused by “Ca. L. asiaticus”.

Ecological studies of D. citri, particularly based on life tables, are scarce but may lead to improved management strategies by providing a better understanding of psyllid population biology. Ecological life tables are useful for studying insect population dynamics, as they provide an understanding of changes in the populations of a species over time in its natural environment, under natural weather conditions and in the midst of other local organisms (Harcourt 1969). Biotic mortality factors such as pathogens, predators, and parasitoids have been reported in the regulation of D. citri populations (Halbert & Manjunath 2004). Coccinelids have been noted as important predators of this pest and play an important role in its biological control (Michaud 2004). The parasitoid, Tamarixia radiata Waterston (Hymenoptera: Eulophidae), is also acknowledged as an important natural enemy of D. citri worldwide (Étienne et al. 2001), and was recently found in Brazil (Gómez Torres et al. 2006). We studied the changes in population densities and the main factors causing natural mortality of D. citri in orange orchards (Citrus sinensis Osbeck), based on an ecological life table that we developed in the regions of Araras and Piracicaba over 4 generations, in order to provide information for the control of this insect.

MATERIAL AND METHODS

RESULTS AND DISCUSSION

Considering the data from all shoots in all 4 experiments, the overall means (∓ standard error of the mean) were as follows: 8.3 (∓0.43) eggs, 3.5 (∓0.33) small nymphs and 1.1 (∓0.22) adults per flush shoot. In the Araras experiment, the highest mortalities were observed for both the small and large nymphal stages. As a consequence, egg-adult survivorship was very low. From the 406 eggs observed at the beginning of the experiment, only 7 adults were obtained (Table 1). Of the 91 large, live nymphs, 6 were parasitized by T. radiata. Thus the estimated parasitism rate was 6.6%. In addition to parasitizing D. citri nymphs, T. radiata females may kill this host by their feeding activity (Skelley & Hoy 2004). However, no killed nymphs other than those that were parasitized (mummified) could be identified in this experiment.

TABLE 1.

LIFE TABLE FOR IMMATURES OF THE PSYLLID DIAPHORINA CITRI ON CITRUS SINENSIS, FROM XI-2006 TO V-2007 IN 2 REGIONS OF THE STATE OF SÃO PAULO.

In the 3 generations of the Piracicaba population, the egg stage showed the lowest survival (Table 1), with rates of 38.8, 30.2, and 35.7% in the first, second, and third generations, respectively. In Araras also, only a small number of adults emerged due to high mortality of small nymphs. A large population of eggs and nymphs were observed but the mortality rates of these stages were very high, and thus survivorship of eggs to adults was minimal. Increased control of eggs and small nymphs would have reduced survivorship even more.

No predator was observed in association with D. citri eggs or nymphs at either site. In other citrus groves in São Paulo, Cycloneda sanguinea (L.), Harmonia axyridis (Pallas), and Scymnus sp., all coccinellids, have been found on shoots with D. citri, but they were often observed preying aphids. Only a larva of the green lacewing, Chrysoperla externa (Hagen), was observed preying nymphs and adults of D. citri. However, there could be additional mortality as a result of nocturnal predation.

The observed mortalities for large nymphs in Piracicaba were quite similar among generations, with 24.3, 33.3, and 25.1% in the first, second, and third generations, respectively. The ectoparasitoid, T. radiata, apparently made only a small contribution toward high nymphal mortality in the Piracicaba experiments, with 0.0, 2.2, and 2.3% parasitism in the first, second, and third generations, respectively. Additional mortality may have occurred due to parasitoid feeding, but insects killed by this action could not be identified, nor could the contribution of feeding be quantified.

Although egg survival was low in the field, less than 40% in the 3 generations of the Piracicaba population, the laboratory experiment showed egg survival above 70%. Egg survival of the same population was 75.3% at 30 °C and RH below 50%, conditions that are similar to those observed in the field in the period from 12:00 to 17:00 h.

The hypothesis was tested that shoots infested with numerous small nymphs would result in reduced survival from egg to adult because of a competition effect. However, the opposite was observed, since there was a positive relationship between initial small nymphal density and survivorship up to the adult stage (F = 26.9, df = 1,189, P < 0.0001, r2 = 0.12), i.e., survivorship for that stage on shoots with a higher initial number of nymphs was higher. As a result, apparently there was no competition among the nymphs. It is likely that a no competition effect occurs at densities such as those observed in the Piracicaba populations, i.e., 7.9 ∓ 0.56 nymphs per shoot (considering only infested shoots). Hodkinson (1974) suggested that, in general, higher nymphal densities of psyllids would result in higher mortalities, whereas lower densities would be conducive to survival. It would be interesting to determine at what nymphal density per shoot such competition would occur in D. citri populations.

The sex ratios of emerged adults were similar in the 3 generations of the Piracicaba population, being 0.504, 0.613, and 0.561 for the first, second, and third generations, respectively. This indicates that field populations have ratios of not less than nearly 1 female to each male, as observed in the study of Nava et al. (2007).

Since T. radiata preferentially parasitizes third-, fourth- and fifth-instar nymphs (Étienne et al. 2001), the use of cages to obtain emerged D. citri adults may have limited the effect of the parasitoid in these experiments. In order to quantify parasitism in the area without cage interference, a parasitism estimate was obtained on 17-IV-2007 with insects collected from the same orchard, but outside the experiment area. Of 126 fourth- and fifth-instar nymphs inspected for parasitoid presence, only 2 were parasitized (1.6%), supporting the small contribution of parasitism toward D. citri nymphal mortality.

Larger nymphs, especially fifth instars, were found on the petioles of larger leaves. This behavior could be for feeding or to prevent parasitism, since no parasitized nymphs were observed at that site; on the other hand, this position on leaf petioles could prevent defoliation. This should be investigated in future studies.

Despite the higher incidence of defoliation, 74 females were obtained in the third Piracicaba generation, the largest population among the 4 generations studied. Defoliation reduced the plant surface that could be exploited by young insects with a possible effect on the amount of food. In addition to defoliation itself, the effect of defoliation on the nutritional quality of leaves for D. citri could be studied in the future. Nevertheless, defoliation is not a controllable factor, since in practice it cannot be increased and leaves are essential for citrus tree development and for the production of fruits.

In addition to weather factors, D. citri nymphal survival seemed to be influenced by the host tree and shoots where the insects develop. Nutritional aspects of the shoots may be important for insect development, since grafted trees are “unnatural” host plants for the insect. Two combined citrus species or compatible genera, such as C. sinensis orange grafted onto Rangpur lime C. limonia, or the ‘Swingle’ citrumelo, a hybrid between C. paradisi and P. trifoliata, present as an “unknown” host for the insect during its evolution. The evolution of D. citri occurred on seed-propagated hosts represented by species of Citrus and other members of Rutaceae. To our knowledge no information on this “unknown” host factor is available.

Egg-adult survival varied greatly between shoots; in Piracicaba population 70% of the shoots showed zero egg-adult survival and only 5% of the shoots had more than 80% survival. No relationship was observed between final shoot length and egg-adult survival in the first and third Piracicaba generations (first - F = 0.002, df = 1, 69, P = 0.9635, third - F = 0.25, df = 1, 57, P = 0.6167). The mean value for final shoot length (∓ sem) in the first and third generations were 25.0 ∓ 0.95 and 17.0 ∓ 1.11 cm, respectively. However, there was a significant and positive correlation between final shoot length, 21.2 ∓ 0.95 cm, and survivorship of the immature stages of D. citri in the second generation studied (F = 5.64, df = 1, 55, P = 0.0210, r2= 0.09). Considering the data from the 3 Piracicaba generations, survival from egg to adult did not depend on the final length (21.3 ∓ 0.66 cm) of the orange tree shoots where the insects developed (F = 0.94, df = 1,185, P = 0.3318).

The hypothesis was tested that freshly sprouted buds or shoots beginning to grow, which produced larger shoots, would harbor more eggs. No relationship was found between number of eggs and final shoot length (F = 0.34, df = 1, 185, P = 0.5583). The manner by which D. citri females select a host for oviposition is still unknown. However, oviposition is performed on freshly sprouted buds, which may have distinct physical and nutritional characteristics; this would explain the variation in the number of eggs laid on each shoot.

Duration of the psyllid egg-to-adult period was estimated from the interval between the beginning of the experiment (egg masses) and adult emergence. Survivorship for this period was obtained from the relationship between the number of emerged adults and the number of eggs in the beginning of the experiment (Table 2).

The estimated survival from egg to adult differed among the 4 D. citri generations studied (F = 6.34, df = 3, 241, P = 0.0004). The lowest survival rates were observed in Araras and in the second Piracicaba generation, at 1.7 and 6.0%, respectively. The highest survivals occurred in the first and third generations of the Piracicaba population, at 17.6 and 21.4%.

Laboratory studies have found shorter durations for the egg-adult period, from 17 to 18 d at 25 °C (Tsai & Liu 2000) and 17 d at the same temperature (Nava et al. 2007) in relation to those obtained in the present work. Survival values of D. citri immatures observed under controlled conditions, in the same studies, were significantly higher, i.e., 70% (Tsai & Liu 2000) and 70 to 80% (Nava et al. 2007) than survival estimates obtained by us in the field. Survival rates were lower than 10% in Araras and in the second generation in Piracicaba, and lower than 25% in the first and third generations in Piracicaba.

The times of adult emergence peaks were similar between the Araras populations (19 d) and the first and second generations from Piracicaba (19 and 20 d) (Fig. 1). In the third Piracicaba generation, the greatest emergence of adults occurred 25 d after the beginning of the experiment, due to lower temperatures. The emergence of adults over several d could be an evolutionary survival strategy in this insect species, since the development of growing citrus shoots, an essential resource for D. citri feeding and reproduction, also takes place over time. The emergence of adults over several d may make insect control more difficult.

The expected values obtained from the mathematical model proposed by Nava et al. (2007) with a Brazilian D. citri population did not differ from the estimated values for duration from egg to adult obtained in this study (Table 3). Similarly, the models proposed by Nakata (2006) with a population in Japan, and Liu & Tsai (2000) with a D. citri population in Florida, USA, also showed egg-adult durations similar to those observed in Brazilian orchards (Table 3). This information indicates that the São Paulo (Brazil) populations, both, those studied by Nava et al. (2007) and those from Araras and Piracicaba, as well as the populations from Japan and Florida had similar temperature requirements for the immature stages.

TABLE 2.

MEAN DAILY MAXIMUM AND MINIMUM TEMPERATURES (∓STANDARD ERROR OF MEAN), CUMULATIVE RAIN (MM), PRECIPITATION DAYS DURING THE EXPERIMENTS, AND BIOLOGICAL DATA FOR DIAPHORINA CITRI ON CITRUS SINENSIS BETWEEN XI-2006 AND V-2007, IN 2 REGIONS OF SÃO PAULO.

TABLE 3.

DURATION OF DIAPHORINA CITRI FROM EGG TO ADULT, IN DAYS, IN 4 GENERATIONS STUDIED, AND VALUES ESTIMATED ACCORDING TO MATHEMATICAL MODELS FOR THE TEMPERATURE REQUIREMENTS OF DIFFERENT POPULATIONS.

Fig. 1.

Duration of the egg-to-adult period and number of emerged Diaphorina citri adults on Citrus sinensis in 4 generations between XI-2006 and V-2007.

The estimates for survival in the field were much lower than those observed under laboratory conditions. On Murraya paniculata L. Jack at temperatures from 15 to 30 °C, egg-adult survival values were higher than 60% (Liu & Tsai 2000), while on C. limonia at temperatures from 18 to 30 °C, survival values were higher than 70% (Nava et al. 2007). Consequently, weather, host plant, interspecific and intraspecific competition factors, among others, may reduce survivorship in the field.

Although in this investigation the insect was studied for 4 generations in less than a yr, the study period comprised the spring, summer, and fall seasons, when the highest populations of this insect have been found. The observations that the highest mortalities occur in the initial life stages (eggs and initial nymphal instars) demonstrate that they are more susceptible to adverse weather and host conditions. Additional mortality rates for late-instar nymphs caused by the parasitoid T. radiata were low.