Species association

In order to study the species association of the two fruit flies, two types of habitats (e.g., fruit orchards and melon farmland) where these flies frequently occur, were selected for investigation from July to October 2022. Five survey sites were established in each type of habitat, with each site covering an area of more than 10 mu and ensuring a distance of > 5 km between sites. The 5 melon farmlands were located in Fengmu and Xinxing Town, Tunchang, Wancheng and Liji Town, Wanning, and Longtang Town, Haikou. The first 3 sites were planted with Momordica charantia (L.), while the last 2 were planted with Cucumis sativus (L.). The 5 fruit orchards were located in Baocheng Town, Baoting (planted with Nephelium lappaceum (L.)), Dazhipo and Sanmenpo Town, Haikou (planted with Psidium guajava (L.)), and 2 orchards in Poxin Town, Tunchang (one was planted with Artocarpus heterophyllus (Jean-Baptiste) and the other was planted with Citrus maxima (Bum.)).

The 5-point sampling method was used to determine the occurrence of B. cucurbitae and B. tau in each site for both habitats, thus, there were 25 survey points for each habitat. The attractant, 1 mL of ethyl acetopyruvate (Kunke Trading Co., Ltd, Heze, China), was added dropwise into the cotton inside a white trap bottle (h = 15 cm, d = 6 cm, Duoyuduo Biotechnology Co., Ltd, Guangzhou, China). The trap bottle was hung at each survey point. In the melon farmlands, the trap bottle was suspended on the central support pole of covered structure at a height of 1 to 1.5 m from the ground, while in the fruit orchards, the trap bottle was hung on the main trunk of the fruit tree at the same height. After 24 h, the trap bottle was removed, and the number of captured fruit flies of both species was recorded.

Ecological niche width and overlap

Two methods, literature review and oviposition preference tests, were used to evaluate the ecological niche width and overlap of B. cucurbitae and B. tau. For the literature review method, we used academic databases such as China National Knowledge Infrastructure (CNKI), Google Scholar, Web of Science (WOS), and PubMed as data sources. This approach ensures the comprehensiveness and thematic relevance of the literature review and analysis. The search formula was: “Bactrocera cucurbitae”, “Bactrocera tau” combined with “oviposition frequency”, “ecological and physiological character”, “specific interactions”, “management”, “population”, “behavior”. The study screened literature based on titles, abstracts, and full texts, and extracting data on the oviposition of the two fruit fly species on different hosts from the selected literature.

For host preference tests, these two species of fruit flies were collected and reared for use. Bactrocera cucurbitae, originating from the Chinese Academy of Tropical Agricultural Sciences, was reared in the laboratory for 10 generations to establish a stable laboratory population. Bactrocera tau, collected from the fruits of M. charantia in the Haidian campus of Hainan University, Haikou, were also reared in the laboratory for 10 generations to establish a laboratory population.

In the laboratory, larvae of both B. cucurbitae and B. tau were fed with Cucurbita pepo (L.). C. pepo containing larvae were placed in plastic trays, with sand as the substrate for pupation. Every 3 d, the sand was sifted to collect the pupae, which were then placed into plastic boxes and reared under controlled conditions of 26 ± 1°C; relative humidity (RH) of 65 ± 3%; photoperiod of 16 L:8 D. After emergence, adults were immediately transferred to 60 × 60 × 60 cm insect-rearing cages and provided with food and water. The adult diet consisted of a mixture of yeast extract and glucose (1:3).

According to the results of the literature review (see above), 10 common host fruits were selected. These included Cucurbita pepo (L.), Cucumis sativus (L.), Momordica charantia (L.), Luffa cylindrica (L.), Cucurbita spp (Joseph), Citrullus lanatus (Mansfeld), Solanum lycopersicum (L.), Musa spp (L.), Lagenaria siceraria (L.), Citrus sinensis (Osbeck). All host fruits were from local supermarkets and washed several times to reduce any pesticide residue.

Host preference tests

Ten host fruits (see above) were cut into small cuboids each with a volume of 2 × 2 × 3 cm³, and specimens of each species were placed in a Petri dish (d = 7 cm). Thus, 10 Petri dishes were used and randomly arranged in a circular pattern within the cage (30 × 30 × 30 cm) to avoid errors caused by different placement positions. Then, ten 15-d-old female adults of B. cucurbitae or B. tau were introduced into the cage. The test hosts were removed after 24 h of oviposition and examined under an anatomical microscope to observe and record the number of eggs on the fruits. Twenty replicates were conducted for each of host.

Data analysis

To determine the difference in the number of fruit flies (B. cucurbitae and B. tau) captured between the two habitats, the Kruskal-Wallis test was used, followed by Šídák multiple comparison tests. Then, the variance ratio (VR) test, which is based on species presence or absence, was used to examine the overall association (Chow et al. 1993, Schluter and Dolph 1984). The VR was calculated using the following formula: $${\delta }_{\mathrm{T}}^2={\displaystyle \sum }_{\mathrm{i}=1}^{\mathrm{S}}{\mathrm{P}}_{\mathrm{i}\left(1-{\mathrm{P}}_{\mathrm{i}}\right), {\mathrm{P}}_{\mathrm{i}}=\frac{{\mathrm{n}}_{\mathrm{i}}}{\mathrm{N}}}$$ $${\mathrm{S}}_{\mathrm{T}}^2=\frac{1}{\mathrm{N}}{\displaystyle \sum }_{\mathrm{j}=1}^{\mathrm{N}}{\left({\mathrm{T}}_{\mathrm{j}}-\mathrm{t}\right)}^2$$ $$VR=S_T^2/{\delta }_T^2$$where S = total number of species, N = total number of quadrats, n_i = number of quadrats where species i appears, T_j = total number of species in quadrat j, and t = average number of species in the quadrat, that is, t is the average number of species in the quadrats (Pandey et al. 2023). When VR = 1, there is independence of the species, VR > 1 indicates a positive association of the species, and VR < 1 indicates a negative association of the species in the community. The statistic W was used to test for the significance of the deviation of VR value from 1 using the formula: $$W=VR\times N$$where W follows the chi-squared distribution, with 0.00393 < X ²< 3.841. If W < X²_0.095N or W > X²_0.05N, the interspecific association is significant (P < 0.05). Conversely, if X²_0.95N < W < X²_0.05N, the interspecific association is not significant (P > 0.05) (Gu et al. 2017, Ludwig et al. 1988, Zhou et al. 2015, Zou et al. 2024).

The searched literature was categorized into four main topics: management strategy, ecological and physiological characteristics, host preference and literature reviews. We further analyzed and compared the number of papers that studied the type of host of two fruit flies. The fecundity data across various hosts of two fruit flies were extracted from chosen papers and standardized by converting the data into the number of eggs laid per female in 24 h. Finally, the niche breadth and niche overlap of B. cucurbitae and B. tau were calculated using Levins niche breadth index, Simpson diversity index, and Pianka niche overlap index.

The niche breadth calculations for B. cucurbitae and B. tau utilized the niche breadth by Levins (Levins 1968, Regmi et al. 2022): $${\mathrm{B}}_{\mathrm{i}}=\frac{1}{\mathrm{r}{{\displaystyle \sum }}_{\mathrm{j}=1}^{\mathrm{r}}{\mathrm{P}}_{\mathrm{i}\mathrm{j}}^2}$$where B_i = the niche breadth of species i, j = the quadrat, P_i indicates the proportion of individuals consuming the i type of resource within the total population. A higher value of the Levins’ ecological niche breadth index signifies greater variation in resource utilization by the species within the niche, while a lower value suggests less variation (Song et al. 2024).

Host diversity was described by Simpson diversity (Simpson 1949): $$D=1-{\displaystyle \sum }{\mathrm{P}}_i^2$$where, D = the Simpson diversity index, and the value of the Simpson index ranges from 0 to 1. Over all i, where P_i represents the proportion of each category i. The closer the value is to 0, the higher the diversity of the community, because the distribution of individuals among different species is more uniform. On the contrary, the closer the value is to 1, the lower the diversity of the community, which indicates that most individuals belong to a few species (i.e., there are dominant species) (Measey et al. 2017).

Niche overlap was described by Pianka niche overlap (Pianka 1973): $${\mathrm{Q}}_{\mathrm{i}\mathrm{k}}=\frac{{{\displaystyle \sum }}_{\mathrm{j}=1}^{\mathrm{r}}{\mathrm{P}}_{\mathrm{i}\mathrm{j}}{\mathrm{P}}_{\mathrm{k}\mathrm{j}}}{\sqrt{{\left({{\displaystyle \sum }}_{\mathrm{j}=1}^{\mathrm{r}}{\mathrm{P}}_{\mathrm{i}\mathrm{j}}\right)}^2{\left({{\displaystyle \sum }}_{\mathrm{j}=1}^{\mathrm{r}}{\mathrm{P}}_{\mathrm{i}\mathrm{j}}\right)}^2}}$$where, Q_ik represents the ecological overlap value between species i and species k. P_ij and P_kj, respectively, denote the proportion of individuals of species i and species k in resource state j relative to the total number of individuals of species i and species k. r is the total number of units in the resource set. The value of Q_ik ranges from 0 to 1, with a higher Q_ik indicating a greater ecological niche overlap between the two species (Bewick et al. 2015).

In the host preference tests, the number of eggs laid by B. cucurbitae and B. tau on different hosts did not follow a normal distribution. A general linear model (GLM) was used for the analysis, and the fly species and host types were considered as the predicted variables, and assumed a Poisson distribution. The Tukey’s tests were used for multiple comparisons. The calculation method of niche breadth and overlap was the same as that proposed above.

All data analyses were completed in R 4.3.3 (R Core Team 2014).

Species association

The number of fruit flies captured was affected by the types of fruit flies (χ² = 13.48, df = 1, P < 0.0001, Fig. 1A) and the habitat type (χ² = 8.36, df = 1, P < 0.05). The number of B. cucurbitae captured was higher than that of B. tau in both habitats, and the number of both fruit flies captured in the melon farmlands was higher than those captured in the fruit orchards (Fig. 1A).

View Figure

• Download as Powerpoint
• Download Figure

In both melon farmlands and fruit orchards, there was a negative association between B. cucurbitae and B. tau, indicated by a Variance Ratio (VR) of 0.30 (< 1.00). The W statistic for melon farmlands was 7.84, and for fruit orchards it was 10.78. Both values exceeded the critical values for the 95% confidence interval (X²_0.95N < W < X²_0.05N).

Ecological niche breadth and overlap

A total of 205 articles were collected, including 154 studies on B. cucurbitae and 51 studies on B. tau. For B. cucurbitae, the proportion of research on management strategy was the highest, reaching 41.56% (Fig. 1B). The proportions for ecological and physiological characteristics, host preference, and literature review were 40.26%, 12.99%, and 5.19%, respectively. For B. tau, the highest proportion of research was focused on management strategy, at 46.77%, while the proportions for host preference, ecological and physiological characteristics, and literature review were 24.19%, 19.35%, and 9.68%, respectively (Fig. 1C).

Regarding host preference, M. charantia was the most frequently mentioned host for B. cucurbitae, followed by Cucurbita spp., C. sativus, and L. cylindrica (Fig. 2A). The most frequently reported ovipositional host was L. siceraria, followed by C. melo, L. cylindrica, and C. sativus (Table 1). For B. tau, C. sativus and Cucurbita spp. were the most frequently mentioned hosts, followed by M. charantia and L. cylindrica (Fig. 2A). The most frequently reported ovipositional host was C. melo, followed by M. charantia, L. cylindrica, and S. lycopersicum for B. tau (Table 1).

View Figure

• Download as Powerpoint
• Download Figure

Table 1.Mean (±SD) number of eggs laid by B. cucurbitae and B. tau on different hosts. These data were extracted from selected papers and converted to the number of eggs laid per 24 h.

View Fullscreen

Based on data extracted from the literature, we calculated the niche breadth for B. cucurbitae and B. tau, and their niche overlap. The niche breadths for B. cucurbitae and B. tau were 7.50 and 7.42, respectively. The host diversity index for B. cucurbitae and B. tau was 0.87 for both. Additionally, the niche overlap between the two fruit flies was 0.89.

Host preference tests

The number of eggs laid by B. cucurbitae and B. tau was affected by host types (χ² = 5961.20, df = 9, P < 0.001), but not by the types of fruit flies (χ² = 2.80, df = 1, P = 0.093, Fig. 2B). For B. cucurbitae, a significantly higher number of eggs was laid on C. sativus, Cucurbita spp, and C. pepo than on other hosts. The number of eggs laid on M. charantia, C. pepo, L. siceraria, and L. cylindrica was also significantly higher than that on M. spp. No eggs were laid on C. sinensis. For B. tau, a significantly higher number of eggs was laid on Cucurbita spp. and C. pepo than on other hosts, followed by C. sativus and L. cylindrica, which have a significantly higher number of eggs than M. charantia, L. siceraria, and Momordica spp. No eggs were laid on C. sinensis (Fig. 2B).

The niche breadths for B. cucurbitae and B. tau were 4.95 and 4.44, respectively. The host diversity index for B. cucurbitae and B. tau were 0.80 and 0.77, respectively. Additionally, the niche overlap between the two fruit flies was 0.95.