Intra- and interspecific competition between western flower thrips and sweetpotato whitefly.

The western flower thrips, Frankliniella occidentalis (Pergande), and the sweetpotato whitefly, Bemisia tabaci (Gennadius), are both invasive insect pests and are present in most of the same agricultural crops without a clear dominance of either species. Here, intra- and interspecific competition in B. tabaci and F. occidentalis was determined under controlled experiments. The results showed that intraspecific competition was distinct in F. occidentalis and that the co-occurrence of B. tabaci had a strong effect on F. occidentalis, resulting in a decrease in oviposition. Significant intraspecific competition was found in B. tabaci, and the coexistence of F. occidentalis had limited effect on the oviposition of B. tabaci. In a selective host plant preference experiment, both F. occidentalis and B. tabaci preferred eggplants most, followed by cucumbers and tomatoes. On cucumber plants, B. tabaci was predominant, whereas on eggplant and tomato plants, F. occidentalis and B. tabaci exhibited comparative competitive abilities during the initial stage. However, over time, higher numbers of B. tabaci than that of F. occidentalis were found on the two host plants. Our in vitro and potted plant experiments indicate that B. tabaci is competitively superior to F. occidentalis, which might help to explain their differential distribution patterns in China.

The western flower thrips, Frankliniellaoccidentalis (Pergande) (Thysanoptera: Thripidae), was reported in China for the first time in 2003 ( Zhang et al. 2003 ). The sweetpotato whitefly, Bemisia tabaci (Gennadius) (Hemiptera: Aleyrodidae), was first recorded in China in 1949, but no significant damage was caused by this insect until the late 1990s, when a new B-biotype from Middle East-Minor Asia 1 invaded China ( Luo et al. 2002 ). Since then, this B-biotype and a later invasive Q-biotype of the Mediterranean group have become the dominant B. tabaci biotypes in China ( Chu et al. 2006 , Pan et al. 2011 ). F.occidentalis and B. tabaci are small hypermetamorphic insects with many common characteristics. The adults mainly reproduce sexually but also by parthenogenesis. Both species are globally distributed invasive insect pests ( CABI 2012 ). They both feed directly on leaves and have an adverse impact on leaf size and photosynthesis. Both can cause indirect damage, which is especially important due to the transmission of plant viruses, such as tomato yellow curl virus by B. tabaci ( Jones et al. 2003 ) and tomato spotted wilt virus by F. occidentalis ( Whitfield et al. 2005 ). Furthermore, they have a wide range of host plants and favor various horticultural crops, and can be found on vegetable crops such as tomatoes, cucumbers, peppers, and melons.

Generally, two species with highly similar fundamental niches would often compete strongly with each other upon first encounter. Interspecific competition is defined as a reduction in individual fecundity, survival, or growth as a result of exploitation of resources or by interference with individuals of another species. Some studies have shown that interspecific competition is widespread among insects ( Denno et al. 1995 , Stewart 1996 , Reitz and Trumble 2002 ). The most severe outcome of interspecific competition is competitive displacement, about 78% of which is triggered by introduction or invasion of an exotic species ( Reitz and Trumble 2002 ). Well-documented cases of competitive displacement have occurred after the importation of Aphytis spp. parasitic wasps ( DeBach 1966 ). The introduced Aphytismelinus , the most widespread natural enemy of the California red scale ( Aonidiella aurantii ) and a superior competitor, has displaced the native Aphytischrysomphali from most citrus areas of the Mediterranean basin and other citrus areas across the globe ( Sorribas et al. 2010 ). Moreover, competitive displacement mediated by different mechanisms also takes place between invasive insects, such as beetles ( Lee et al. 2010 ), leaf miner flies ( Gao et al. 2011 ), and whiteflies ( Liu et al. 2012 ). More commonly, different phytophagous insect species usually co-occur, although interspecific competition is often asymmetrical, so that the effect of one species on another is much more prominent than the reverse ( Lawton and Hassell 1981 ). For instance, the current distribution of Aedes albopictus and Aedesjaponicus in the United States overlaps considerably, and A. albopictus has a competitive advantage over A. japonicus ( Armistead et al. 2008 ). Peixoto and Benson (2009) reported the co-occurrence of two tropical satyrine butterflies through completely different daily activity patterns to avoid stressful climatic conditions. Klapwijk and Lewis (2011) investigated the spatial ecology of multiple parasitoids. Venner et al. (2011) observed four competing weevil sister species to commonly co-occur on the same oak trees. Here, the marked time partitioning of the resource use appeared as a keystone of their coexistence.

As invasive insect pests, both intra- and interspecific competition between F. occidentalis or B. tabaci and native species has been widely studied and discussed ( Brown et al. 1995 , Pascual and Callejas 2004 , Paini et al. 2008 ). However, it is not clear whether interspecific competition exists between the invasive F. occidentalis and B. tabaci . In China, invasive F. occidentalis has been reported from Beijing, Yunnan, Shandong, and Guizhou ( Wu et al. 2007 , Zheng et al. 2007 , Yuan et al. 2010 ), whereas B. tabaci has been found almost everywhere ( Pan et al. 2011 ). We speculate that apart from other factors affecting the different range of their distribution, B. tabaci might outcompete F. occidentalis . During indoor rearing of F. occidentalis and B. tabaci on host plants, we found that F. occidentalis on some seedlings were often contaminated with or sometimes replaced by B. tabaci , and vice versa, which seems to indicate that the competition between the two species is host plant dependent. The coexistence of F. occidentalis and B. tabaci on the same vegetable crops might result in them competing for similar resources. Therefore, the major objective of this study was to determine whether both intra- and interspecific competition would occur between F. occidentalis and B. tabaci under controlled experimental conditions. The fecundities of F. occidentalis and B. tabaci at different densities were compared in the laboratory, and their host preference was compared in a cage experiment.

Materials and Methods

Insect Pests

F . occidentalis and B. tabaci B-biotype were used in this study. F.occidentalis has been reared on bean pods in our laboratory since 2003 ( Zhang et al. 2007 ). B.tabaci B-biotype was originally collected on cabbage plants from a field of the Institute of Vegetables and Flowers, Chinese Academy of Agricultural Sciences, China, in 2000, and has subsequently been maintained on cabbage plants in a glasshouse ( Feng et al. 2009 ).

Competition Experiment on Cut Leaves

Intraspecific Competition Within F. occidentalis.

The experiment containers consisted of agar tubes and were prepared as follows: agar was diluted with distilled water into a concentration of 15–17 g/ml in a triangle flask and then heated in a microwave oven at medium level for 3–4 min. The agar was allowed to cool down, and 5 ml of liquid agar was transferred to a flat-bottomed glass tube (78 mm length, 22 mm diameter) by using a micropipettor (Eppendorf AG, Germany). Great care was taken to avoid liquid tainting of the inner tube wall or producing air bubbles. Leaf discs (22 mm diameter) were cut from Phaseolus vulgaris leaves and placed with their abaxial side down on the agar bed. One-day-old female and male adults of F. occidentalis with densities of 3, 6, and 10 pairs, respectively, were placed into each tube with an aspirator. Each density had 15 replicates. The open ends of the tubes were sealed with cotton plugs, and the tubes were placed in growth chambers upside down at a photoperiod of 14:10 (L:D) h at 26°C. The leaves were changed daily after the F. occidentalis on the leaves had been carefully removed by an aspirator. The displaced leaves were put in Petri dishes (10 cm diameter) filled with moistened filter papers. The Petri dishes were maintained in the same growth chambers. Because eggs are laid inside the leaf and are not visible, the number of the first instars hatching on replaced leaves was counted and used to calculate the daily fecundity of females ( Watts 1934 , Zhang et al. 2007 ). The observation lasted for 7 d.

Intraspecific Competition Within B. tabaci.

The experimental containers were similar to the agar tubes described earlier. Newly emerged female and male adults of B. tabaci with densities of 3, 6, and 10 pairs, respectively, were placed into each tube with an aspirator. Each density had 10 replicates. The open ends of the tubes were sealed with cotton plugs, and the tubes were placed in growth chambers upside down at a photoperiod of 14:10 (L:D) h at 26°C. The leaves were changed daily after the B. tabaci on the leaves had been carefully removed by an aspirator. The displaced leaves were placed in Petri dishes (10 cm diameter) filled with moistened filter papers. The eggs on leaves were counted under a stereomicroscope. The observation lasted for 7 d.

Interspecific Competition Between F. occidentalis and B. tabaci.

The experimental containers were similar to the agar tubes described earlier. B.tabaci were first inoculated, followed by F. occidentalis . The B. tabaci used in this experiment were the newly emerged female and male adults, and the F. occidentalis were 1-d-old female and male adults. All insects were transferred to the tubes by using an aspirator. The procedure was similar to that described in the previous section. The densities were 3 pairs of B. tabaci with 3 pairs of F. occidentalis , 6 pairs of B. tabaci with 6 pairs of F. occidentalis , and 10 pairs of B. tabaci with 10 pairs of F. occidentalis , respectively. Each density had 10 replicates. The number of the first instars of F. occidentalis and the number of eggs of B. tabaci on leaves were counted under a stereomicroscope. The observation lasted for 7 d.

Host Plant Preference Experiment.

Eggplants ( Solanum melongena ) with three to four leaves, cucumber ( Cucumis sativus ) plants with three to four leaves, and tomato ( Solanumlycopersicum ) plants with three to four compound leaves were used in the experiment.

Multiple Choice Experiment

Two plants of each of the three plant species ( S. melongena , C. sativus , and S. lycopersicum ) were placed in one cage (60 by 60 by 60 cm) in an alternating arrangement, i.e., six host plants were placed in one cage in this experiment. Fifteen pairs of newly emerging female and male adults of F. occidentalis and 15 pairs of 1-d-old female and male adults of B. tabaci were collected with an aspirator. The insects were put in Eppendorf tubes and then simultaneously released onto plant leaves in the cage. The experiment was conducted in three replicates, and the cages were placed in a glasshouse. The numbers of B. tabaci and F. occidentalis on every plant were investigated for a month at a 5-d interval. The plants were watered when necessary.

Nonchoice Experiment

Three plants of S. melongena , C. sativus , or S. lycopersicum were placed in a cage, i.e., only one species of host plant was used per cage. Five pairs of F. occidentalis and five pairs of B. tabaci were collected and released by using the same method as described earlier. The experimental conditions and investigation methods were the same as in the previous section.

Statistical Analysis

Statistical analyses were performed with SPSS (version 13.0; SPSS, Chicago, IL). In the intraspecific competition experiments, the effect of densities on female fecundity was tested by one-way Analysis of Variance (ANOVA). In the interspecific competition experiments, the effects of densities and insect species on female fecundity were tested by two-way ANOVA. In the multiple-choice experiment, the effects of insect species and host plants on insect settling preferences were tested by repeated-measures ANOVA. In the nonchoice experiment, the effect of mutual interferences between F. occidentalis and B. tabaci on insect feeding was tested by repeated-measures ANOVA. Tukey’s test was used to separate treatment means when the main effect or the interaction was significant.

Results

Fecundity of F. occidentalis Alone and Coexisting With B. tabaci

For intraspecific competition in F. occidentalis , the number of eggs laid per F. occidentalis per day decreased with increased density ( Fig. 1 A), resulting in a significant decrease in the total fecundity (one-way ANOVA: Fig. 1 B, F 2,42 = 47.119, P  < 0.0001). The highest fecundity (19 eggs per F. occidentalis ) was observed under the density of three pairs, whereas only eight eggs per F. occidentalis at the density of 10 pairs. When F. occidentalis coexisted with B. tabaci , its fecundity was gradually decreased with the increased density of B. tabaci (two-way ANOVA: Fig. 1 C and D, F 1,69 = 380.486, P  < 0.0001). The total fecundity of F. occidentalis at a density of 10 pairs was significantly lower than that at six and three pairs ( Fig. 1 D, F 2,27 = 11.785, P  < 0.0001).

Fig. 1.

Daily and total fecundity per F. occidentalis alone (A, B) or coexisting with different densities of B. tabaci (C, D). WFT: western flower thrips, F. occidentalis ; SW: sweetpotato whitefly, B. tabaci. Bars represent mean ± SE. Different lowercase letters above the error bars in (B) or (D) indicate significant difference (Tukey’s test, P  < 0.05).

Fecundity of B. tabaci Alone and Coexisting With F. occidentalis

For intraspecific competition in B. tabaci , the fecundity was uncorrelated with densities ( Fig. 2 A). The total fecundity at a density of 10 pairs was significantly lower than that at six pairs (one-way ANOVA: Fig. 2 B, F 2,27 = 4.974, P  = 0.014). The existence of F. occidentalis resulted in significant decreased fecundity of B. tabaci and higher density of F. occidentalis had a stronger effect on the fecundity of B. tabaci (two-way ANOVA: Fig. 2 C and D, F 1,54 = 36.142, P  < 0.0001). The total fecundity of B. tabaci gradually decreased with densities ( F 2,27 = 64.743, P  < 0.0001).

Fig. 2.

Daily and total fecundity per B. tabaci alone (A, B) or coexisting with different densities of F. occidentalis (C, D). WFT: western flower thrips, F. occidentalis ; SW: sweetpotato whitefly, B. tabaci. Bars represent mean ± SE. Different lowercase letters above the error bars in (B) or (D) indicate significant difference (Tukey’s test, P  < 0.05).

Selective Host Plant Preferences of F. occidentalis and B. tabaci

Equal numbers of eggplant, tomato, and cucumber plants were placed in one cage, and at the same time, equal numbers of F. occidentalis and B. tabaci were released onto the plants in the cage. Insect settling preference was significantly affected by insect species (repeated-measures ANOVA: F 1,12 = 265.186, P  < 0.0001), by host plant (repeated-measures ANOVA: F 2,12 = 50.092, P  < 0.0001), and by the interaction between insect species and host plant (repeated-measures ANOVA: F 2,12 = 50.349, P  < 0.0001). Among the three tested host plants, the total numbers of B. tabaci were significantly higher than that of F. occidentalis ( Fig. 3 ). In addition, significant higher numbers of insects (both F. occidentalis and B. tabaci ) were observed on eggplants than on tomato or cucumber plants ( Fig. 3 ). Furthermore, F. occidentalis preferred cucumber plants over eggplant and tomato plants, whereas B. tabaci preferred eggplants over cucumber and tomato plants ( Fig. 3 ).

Fig. 3.

Mean number of F. occidentalis and B. tabaci settled on different host plants over time in choice experiment.

Nonselective Host Plant Performance of F. occidentalis and B. tabaci

On eggplant and tomato plants, the population increases of F. occidentalis and B. tabaci showed a similar trend. The F. occidentalis population increased faster than that of B. tabaci , although at 5 d after treatment its size was smaller ( Fig. 4 ). However, the mean total numbers of F. occidentalis and B. tabaci on eggplants and tomato plants were not significantly different (repeated-measures ANOVA: eggplant, F 1,4 = 0.018, P  = 0.90; tomato, F 1,4 = 3.875, P  = 0.120). On cucumber plants, the numbers of F. occidentalis and B. tabaci increased slowly for 15 d after inoculation. F.occidentalis showed the same trend during the entire test, with a mean total number of 0.8 individuals per plant. B. tabaci exhibited a different trend and increased quickly to a mean number of 23.7 individuals per plant at 20 d after treatment ( Fig. 4 ). The mean total number of B. tabaci individuals on cucumber plants was significantly higher than that of F. occidentalis (repeated-measures ANOVA: F 1,4 = 488.058, P  < 0.0001). The results indicate that on eggplants and tomato plants, F. occidentalis and B. tabaci have a similar population increase potential, whereas on cucumber plants, B. tabaci might be stronger.

Fig. 4.

Mean number of F. occidentalis and B. tabaci settled on different host plants over time in nonchoice experiment.

Discussion

Invasion success and spread of nonnative species can be enhanced by superiority in interspecific competition, particularly when similar species and limited resources are encountered ( Williamson 1996 ). F.occidentalis and B. tabaci are superior competitors compared with some native species, which is often cited as the main reason for their invasive success ( Kirk and Terry. 2003 , Chu et al. 2010 ). Our results indicate that under the conditions of limited food and space (such as a glass tube), the daily reproduction and the total mean reproduction of F. occidentalis decrease with increasing density, and that the intraspecific competition is distinct in F. occidentalis ( Fig. 1 A and B). The coexistence of B. tabaci had great effect on F. occidentalis , resulting in a decreased fecundity ( Fig. 1 C and D). Although significant intraspecific competition was found in B. tabaci ( Fig. 2 A and B), the coexistence of F. occidentalis had limited effect on the reproduction of B. tabaci ( Fig. 2 C and D). Intraspecific competition is a particular form of competition in which members of the same species vie for the same resource in an ecosystem. Our results indicate that for F. occidentalis , both the intraspecific competition and the interspecific competition from B. tabaci were stronger . In B. tabaci , on the other hand, the interspecific competition from F. occidentalis was a little stronger than the intraspecific competition, in which the egg-laying amount decreased more in interspecific than in intraspecific competition. Northfield et al. (2011) reported significant intraspecific competition occurring in F. occidentalis . A likely explanation might be that F. occidentalis is a facultative predator of spider mites ( Wilson et al. 1996 ) and that higher densities could cause it to become more aggressive. In B. tabaci , however, intraspecific competition occurred among immature insects rather than among adults ( Pascual and Callejas 2004 ).

Both F. occidentalis and B. tabaci are small-size insects and have a close relationship with plants ( Morse and Hoddle 2006 , Inbar and Gerling 2008 ). They both have a wide range of host plants but different host preferences. Our multiple-choice experiment showed that F. occidentalis and B. tabaci favored eggplants over cucumber and tomato plants ( Fig. 3 ). On cucumber plants, B. tabaci was predominant, whereas on eggplant and tomato plants, the F. occidentalis populations increased faster than those of B. tabaci . The F. occidentalis population also showed a stronger capacity of population growth during the initial stage. However, over time, higher numbers of B. tabaci than F. occidentalis were found on the two host plants ( Fig. 4 ). This difference might be due to the biological differences between the two insect species. F.occidentalis develops faster but has a lower fecundity than B. tabaci . On average, the generation time of B. tabaci is about 25 d, and its lifetime fecundity is about 100 eggs per female ( Mansaray and Sundufu 2009 ), whereas for F. occidentalis , the mean generation time is about 15 d, and the lifetime fecundity is 70–80 eggs per female ( Zhang et al. 2007 ). The crop growth period could be another possible reason. Some studies have reported that pollen can have a positive effect on growth, development time, and fecundity of F. occidentalis ( de Jager and Butôt 1993 , Hulshof et al. 2003 ). The host plants used in this experiment were at the stage of vegetative growth and not many flowers were present. The fewer flowers might affect the population growth of F. occidentalis but not of B. tabaci.

It is possible that competition from B. tabaci reduced F. occidentalis density on the plants. The mechanisms that lead to the competitive superiority of B. tabaci over F. occidentalis are yet unclear. Interference-type competitive interactions could have occurred between the two species, thus showing that the presence of B. tabaci resulted in the absence of an oviposition peak in F. occidentalis , whereas the presence of F. occidentalis resulted in a decreased reproduction time in B. tabaci . Thus, our in vitro and potted plant experiments contribute to explaining their different distribution patterns in China.