The silkworm, Bombyx mori L. has a long history of domestication as an economically important insect. Intensive and careful domestication over centuries has apparently deprived this commercial insect of the opportunity to acquire thermotolerance (Kumari et al., 2011). This vulnerability is more pronounced in bivoltine breeds compared to multivoltine breeds. Thus, among many factors responsible for poor performance of the bivoltine breeds under tropical conditions, the major one is lack of thermotolerance (Kumari et al., 2011). Many quantitative characters decline sharply at higher temperature (Krishnaswami et al., 1973). Therefore, high temperature is one of the main factors for bivoltine crop failure in the tropics (Kumar et al., 2001).
Sericulturists from the tropical countries like India relay on climatic conditions prevailing in the field and restrict themselves to rear silkworm breeds suitable for high temperature conditions. This prompts the commercial exploitation of multivoltine x bivoltine hybrids in India, as they are hardy and ably to survive and reproduce under fluctuating climatic conditions (Kumari et al., 2011). The quantity of silk produced by multivoltine x bivoltine hybrids meet the domestic needs but the quality of silk is low compared to bivoltine silk of international grade. High quality, internationally recognised silk production can be achieved only by rearing bivoltine breeds (Begum et al., 2001). To overcome this drawback, compatible bivoltine hybrids for rearing under tropical conditions were developed (Datta et al., 2001; Kumar et al., 2002, 2006; Lakshmi et al., 2011) and selected for rearing in field conditions. However, the attempts to spread bivoltine silkworm breeds / hybrids throughout the sericulture belt of India resulted in extensive crop loss, especially in the hot and humid climatic conditions of tropics (Chavadi et al., 2006). These productive hybrids were vulnerable to varied environmental conditions, as they originated from temperate regions (Kumari et al., 2011).
Performance of each organism is determined and affected by genotype, environment and interactions between them. Therefore, individuals with high genetic potential must be selected in order to improve the quantity and quality of silk production, besides improving environmental conditions (Nezhad et al., 2010). Knowledge of genetic diversity and genetic relationships between different pure lines can be a valuable factor to improve silkworm breeding strategies. The selection of divergent parents present in the breeding stocks forms the base for any breeding program (Vijayan et al., 2010). Genetic diversity in silkworm is generally measured based on the phenotypic traits that have been used for differentiation of silkworm genotypes because most of the traits are either qualitative or quantitative (Bindroo and Moorthy, 2014). However, molecular markers are other types to estimate genetic diversity, which can provide a rapid means of analysis of genetic diversity in germplasm (Bindroo and Moorthy, 2014). Anyhow, many important phenotypic traits in silkworms decline above 28℃ (Krishnaswami et al., 1973). Therefore, the estimation of phenotypic changes at high temperature is considered of great importance for sustainable progress in bivoltine breeding (Kumari et al., 2011).
Euclidean or straight-line measure of distance is the most commonly used statistics for estimating genetic distance between individuals by morphological data (Mohammadis and Prasanna, 2003). Numerous studies represent the hierarchical clustering of different silkworm breeds based on their performances with respect to quantitative traits. The economically important quantitative traits like fecundity, larval weight, yield/ 10000 larvae by weight, yield/ 10000 by number, shell weight, cocoon weight, shell %, filament length, filament size, neatness, etc., have been traditionally used for characterization / diversity studies (Thangavelu et al., 2000). Applying D2 (Mahalonobis’ distance) statistics, Kumaresan et al. (2007) categorized 58 multivoltine silkworm genotypes traits into nine clusters using ten economic. Nezhad et al. (2010) classified all 51 pure lines of Iran silkworm germplasm based on larval traits into two distinct groups using Ward’s method and UPGMA (Unweighted Pair-group Method using Arithmetic mean) method. Pal and Moorthy (2011) studied variability in larval and cocoon traits in 19 bivoltine silkworm breeds and reported their division into three distinct clusters. Such classification is also important for breeding work aimed at improving adaptability and other yield components. The aim of this study was to identify high temperature tolerant bivoltine breed and an attempt was also made to classify 20 silkworm breeds based on the rearing performance under high temperature conditions using Euclidean distance.
Materials and Methods
Silkworm breeds and thermal treatment
Twenty silkworm breeds consisting of two multivoltine (Nistari and Cambodge) and eighteen bivoltine breeds (CSR2, CSR17, S5, D13, 5HT, 8HT, BHR2, BHR3, D6(P), SK3, SK4, D6(P)N, SK4C, B37, S38, NN6D, ATR16, ATR29) were used for the study. Multivoltine breeds were considered as reference to compare with the performance of bivoltine breeds. Silkworm rearing was conducted following the standard method under recommended temperature and humidity conditions till 2nd d of 5th instar as suggested by Krishnaswami et al. (1973). With regard to thermal exposure, the 3rd d 5th instar larvae were reared in a SERICATRON (Environment chamber with temperature and humidity control) at three different temperatures viz., 32 ± 1℃, 34 ± 1℃ and 36 ± 1℃ with an exposure duration of six h (10:00-16:00) per d. Fresh mulberry leaves were given twice a day and silkworm rearing was continued as suggested by Kato et al. (1989) and Kumar et al. (2001). The larvae reared at 25 ± 1℃ under normal conditions were considered as control. The experiment was performed in triplicate for each temperature treatment with 100 larvae per replicate. The ripened larvae were mounted on plastic mountages. After 48 h of mounting, when the larvae formed hammock, the mountages were turned upside down. Cocoon harvesting was carried out on the 7th d of spinning. Cocoons were deflossed and the defective ones were sorted out. Assessment of all the breeds was carried out on the same day of development.
During silkworm rearing, data on six quantitative traits were collected separately for control and each treatment batch (32 ± 1℃, 34 ± 1℃ and 36 ± 1℃). The traits like pupation percentage (%), larval weight (g), cocoon yield for 10,000 larvae by weight (kg), cocoon weight (g), shell weight (g) and shell percent (%) were noted. The overall mean of each trait was calculated for each breed at all four temperatures. Pupation percentage was estimated as live pupa present inside the cocoon during metamorphosis of larva into pupa expressed as a percent. The percentage change in the performance of the breeds in treated over control with respect to their pupation percentage was calculated. This genetic trait was considered as the measure of index of thermotolerance in silkworm. Percentage change / reduction in the pupation percentage was calculated as follows (Kumari et al., 2011)
Hierarchical clustering was done by using the rearing data of twenty silkworm breeds at four temperatures. Euclidean distance was calculated as a measure of genetic diversity by using the rearing parameters of the twenty silkworm breeds at 25 ± 1℃, 32 ± 1℃, 34 ± 1℃ and 36 ± 1℃. Based on Euclidean distance dendrogram was constructed using Ward’s method with a bootstrap value of 10000. Analysis of variance was performed on pupation percentage, larval weight, cocoon yield for 10,000 larvae by weight, cocoon weight, shell weight and shell percent of silkworm breeds using temperature as factor. All the analysis was done with SPSS software package.
Eighteen bivoltine and two multivoltine breeds were considered for the study. Variations in the larval marking, cocoon shape and cocoon colour were observed in experimental breeds. Among multivoltines, Nistari and Cambodge spun yellow coloured and oval elongate cocoons. Nistari larvae are marked and Cambodge larvae are plain. Among bivoltines, ATR16, B37, CSR2, CSR17, 8HT, S5, NN6D, S38 and SK3 larvae were plain and spun oval shaped cocoons. Other bivoltines viz., ATR29, BHR2, BHR3, 5HT, D13, D6(P), D6(P)N, SK4 and SK4C larvae were marked and spun dumbbell shaped cocoons. All the bivoltines spun white coloured cocoons.
Overall performance of silkworm breeds under normal and high temperature treatments
Six important rearing traits viz., pupation percentage (%), larval weight (g), cocoon yield, cocoon weight (g), shell weight (g) and shell percent (%) were analysed in twenty silkworm breeds by exposing them to temperatures at 32 ± 1℃, 34 ± 1℃ and 36 ± 1℃ (Treated) against 25℃ ± 1℃ (Control). Differences in the studied parameters over temperature suggested the response of each breed to high temperature. Generally the control larvae performed better than the treated. In order to know the breed that is performing better in tested temperatures, the overall mean of particular trait of a breed at all temperatures (25℃ ± 1℃, 32 ± 1℃, 34 ± 1℃ and 36 ± 1℃) was considered. The performance of the breeds under normal and treated temperatures and their overall averages are presented in Tables 1-5. Among bivoltines, average pupation percentage (indicator of thermotolerance) was high in SK4C (73.34%) followed by BHR3 (73.05%) and low in CSR2 (42.02%) followed by B37 (43.53%) with 86.82% and 83.53% for Nistari and Cambodge, respectively, which were used as reference for thermotolerance. The rearing traits studied were low in multivoltines than bivoltines; therefore the comparison has been made only among bivoltines. The overall mean for weight of ten larvae was highest in CSR2 (38.39 g) and lowest in D6(P)N (28.05 g) followed by NN6D (28.5 g). Highest average cocoon yield per 10,000 larvae by weight was demonstrated in S5 (9.91 kg) followed by 8HT (9.69 kg) and lowest was noted in B37 (5.28 kg). The highest average cocoon weight and shell weight of 1.57 g and 0.32 g were observed in CSR17 and lowest of 1.28 g and 0.22 g in B37. Average shell percent was highest in CSR2 (20.38%) and lowest in BHR2 (16.17%) followed by NN6D (16.48%). Significant differences were noticed in all the studied rearing traits of silkworm breeds with regard to temperature (Table 6).
Table 1.Rearing performances of twenty silkworm breeds at 25 ± 1℃
Table 2.Rearing performances of twenty silkworm breeds at 32 ± 1℃
Table 3.Rearing performances of twenty silkworm breeds at 34 ± 1℃
Table 4.Rearing performances of twenty silkworm breeds at 36 ± 1℃
Table 5.Overall rearing performances of twenty silkworm breeds at four temperatures
Table 6.ANOVA on rearing traits with temperature as a factor
Percentage reduction in pupation percentage of treated groups over control
The negative sign in the percent changes over control in treated groups indicate a reduction over control and the value near to zero suggests a better performance of the breed. At all the treated temperatures, among multivoltines, the highest reduction in pupation percentage was in Cambodge against Nistari (Table 7). At 32 ± 1℃, 34 ± 1℃ and 36 ± 1℃ reduction in pupation percentage in Cambodge was 11.14%, 14.66% and 19.5% over control, respectively. Similarly, reduction in pupation percentage in Nistari was 9.08%, 10.13% and 15.04% at 32 ± 1℃, 34 ± 1℃ and 36 ± 1℃ over control, respectively. Among bivoltines, highest reduction of 65.15% in pupation percentage was demonstrated in CSR2 under 32 ± 1℃ and lowest reduction of 21.31% in SK4C. At 34 ± 1℃, B37 was highly affected by highest reduction in pupation percentage of 74% followed by CSR2 with 67.15%. At 36 ± 1℃, CSR2 had highest reduction in pupation percentage of 86.53% followed by B37 with 85.78%. BHR3 was more stable at 34 ± 1℃ and 36 ± 1℃ with lowest reduction in pupation percentage of 26.1% and 34.24%, respectively followed by SK4C with 26.7% and 35.76%, respectively (Table 7).
Table 7.Reduction in pupation percentage in silkworm breeds at high temperatures over control
Pairwise Euclidean distance was estimated between silkworm breeds based on their rearing traits under normal and treated temperatures. Using Euclidean distance as measure of diversity the dendrogram was constructed by Ward’s method. Twenty silkworm breeds were broadly divided into four clusters. Cluster I contained SK4C, BHR3, Nistari and Cambodge, which were thermo tolerant with pupation percentage of >70%. Cluster II was a large cluster with 8 silkworm breeds viz., 8HT, S5, CSR17, 5HT, ATR29, BHR2, NN6D and D6(P)N with pupation percentage of more than 60% and less than 70% except CSR17, which has a pupation percentage of 57.86%. Cluster III had 6 silkworm breeds namely, SK4, D6(P), SK3, S38, ATR16 and D13, and the pupation percentage of the members ranged between 51.48 and 58.58%. Cluster IV had 2 silkworm breeds namely, B37 and CSR2 with lowest pupation percentage of 43.53% and 42.02%, respectively (Fig. 1).
Fig. 1.Phylogenetic relationship of silkworm breeds based on six rearing traits at four different temperature regimes.
The genetic distance ranged between 5.49 (between SK4C and BHR3) and 109.84 (between Nistari and CSR2) with an average of 36.74. In cluster I, four silkworm breeds were classified with an average genetic distance of 29.8 with 37.51 being highest between Nistari and SK4C and 5.49 being lowest between SK4C and BHR3. Similarly, in cluster II, eight silkworm breeds were divided with an average genetic distance of 17.63 with 30.92 being highest between CSR17 and NN6D and 9.78 being lowest between D6(P)N and NN6D. Silkworm breeds in cluster III were differentiated with an average genetic distance of 19.14 ranging from 10.58 between ATR16 and D13 to 28.21 between ATR16 and SK4. The two silkworm breeds, CSR2 and B37 formed cluster IV, which were distanced by a genetic distance of 31.38 (Supplementary material 1).
On increase in the rearing temperature above 25℃, all the breeds showed a decline in rearing parameters which is similar to that reported by Kumar et al. (2002). The decline in the rearing parameters with increase in temperature was probably due to the low feeding activity of the silkworms during high temperature. As a fact, high temperature reduces the moisture content in the leaf through evaporation thereby drying the leaf sooner and therefore not enabling to feed by silkworms (Kumar et al., 2012). Water is an essential requirement for metabolic activity and optimum growth of silkworm. At higher temperature, evapo-transpiration at body surface and respiratory epithelium of tracheal system significantly increases. The problem of water balance in silkworm at ambient temperature is further complicated by poor moisture content of the leaf, which finally affects the growth and productivity of silkworm (Rahmathulla, 2003). Moreover, heat stress affects biologically important molecules like DNA, RNA, lipids and halts normal protein synthesis mechanism. It even causes unfolding of the typical folded functional cellular proteins making it non-functional (Feder, 1996). All these changes tend to increase the vulnerability of the silkworm to other biological processes resulting in the decline in its performance (Kumari et al., 2011).
Abiotic and biotic factors have a considerable impact on the success of sericulture industry and thus they are of great importance (Kumar et al., 2012). Among the abiotic factors, temperature plays a key role in growth and productivity of the silkworm, as it is a poikilothermic insect (Benchamin and Jolly, 1986). It is a well known fact that the late age silkworms prefer relatively lower temperature than young age and fluctuation of temperature during different stages of larval development is more favourable for growth and development of larvae than constant temperature. There is ample literature stating that good quality cocoons are produced within a temperature range of 22-27℃ and above these levels the cocoon quality becomes poorer (Krishnaswami et al., 1973) indicating that studies on cocoon traits is basis for appropriate selection for required environment. The effect of high temperature in terms of cocoon crop depends on several factors that operate within and outside the body of the silkworm.
It is a difficult task to assess thermotolerance trait in silkworms as it is governed by genetic and environmental factors (Kumar et al., 2012). The effect of temperature higher than 30℃ on silkworm larvae was reported earlier by Shirota (1992). Tazima and Ohnuma (1995) further confirmed the genetical nature of thermotolerance by selection based on pupation rate of silkworms reared under high temperature conditions in 5th instar. The experiments conducted by Kato et al. (1989) showed that the thermotolerance character in silkworm is heritable. Therefore, it is important to understand the performance of the breeds under a set of different high temperatures. In this study thermotolerance in twenty silkworm breeds was assessed under three high temperatures viz. 32 ± 1℃, 34 ± 1℃ and 36 ± 1℃.
One of the objectives of the silkworm breeder is to recommend stable breeds to the farmers for rearing under different environmental conditions. Silkworm breeds reared over a series of environments exhibiting less variation are considered stable. According to Allard and Bradshaw (1964), performance of the breed itself in a given environment indicates its superiority. As the majority of the economically important genetic traits of silkworms are qualitative in nature and phenotypic expression is greatly influenced by environmental factors such as temperature, relative humidity, light and nutrition (Pillai and Krishnaswami, 1980, Ramesha et al., 2010; Zhang et al., 2002), it is essential to measure the degree of phenotypic difference of the economical traits under varied environmental conditions to understand the genetic steadiness. In the present study, the phenotypic expressions of six traits of economic importance were studied in twenty silkworm breeds exposed to different temperatures viz. 32 ± 1℃, 34 ± 1℃ and 36 ± 1℃. The variations observed in the phenotypic manifestation for the traits measured can be attributed to the genetic constitution of the breeds and their degree of expression to the conditions to which they are exposed during their rearing. Such variations in the manifestation of phenotypic traits of the breeds studied can be ascribed to the influence of environmental conditions. Variable gene frequencies at different loci make them to respond differently. The results of the present study are in line with the findings of Kumar et al. (2012).
The multivoltine breeds were more tolerant to high temperature than the bivoltines. It is well established that multivoltine breeds originated and reared in tropical countries are tolerate slightly higher temperature, as are cross breeds that have been evolved for a tropical climate (Ramesha et al., 2009). In the present investigation, the productive bivoltine breeds like CSR2, CSR17 and SK3 showed low pupation percentage at high temperatures. This was also noticed by Kumar et al. (2002) and Kato et al. (1989) by observing higher survival in the hybrids than the pure races under high temperature conditions. The bivoltine breeds exposed to high temperature revealed a drastic reduction in the pupation rate and cocoon traits suggesting that bivoltines are more prone to high temperature.
In this study, the highest overall means of larval weight and shell percent were noticed in CSR2, while, the highest overall mean of cocoon and shell weights were observed in CSR17. Highest cocoon yield per 10,000 larvae by weight was demonstrated in S5. The current study observed SK4C (73.34%) and BHR3 (73.05%) as two bivoltine breeds tolerant to high temperatures as they showed highest pupation percentage at all the temperatures tested. These breeds not only performed well under different high temperatures but also were stable with low percentage reduction in pupation rate over their controls indicating their ability to withstand high temperatures.
Classification of silkworm breeds based on six rearing parameters at high temperatures
In the current study, the genetic relationship between 20 silkworm breeds was determined by using rearing data of silkworm breeds at four different temperatures. Based on the data Euclidean distance was estimated and dendrogram was constructed using Euclidean distance and Ward’s method. The lowest genetic distance was 5.49 between SK4C and BHR3. These two breeds had the highest pupation percentage at each high temperature as well as overall temperatures tested. They were developed by screening each generation at high temperature conditions and selection criteria for further generation was based on heat-tolerant traits (Moorthy et al., 2007). The highest genetic distance was 109.84 between Nistari and CSR2 that had contrasting characters with respect to cocoon traits and thermotolerance as revealed by this study. Nistari and CSR2 had highest and lowest pupation percentage at four tested temperatures, respectively, which probably might be the factor that separated Nistari and CSR2 with high genetic distance.
Cluster analysis revealed that SK4C and BHR3 (>70%) breeds were grouped together with the multivoltines (>80%), which had highest pupation percentage at all the tested high temperatures including overall means. Similarly, CSR2 (42.02%) and B37 (43.53%) showed an overall mean of lowest pupation percentage and high percentage reduction in pupation rate over control. Thus, CSR2 and B37 form a cluster. The clustering pattern and genetic distance suggests that the breeds were primary grouped based on their ability of thermotolerance epresented by pupation percentage. However, in cluster II, the silkworm breeds had pupation percentage between 60-70% except for CSR17, which had a pupation percentage of 57.86%, typical for cluster III (i.e. 50-60%). This clearly indicates the role of rearing traits other than pupation percentage in classification of silkworm breeds because CSR17 though had a pupation percentage of cluster III members; it shared very nearer values for other rearing traits with the members of cluster II. Therefore, it was grouped in cluster II.
In this study, the breeds reared in Southern India (S38 and D13) clustered with the breeds reared in Northern India (SK4, D6(P), SK3 and ATR16) to form cluster III. According to Chatterjee and Datta (1992) for the silkworm, domestication has played a major role in genetic diversification. As sericultural regions of the world have different climatic conditions, physiological diversification has also been influenced by agro-climatic factors. Thus, given geographic isolation and limited cultural exchange, some breeds of South India and North India inclusive in cluster III may have acquired similar genetics due to similar selection pressures. The above results revealed that the inclusion of silkworm breeds of the same region in different clusters clearly indicate the presence of considerable genetic diversity among the population used in this study. All these cases indicate that clustering on the basis of estimates of phenotype does not always reflect geographical distance, as has also been pointed out by researchers working on the clustering of plant materials (Harlan, 1971). Likewise Spagnoletti and Qualset (1987) pointed out that geographical position does not correspond with the phenotypic grouping for the origin of spike characteristics in Duram wheat. Kumaresan et al. (2007) reported the inclusion of silkworm genotypes of same origin in different clusters by applying D2 (Mahalonobis’ distance) statistics on 10 economical traits of 58 polyvoltine strains. In contrast, Nezhad et al. (2010) reported that 51 pure lines of Iran silkworm germplasm were clustered into different groups according to the geographic areas that were initially observed by using data on larval traits. However, in this study, clustering on the basis of rearing parameters at high temperature conditions separated most tolerant and most susceptible silkworm breeds in distinct clusters but failed to separate the breeds with moderate thermotolerance into distinct clusters. This is evident from the breed CSR17 (57.86%), which was grouped in cluster II but had a pupation percentage of members of cluster III (60-70%). This indicates that more rearing and economically important traits of silkworm have to be assessed at different high temperatures to classify silkworm breeds with moderate thermotolerance into distinct clusters.
Current study reveals a negative relationship between the tested temperature and the rearing parameters of silkworm breeds. SK4C and BHR3 with high pupation percentage at three different high temperatures and less percentage reduction in pupation percentage over controls, were identified as high temperature tolerant and stable bivoltine silkworm breeds over a range of 32 to 36℃. So far in silkworm, phenotypic studies to identify high temperature tolerant bivoltine breeds were carried out with a single high temperature treatment. In the present report, the high temperature tolerant bivoltine breeds were identified based on their rearing performance at three treated temperatures, which a breeder can rely on. Therefore, these bivoltine breeds can be recommended as potential breeding material for the development of thermotolerant silkworm breeds/hybrids. Furthermore, this study also suggests that Euclidean distance estimated on the basis of rearing data of silkworm breeds under high temperature conditions can be an effective approach in classifying the silkworm breeds based on their capacity of thermotolerance. Based on the Euclidean distance 20 silkworm breeds were divided into 4 clusters, which represented one each of thermotolerant, thermo susceptible and two clusters of moderately tolerant silkworm breeds. The utilization of breeds from different clusters for breeding will increase heterosis. These results will enhance the knowledge of the silkworm breeders in developing thermotolerant breeds or hybrids with maximum heterosis.