Antheraea mylitta Drury (Lepidoptera: Saturniidae) is a “tasar silk” producing polyphagous forest insect of India, but it has never been experimented in any Himalayan states of India including Uttarakhand. This is the first study on effect of forest tree species, rearing seasons and their interaction on Silk Gland Weight (SGW) of A. mylitta. Silk gland of A. mylitta is an ectodermal organ, which produces liquid silk that forms cocoon fibre. Silk gland become secretory prior to hatching and increase their secretory potential by growing in size during larval life (Aruga, 1994).

Forestry host plants differ greatly with their nutrient profile (Kohli et al., 1969; Sinha and Jolly, 1971; Agarwal et al., 1980; Sinha and Chaudhury, 1992; Puri, 1994) that has direct influence on health, growth, and survival of A. mylitta (Sinha et al., 1986). Instances of poor growth might not be due to the nutritional inadequacy of the diet but to a low rate of intake due to the absence of a non-nutrient phagostimulant (Dadd, 1960). Lokesh et al. (2012) studied variability in silk gland weight of three ecoraces of A. mylitta viz., Daba, Sukinda and Sarihan by rearing on Terminalia arjuna and found that Daba ecorace showed highest silk gland weight of 8.14 g.

Studies have proven the significance of seasonal variations on biology and development of insects (Odum, 1983; Ouedraogo et al., 1996). Temperature influences everything that an organism does (Clarke, 2003), humidity affects embryonic development (Tamiru et al., 2012), and rainfall affects both (Getu, 2007). Rainfall alters functioning of microhabitat, which along with soil and other environmental factor affect foliage quality that consequently reflect distinct impact on performance of insect (Mattson and Haack, 1987).

In commercial tasar silkworm rearing, larvae should be healthy and their silk glands should be free from any abnormalities to reduce presence of non-spinning silkworms to zero. In the present work, expression of silk gland weight by A. mylitta larvae reared on seven forest tree species in two rearing season has been analyzed.

The study was conducted at New Forest, Forest Research Institute (FRI), Dehra Dun, Uttarakhand, which is situated at 30° 19’ 56.21’’ N to 30° 21’ 5.35’’ N and 77° 58’ 56.81’’ E to 78° 0’ 59.73’’ E at 640.08 AMSL (Fig. 1). The climate of New forest, Dehra Dun is moderate due to its location at foot of the Himalayas (WMD, 2011) and is monsoon-dominated like Gangetic Indian Plain. In the foothills, altitude is the main factor for variation in temperature and vegetation. The climatic variables of the study site is summarised in Fig 2, 3, 4, & 5 and the normal climatic data of Doon Valley for last thirty years is shown in table 1.

Outdoor tasar silkworm rearing was conducted on 5-6 year old bushes of seven forest tree species viz., Lagerstroemia speciosa, L. tomentosa, T. alata, T. arjuna, T. bellirica, T. chebula and T. tomentosa. Four outdoor experimental rearings of A. mylitta, each in July-August (first rearing season) and September-November (second rearing season) were conducted during the year 2012 and 2013 as per rearing technology of Mathur et al. (1998) and data were collected. Three hundred neonate larvae of A. mylitta were allowed to crawl on tender foliage (Fig. 6) of each of the six bushes of all the treatments in the morning hours. After second moult (Fig. 7), net was removed from all the bushes and the late age rearing of third (Fig. 8), fourth (Fig. 9) and fifth instar larvae (Fig. 10) were carried out in open forest conditions, as it is done by the forest dwellers in Central India.

The silkworm larvae, grown on different forestry host plants were collected one day before to the spinning. The individual larvae were dissected for silk glands (Fig. 11). The dissected silk glands were allowed for 5-7 min in the buffer and later transferred to thin and neat blotting paper to dry the excess moisture on the surface of gland and immediately thereafter, weight of the silk glands was recorded. The silk glands are the most conspicuous part of the larval anatomy. Morphologically, the silk gland is divided into two distinct regions. The narrow proximal region is the excretory duct, whereas the thick, long and highly convoluted distal region is secretary region.

Tested forest tree species were taken as treatments, so there were 07 treatments, and the number of replications were 06. Normality of data was checked before to the statistical analysis. Descriptive statistics were calculated by using Microsoft Excel. Data of first and second rearing seasons were combined together, then treatment wise descriptive statistics were calculated and mean tables were prepared. A two-way completely randomized block factorial design was used to test the significance of difference in the means of variable. We did Factorial ANOVA by using advance statistical software, STATISTICA 10. Rearing season, host plant and their interactions were treated as the main (fixed) effects and silk gland weight served as dependent variables for block effect. The level of significance was fixed at p=0.05. Post HOC test was carried out by using Tukey’s HSD test to compare the homogeneous pairs of means.

As forest based commercial rearing of A. mylitta may be a new forestry employment avenue for the people of Uttarakhand, so it was considered to assess the major economic traits that influence silk yield of A. mylitta. Therefore, to find out the superiority of forestry host plants on different economic traits of A. mylitta, a common evaluation index (E.I.) method was applied as laid down by Mano et al. (1993). The use of E.I. in tropical tasar research sector is relatively new, and we have applied it for comparing the performance of A. mylitta on different forestry host plants, based on multiple traits E.I. analysis.

Evaluation Index (E.I.) shows an aggregate unit of cocoon yield of A. mylitta reared on a particular host plant in different rearing seasons, which was calculated as per the procedures outlined by Mano et al. (1993, 1998). For cocoon yield, E.I. of 50 or more than (>) 50 is considered suitable.

Where, A = Mean of a variable on a particular treatment; B = Over all mean of that variable on all the treatments; C = Over all standard deviation of that variable on all the treatments; 10 = Standard unit and; 50 = Fixed value

We carried out multiple regression analysis for dependent variable of weight of silk gland in matured larvae of A. mylitta over two independent variables of Weight gain (g) from I to V instar and Larval period(days).

Table 2 reveals that null hypothesis (H₀) for no difference on effect of rearing seasons, host plants and their interactions on mean weight of silk gland in mature larvae of A. mylitta was rejected by ANOVA. Table indicates that silk gland weight differed significantly between rearing seasons (DF=1, F=2333.98, p<0.05), host plants (DF 6, F= 1516.25, p<0.05) and their interactions (DF=6, F=7.10, p<0.05).

Fig. 12 indicates that silk gland weight in matured larvae of A. mylitta was found higher on all the forestry host plants during second rearing season as compared to the first rearing season. Further, table 3 reveals that during first rearing season, T. alata fed A. mylitta larvae recorded the highest weight of the silk gland (8.03 g), followed by T. tomentosa (7.19 g), T. arjuna (6.8 g) and L. speciosa (6.57 g) fed larvae. However, L. tomentosa fed larvae showed the lowest weight of silk gland (3.72 g), followed by T. chebula (4.42 g) and T. bellirica (5.3 g). In second rearing season, the corresponding figures were 9.47, 9.01, 8.08 and 7.83 g on T. alata, T. tomentosa, T. arjuna and L. speciosa, fed larvae respectively.

Results of Tukey HSD test in table 4 indicates that weight of silk gland differed significantly between rearing seasons and formed two homogeneous groups. Table 5 on Tukey HSD test for the effect of host plants on silk gland weight indicated seven homogeneous groups of means that differed significantly from one another.

Further, Tukey HSD test for the effect of interactions between rearing seasons and host plants (table 6) indicated ten homogeneous groups of means that differed significantly from one another. Table also showed that silk gland weight of A. mylitta larvae fed on L. speciosa (6.57 g) and T. Arjuna (6.80) did not differ significantly and formed a homogeneous group. Similarly, silk gland weight of T. arjuna fed larvae in second rearing season (8.08 g), T. alata fed larvae in first rearing season (8.03) and L. speciosa fed larvae in second rearing season (7.863 g) did not differ significantly from one another and formed one homogeneous group.

Weight of silk gland is a desirable character; therefore, a higher value of E.I. is required. Table 7 indicates that four forestry host plants viz., T. alata, T. tomentosa, T. arjuna and L. speciosa have registered higher indices of E.I. (>50) in both the rearing seasons and therefore, it is concluded that T. alata, T. tomentosa, T. arjuna, and L. speciosa are superior to T. bellirica, T. chebula and L. tomentosa, which scored lower indices of E.L. (<50).

Multiple regression analysis for dependent variable of mean weight of silk gland in matured larvae of A. mylitta with its two predictor viz., weight gain, and larval period from first to fifth instar larvae is presented in table 8. The multiple regression model produced R= 0.998; F (4,835) =84239; p<0.000. Table 5.242 indicates that the larval weight gain (g) from first to fifth instar larvae had significant positive regression weight (β=1.002, t=346.777, p= <0.05) on mean weight of silk gland; which indicates that with increasing weight gain, the silk gland weight is expected to increase. This regression equation indicates that weight gain from first to fifth instar larvae is a strong predictor for the silk gland weight of A. mylitta.

However, larval duration from first to fifth instar showed significant negative regression weight (β=-0.270, t=-8.436, p=<0.05) on mean weight of silk gland. It proves that less larval duration is a desirable feature for A. mylitta.

It was found that seasons, host plants and their interaction affected the weight of silk gland significantly (p<0.05). Further, in second rearing season, there was an increasing trend in the weight of silk gland on all the forestry host plants. The seasonal variations for this trait are in conformity with the findings of Yokoyama (1963), who demonstrated effect of season on development of silk gland in B. mori.

It was further found that T. alata, followed by T. tomentosa, T. arjuna and L. speciosa yielded silk gland of higher weight; whereas, L. tomentosa fed larvae showed the lower weight of silk gland, followed by T. chebula and T. bellirica. Result of evaluation index analysis also reflected the same trend. Nutritional profile of T. alata, T. tomentosa, T. arjuna and L. speciosa is reported to be superior than T. chebula, T. bellirica and L. tomentosa (Kohli et al., 1969; Sinha and Jolly, 1971; Jolly et al., 1979; Agarwal et al., 1980) that might have influenced the weight of silk gland of A. mylitta on these forestry host plants. Lower nutritional contents and poor palatability of the leaves of L. tomentosa and T. chebula caused lower silk gland weight in the matured larvae of A. mylitta. Findings of Banagade and Tembhare (2002) support this assertion, who concluded that starvation caused significant reduction in synthesis of silk gland protein. Effect of various food plants on concentration of amino acid in the haemolymph of A. mylitta is also reported (Hsiao, 1985), to affect weight of silk gland in matured larvae of A. mylitta (Lokesh et al., 2012).

Banagade and Tembhare (2002) studied the effect of some exogenous factors such as photoperiod, starvation, food plants, and supplementary food on silk gland protein in A. mylitta and concluded that starvation caused significant reduction in total protein concentration. Among the three food plants tested viz., T. tomentosa, T. arjuna and Anogeissus latifolia, the highest total silk gland protein concentration was recorded in case of feeding on T. tomentosa leaves, suggesting the most preferable host plant for A. mylitta.

Rajesh Kumar and Elangovan (2010) had also found significant variations in the silk gland weight of eri silkworm reared on different food plants. Results of present study are also supported by the findings of Lokesh et al., (2012), who studied variability in silk gland of three ecoraces of A. mylitta viz., Daba, Sukinda, and Sarihan by rearing on T. arjuna and found that Daba ecorace showed highest weight of silk gland.

Results of multiple regression analysis indicated that the larval weight gain had significant positive regression on mean weight of silk gland. It indicates that with increase in larval weight gain, a simultaneous increase can be expected in the weight of silk gland. Regression equation indicates that larval weight gain is a strong predictor for corresponding silk gland weight of A. mylitta. Secondly, larval duration showed significant negative regression on mean weight of silk gland. This indicates that with reduction in larval duration, weight of silk gland may also increase. It proves that less larval duration is a desirable feature for A. mylitta to have better silk production efficiency and the host plant that supports faster larval growth is superior.

Our results confirmed that for achieving higher weight of larval silk gland weight of A. mylitta in tropical forest areas of Uttarakhand, T. alata is the best-suited food plant, followed by T. arjuna, T. tomentosa, and L. speciosa. Accordingly, State Forest Department may initiate systematic plantation of T. alata, T. tomentosa, L. speciosa and T. arjuna through their various afforestation, reforestation and plantation programme to create a new forest insect industry of A. mylitta in Uttarakhand to promote its adoption by tribals and rural communities inhabiting in forest fringe areas to improve their livelihood.