1
Host plant relationships, temperature tolerance and mating compatibility studies in Bactrocera cucurbitae and Ceratitis rosa
(Contract No. 16072/R0)
Part of IAEA/FAO Co-ordinated Project: Resolution of Cryptic Species
Complexes of Tephritid Pests to Overcome Constraints to SIT
Application and International Trade
2ndPROGRESS REPORT
icipe
P. O. Box 30772, 00100 GPO, Nairobi, Kenya
Chief Scientific Investigator: Dr. S. Ekesi
June 28, 2013
Program of work
The contract was awarded to cover the following program of work:
- Establish Ceratitis rosa colonies and share materials with relevant CRP participants.
- Assess mating compatibility between lowland and highland populations of C. rosa.
- Assess the effect of temperature on development and survival of lowland and highland populations of C. rosa.
- Establish Bactrocera cucurbitae colonies from different hosts and habitats.
- Catalogue B. cucurbitae host plants and carry out host preference studies, preferably under field cage conditions.
During the period under review, the following progress was made with regard to the stated objectives above.
Ceratitis rosa
- Establish Ceratitis rosa colonies and share materials with relevant CRP participants.
Background and rationale. In coastal Kenya, traditionally, Ceratitis rosa has been reared from the following host plants: Psidium guajava (Myrtaceae), Monodoragrandidieri, Uvaria lucida(Annonaceae), Salacia elegans (Hippocrateaceae) and Dryetes natalensis (Euphorbiaceae). All the listed host plants fruits between September and December and effort to establish colonies of the coastal population of C. rosa was delayed due to lack of available host fruits.. However, in October 2012, collection of 1.5 kg of guava fruits from Mwanjamba, Msambweni, Coastal region yielded 21 adult flies which were utilized for colony initiation. Mwajamba is located at 040 18’21”S and 0390 29’88”E, 106 m above sea level.
In the highlands, C. rosa have been traditionally reared from mango.In June 2012, one hundred and two (102) kilograms of mango fruits (variety Van Dyke) were collected from Kithoka, Meru, Central Province and processed in the laboratory for fruit flies. The fruits yielded 29 adult C. rosa and the insects were utilized for colony establishment. Kithoka is located at 00005’59”N, 037040’40”E and 1425 m above sea level.
Materials and methods
Colony establishment. Due to irregular availabilityof fruits, attempts were made to establish colonies of the two populations of C. rosa on solid diet. Eggs of each population of C.rosa were collected using a guava or mango dome (that had the pulp and the seed removed) that was placed into fly stock colonies for 1 h. Eggs were carefully removed from the underside of the dome with a fine camel hair brush and placed on a 9-cm diameter moist blotting paper. After 36 h, 100 newly emerged larvae from the above lots of eggs were gently introduced with a fine camel’s hair brush onto the surface of 100 g of carrot-based diet (Table 1) (Ekesi et al., 2007) in open 150 ml plastic cups. This cup was nested on a larger 300 ml plastic cup with sterile sand at the bottom. Mature larvae exited the diet and pupariated in the sand.
An experiment consisted of 5 cups of diet containing 100 larvae, with each cup serving as a replicate. At every generation tested, records were kept on the following (1) larval stage duration, (2) % puparia recovered from diet, (3) weight of puparia, (4) adult emergence, (5) fecundity and (6) fertility over a 10-d period. Percent pupal recovery was calculated based on the initial number of larvae introduced into each container of rearing medium. Pupal weight was based on four lots of 20 puparia from each replicate. Adult emergence was based on four lots of 20 puparia from each replicate that were placed in screened 12 cm diameter plastic containers and observed for a period of 21 d. Fecundity and fertility were based on daily egg collections from 10 pairs of flies held after a pre-oviposition period of 7 d. Eggs were collected using a mango or guava dome and hatch rate was assessed after 72 h. In all experiments adults were fed on a diet consisting of 3 parts sugar and 1 part enzymatic yeast hydrolysate ultrapure (USB Corporation, Cleveland, OH), and water on pumice granules. All experiments were carried out in a room maintained at 28 1C, 50 8% RH with a photoperiod of 12:12 (L:D).
Results and discussion.There were significant differences between the two populations of C. rosa evaluated for the following quality control parameters:larval development (F=19.87; 1,40; P=0.0001), % pupal recovery (F=28.90; 1,40; P=0.0001), adult emergence (F=74.11; df=1,40; P=0.0001) and fertility (F=18.19; df=1,40; P=0.0001). Pupal weight (F=2.17; df=1,40; P=0.6541) and fecundity (F=2.13; df=1,40; P=0.6543) did not differ significantly across the two populations.
Significant differences were also observed for the following quality control parameters over the five generations of C. rosa: larval development (F=78.11; 4,40; P=0.0001), % pupal recovery (F=76.12; 4,40; P=0.0.0001), adult emergence: F=81.01; df=4,40; P=0.0989),fecundity (F=52.54; df=4,40; P=0.0001) andfertility (F=60.22; df=4,40; P=0.0001). Pupal weight (F=1.06; df=4,40; P=0.8865)wasnot affected by generation of rearing.
Table 1. Mean larval duration (± SE) and % pupal recovery (± SE) of two populations of adult C. rosa reared on carrot-based artificial diet
______
Larval developmental period (d)% Pupal recovery
______
Gene-HighlandLowlandHighlandLowland
rationC.rosaC. rosaC. rosaC. rosa
______P 22.4 ± 0.8a 18.2 ± 0.5b 30.5 ± 1.2a 15.8 ± 1.1b
121.8 ± 1.1a18.7 ± 1.2b34.6 ± 1.8a20.2 ± 0.8b
222.4 ± 0.6a18.5 ± 0.1b50.4 ± 1.5a30.2 ± 1.8b
320.1 ± 0.5a16.2 ± 1.1b64.3 ± 1.2a38.8 ± 2.4b
420.2 ± 0.4a16.8 ± 0.7b66.4 ± 2.1a42.6 ± 2.4b
______
For each quality parameter, means within a row followed by the same letter do not differ significantly by t test (P=0.05).
Table 2. Mean pupal weight (± SE) and adult emergence (± SE) of two populations of C. rosa reared on carrot-based artificial diet
______
Pupal weight (mg) % adult emergence
______
Gene-HighlandLowlandHighlandLowland
rationC. rosaC. rosaC. rosaC. rosa
______
P8.8 ± 1.1a 9.1 ± 0.5a70.6 ± 2.1a46.2 ± 1.8b
18.7 ± 0.6a 8.6 ± 1.1a72.2 ± 1.4a50.2 ± 1.5b
210.3 ± 1.1a10.6 ± 1.2a72.6 ± 2.5a54.6 ± 4.6b
312.2 ± 1.5a11.9 ± 1.6a74.2 ± 3.8a56.2 ± 2.8b
412.4 ± 1.2a12.2 ± 1.8a78.2 ± 1.8a50.8 ± 2.8b
______
For each quality parameter, means within a row followed by the same letter do not differ significantly by t test (P=0.05).
Table 3. Mean fecundity (± SE) and fertility (± SE) of two populations of C. rosa reared on carrot-based artificial diet ______
Fecundity (10-day)Fertility, % egg hatch
______
Gene-HighlandLowlandHighlandLowland
rationC.rosaC. rosaC. rosaC. rosa
______
P78.8 ± 10.4a80.2 ± 8.6a50.4 ± 1.2b36.2 ± 0.8b
174.2 ± 12.6a78.4 ± 14.4a54.4 ± 2.6b34.5 ± 2.6b
298.2 ± 21.1a104.5 ± 18.5a62.1.0 ± 2.4b44.2 ± 4.1b
3102.5 ± 27.4a104.2 ± 17.4a68.4 ± 1.8a40.8 ± 1.8b
4154.2 ± 18.6a149.5 ± 24.2a70.2 ± 1.8a48.9 ± 3.6b
______
For each quality parameter, means within a row followed by the same letter do not differ significantly by t test (P=0.05).
Across generations, larval duration of the highland population were longer (20.1-22.4 d) compared with the lowland population (16.2-18.7 d) (Table 1). Similarly, across generations, % pupal recovery of the highland population was higher (30.5 to 66.4%) compared to the lowland population (15.8 to 42.6%) (Table 1). In both populations, pupal recovery increased from parent generation to the fourth generation.
Pupal weight was not affected by population or generation but became heavier with increasing generations (Table 2). Significantly higher adults emerged from the highland population (70.6 to 78.2%) compared with the lowland population (46.2 to 56.2%) (Table 2).
Ten day fecundity was not affected by population but increased with increasing generations of rearing (Table 3). Across generations, egg fertility in the highland population was significantly higher (50.4 to 70.2%) than that of the lowland population (34.5 to 48.9) (Table 3).
Both populations of C. rosa can be successfully reared on artificial carrot-based diet but the highland population appears to be easily amenable to artificial rearing than the lowland population. Generally, laboratory colonization and mass production of fruit flies on artificial diet may require several generations for the insects to adapt to the artificial diet (Kamikado et al. 1987, Souza et al. 1988, Economopoulos 1992). It is therefore not surprising that quality control parameters of the highland population differ from that of the highland population. For example, Economopoulos (1992) showed that even after nine generations, fecundity of wild C. capitata did not match the levels of laboratory-adapted flies. Clearly, the lowland C. rosa population require a longer period to adapt to the medium but long term monitoring is required to see how long it will take for this population to reach the plateau of the highland population.
- Assess the effect of temperature on development and survival of lowland and highland populations of C. rosa.
Background and rationale. In Kenya, Ceratitis rosa has been observed to split up into two different groups (the lowland and highland populations) which can be differentiated genetically as well as morphologically (at least in the males). It is probable that they might also both have different physiological patterns with one entity being more cold tolerant than the other and with the potential for varying invasive powers. In this regard we propose to establish colonies of the two populations of C. rosa from known host plants of the insect in Kenya (Copeland et al., 2006). Colony establishment will follow the same methodology as previously described. The different populations will be shared with participants in the CRP as required.
Materials and methods
Insect culture.Two C. rosa populations were used for this experiment and initial stock culture of highland population originated from infested mango fruits collected at a smallholder farm in Meru and the larvae were subsequently reared on a carrot-based artificial diet (hereafter referred to as diet) in the laboratory. The lowland population originated from fallen guava fruits collected from a smallholder farm in Mwamjamba, Kenya and also subsequently reared on carrot-based diet. Both colonies were maintained for 5 generations on the artificial diet at the Animal Rearing and Containment Unit (ARCU) of the International Centre of Insect Physiology and Ecology (icipe), Nairobi, Kenya before commencement of the experiments. Rearing conditions are maintained at 28 ± 10C, 50 8% RH and photoperiod of L12: D12.
Egg collection.Eggs of C. rosa were collected from the stock colony by offering to mature female flies, ripe mango or guava dome (fruit skin that has the seed and pulp scooped out). The domes were placed over a 9 cm diameter Petri dish lined with moistened filter paper. Domes were maintained in 30 x 30 x 30 cm perspex cage at 28 ± 10C , 50 8% RH. Each dome was pierced with an entomological pin (38mm long, 0.3mm diameter) to facilitate oviposition. Eggs were collected within 1 h of oviposition using a moistened fine camel’s hair brush.
Effect of temperature on development and survival of C. rosa. Egg stage: 100 eggs were counted and carefully lined on a rectangular piece of moistened sterilized black cloth in a Petri dish. The Petri dishes were immediately transferred to thermostatically controlled environmental chambers (MLR-153, Sanyo, Japan) set at 5 constant temperatures of 10, 15, 20, 25, 30, 33 and 350C (± 10C) and 50 8% RH, 12:12 L:D photoperiod. Duration of egg stage was observed at 6-hourly intervals under a binocular microscope for determination of egg hatch.
Larval stage: 100 first instar larvae (~1 h old) of each population were counted and carefully transferred on a cellulose sponge in a Petri dish and thereafter placed on top of a 50 g of diet on a 5 x 5 x 3 cm plastic container. Twenty four hours after the larvae have settled in the diet, the tray was transferred in to larger rectangular plastic rearing containers (7 x 7 x 5 cm) carrying a thin layer (~ 0.5 cm) of moist sterilized sand at the bottom for pupation and into the environmental chambers. The top of the plastic containers were screened with light cloth netting material for ventilation. The containers were then maintained at the same constant temperature in the environmental chambers. Normally, mature late third instar larvae leave the Petri dishes containing the artificial diets ad libitum and jump into the sand in the larger containers to pupate. Starting at 5 days after the larvae were exposed to artificial, the number of puparia in the sand was recorded daily by sifting and record of larval duration were kept for each population at each temperature regime.
Pupal stage: One hundred newly formed pupae (~ 1 h old) were obtained from the culture and held in small-ventilated transparent cylindrical plastic cages (5.5 x 12.5 cm). The pupae were transferred to the same 7 constant temperatures and records were kept of duration to adult eclosion.
Statistical analysis
General analysis: To examine the effects of temperature on life history parameters, data for developmental time and survivorship was subjected to two-way analysis of variance (ANOVA). Log10 ± 0.5 and arcsine square root transformation were used respectively, on counts and percentages before statistical analyses (Sokal and Rohlf, 1981). When treatment effects were significant, means were separated using Student-Newman-Keul’s (SNK) test.
Linear model: The linear portion of the developmental rate curve [R(T)=a + bT] was modeled using regression analysis where T was temperature, and a and b were estimates of the intercept and slope, respectively. The lower temperature threshold (Tmin) was estimated by the intersection of the regression line at R(T) = 0, T0 = -a/b. Degree-day (DD) requirements (thermal constant, K) was calculated using the inverse slope of the fitted linear regression line (Campbell et al. 1974).
Nonlinear model: The nonlinear relationship between developmental rate r(T) and temperature T was fitted to the Brière model, which allows the estimation of the upper and lower developmental thresholds (Brière et al. 1999). Brière-1 model was used and described as:
1/D=a x T x (T – T0) x Sqrt(TL – T)
Where T0 (tmin) is the lower threshold, TL(tmax) the lethal temperature is the upper threshold and a is an empirical constant.
The following statistical items were used to assess the goodness-of-fit: the coefficient of determination (for linear model; R2) or the coefficient of nonlinear regression (for nonlinear models; R2) and the residual sum of squares (RSS). Higher values of R2 and lower values for RSS reveal a better fit. For the linear regression, the data points at 350C which deviated from the straight line through the other points were rejected for correct calculation of regression (Campbell et al. 1974). The model fitting was implemented in R 2.13.1 (R Development Core Team, 2011) using the nls function.
Results and discussion
Effect of temperature on stage development. The time required for eggs to hatch ranged from 7.83 days at 150C and decreased to 1.75 days at 330C (F = 29.1.2, d.f = 4, 15 P=0.0001) in the highland population. In the lowland population, egg development was also longest (7.54 days) at 150C and shortest (1.63 days) at 330C (F=42.98; df=5,18; P=0.0001) (Table 4). At 350C, eggs of the highland population of C. rosa, failed to hatch. However, egg developed at 350C in the lowland population and took 1.75 days to hatch. In both populations, eggs did not develop at 100C.
The linear regression model showed a strong positive linear relationship between temperature and egg development rate for highland (R2 =0.98) and lowland (0.94) populations (Fig 1) with a lower development threshold of 10.60C and 10.10C for highland and lowland populations, respectively. The egg stage required 38 degree-days (DD) to complete development in the highland population and 34 DD in the lowland population. For the highland population, the parameter estimates for the Brière-1 nonlinear model predicted the lower temperature threshold of 11.10C and the upper temperature threshold of 34.80C. The rate of development increased with temperature until the curve reached an optimum and then decreased rapidly as temperatures reached the upper temperature threshold (Fig. 1).
At larval stage, the trend was similar to the egg stage with development periods decreasing from 24.91 days at 150C to 9.50 days at 350C (F = 56.63, d.f = 4, 15 P= 0.0001) for highland population; and 22.9 at 150C to 7.69 at 350C (F = 113.96, d.f = 5, 18 P= 0.0001) in the lowland population (Table 4). At 350C, larvae did not develop in the highland population. Also at 100C, no development occurred in the both populations (Table 4). At 300C, the lowland population developed faster than the highland population (Table 4).
The linear regression between temperature and development rate for this stage was positive for the highland population (R2 = 0.94) as well as the lowland population (R2=0.97) (Fig. 1). Highland C. rosa required 203 DD above the development threshold of 10.10C to complete development from larval stage to the pupal stage. The lowland population took 177 DD to develop above a threshold of 10.30C. Parameter estimates for the Brière-1 nonlinear model predicted the lower and upper temperature threshold of 10.40C and 35.00C for the highland population. For the lowland population, Brière-1 nonlinear model estimated lower and upper developmental threshold of 10.60C and 40.80C, respectively.
In the highland and lowland populations, temperature had a significant effect on development of pupae: (F =149.45, d.f = 3, 12 P< 0.0001) and (F = 385.43, d.f = 3, 12 P= 0.0001), respectively. In both populations, the longest duration occurred at 150C: highland =35.42 days and lowland = 34.25 days (Table 4). There was no eclosion at 10, 33 and 350C in both populations (Table 4).
The linear regression between temperature and development rate for this stage was positive for both populations: Highland - R2 = 0.98 and lowland - R2 = 0.96 with a lower development threshold of 9.6 and 10.00C, respectively (Fig.1). The puparium required 228 and 182 DD to complete development in the highland and lowland populations, respectively. Parameter estimates for the Brière-1 nonlinear model predicted the lower and upper temperature threshold of 9.980C and 35.00C for the highland population. For the lowland population, Brière-1 nonlinear model estimated lower and upper developmental threshold of 10.10C and 35.00C, respectively.
In both populations, total developmental duration was longest at 150C (Highland =68.16 days; Lowland=64.69 days) and shortest at 300C (Highland=22.64 days; Lowland=19.76 days) (Table 4).
Survival rates. At egg stage, survival ranged between 43.5% at 330C to 63.5% at 250C (F = 4.96, d.f = 4, 15 P = 0.0095) in the highland population and 72.0% at 350C to 91.5% at 250C in the lowland population (F = 5.42, d.f = 5, 18; P = 0.0033) (Table 5).
In the highland population, survivorship at larval stage ranged between 35.75% at 330C to 77.00 % at 250C (F = 14.36, d.f = 4, 15 P= 0.0001). In the lowland population, survival rate was lowest (48.25%) and highest (89.25%) at 250C (F = 24.24, d.f = 5, 18 P= 0.0001) (Table 5).
During the pupal stage, survival ranged from 61.25% at 300C to 94% at 250C in the highland population (F = 22.76, d.f = 3, 12 P= 0.0001). In the lowland population, survival ranged from 75.5% at 300C to 95.5% at 250C (Table 5).
The ranges in duration of immature stages are in general agreement with those reported for other Ceratitis species (Delrio et al., 1986; Fletcher, 1989; Vargas et al., 1996; Duyck and Quilici, 2002). Duyck and Quilici (2002) published the first report on development of C. rosa at temperatures of 15-350C. The authors reported total development in days of 18.8–65.7 days. In our study, total development of the highland population of C. rosa took 22.6-68.2 days while lowland population took 19.8-64.7 days. Results from the lowland population therefore closely mirrored those reported for populations of C. rosa from La Reunion. Values for the temperature threshold and thermal constant were also consistent with previous studies. Duyck and Quilici (2002) reported lower developmental thresholds for the egg, larval and pupal stages as 9.8, 3.1, and 11.0, respectively. Apart from the larval stage, the estimated thresholds from our study are generally in agreement with these authors for both populations. No previous studies are available in literature with regard to upper developmental threshold for C. rosa. Physiologically, Brière-1 nonlinear model predicted that both populations seem to have similar lower temperature thresholds but vary in their tolerance to higher temperatures. In our study, these values ranged from 35.0-40.80C across the developmental stages in the lowland population but was exactly at 35.00C for the highland population. Clearly, the lowland population tolerates higher temperatures than the highland population.