Drought Mitigation in Cocoa (Theobroma Cacao L.) Through Developing Tolerant Hybrids

Background: Cocoa, being a shade loving crop cannot withstand long periods of water stress. Breeding for drought tolerance is the need of the hour due to change in climatic condition and extension of crop to non traditional area. Hybrids were produced by crossing four tolerant genotypes in all possible combination. The cross GVI 55 x M 13.12 didn’t yield any fruit due to cross incompatibility between these genotypes. Various biochemical parameters act as the true indicators to select tolerant and susceptible types. The major biochemical parameters considered after imposing stress included proline, nitrate reductase activity, superoxide dismutase content and glycine betaine. Results: This experiment showed the role of different biochemical enzymes and osmolytes in giving tolerance to plants during drought stress. Logistic regression analysis selected proline and nitrate reductase as the two biochemical markers for identifying ecient drought tolerant genotypes in the future breeding programmes.

Moreover, due to the increase rate of consumption of cocoa in chocolate industry, the cultivation is extending to non-traditional area. In such area acute water shortage is the most important problem to be addressed. Hence to cop up with the changing climatic scenario and cropping pattern it is essential to bred genotypes that can tolerate water stress.

Experimental site
The hybridization work was conducted at Cocoa Research Centre (CRC), Kerala Agricultural University, India, using four genotypes identi ed as tolerant to drought in a preliminary study [5] as parents ( Table   1). They were crossed in diallel out of which, the cross GVI 55 x M 13.12 yielded no fruits due to cross incompatibility. The pods matured approximately within 5-6 months; mature pods from each cross were harvested separately and raised in the nursery. Since parents are in heterozygous condition F1 population itself was segregating. Hence each individual was considered as separate hybrids. Since parents are in heterozygous condition F1 population itself was segregating. Hence each individual was considered as separate hybrids.

Drought stress imposition and management
Based on initial vigour, hybrids were selected and subjected to moisture stress following gravimetric method [6]. Initial trials in CRC indicated that cocoa cannot tolerate water stress less than 40 per cent eld capacity [7]. Hence, 40 percent eld capacity was maintained for two weeks. This was achieved by supplementing the water lost by evaporation and transpiration twice daily (morning and evening). A control was also kept at fully irrigated condition representing each of the crosses. Based on the percentage of leaves retained after the drought imposition for two weeks, morphological classi cation of these hybrids were done based on the score chart ( Table 2).The humidity and the temperature of the mist chamber were recorded using Berlin's psychrometeron on daily basis. Percent of leaves retained was calculated using the formula.
Percentage of leaves retained = Number of leaves retained x 100 Total number of leaves

5.3.Analysis of biochemical parameters
Biochemical analysis was carried out using standard procedures after two weeks of stress imposition.

Statistical analysis
Analysis of variance was done for biochemical analysis for all selected hybrids following completely randomised design (CRD). Binary regression was carried out to nd the relationship between the dependent biochemical parameters over the independent variable, the number of leaves retained.

Selection of hybrids and classi cation
A total of 1505 hybrid seedlings were raised in the nursery. Based on initial vigour 120 hybrids were selected representing eleven crosses. After imposition of water stress for two weeks the percentage of leaves withered was calculated and hybrids were classi ed as highly tolerant (HT), tolerant (T), susceptible (S) and highly susceptible (HS) as per the score chart given. Table 2. Different biochemical characters recorded from 120 hybrids are statistically analysed and depicted in Table 3. In all the 11 crosses, all the tolerant and highly tolerant hybrids indicated high proline content as compared to the susceptible hybrids. The control recorded minimum amount of proline under full irrigated conditions (Table 3).
In the cross M 13.12 x G I 5.9, the highest content was found in hybrid H27 (1105.64 µg/g) and the lowest value was found in H18 (167.18 µg/g When the progenies of all crosses were compared, it was seen that the content ranged from 85.52 µg/g in H44 (M 13.12 x G VI 55) to 2817.39 µg/g in H43 (M 13.12 x G VI 55

Effect of water stress on Glycine Betaine (GB)
When the hybrids of present study were analysed, the tolerant hybrid, H103 (G II 19.5 x G I 5.9) was having high amount of glycine betaine of about 13.79 µmol/g and the susceptible hybrid, H3 (4.42 µmol/g) of the cross M13.12 x G I 5.9 recorded the least. All the tolerant hybrids had high glycine betaine content as compared to the susceptible hybrids.
However, the control which was kept under fully irrigated condition had least amount of glycine betaine among the crosses indicating the accumulation of glycine betaine under drought stress conditions. In cross M 13.12 x G I 5.9, the highest glycine betaine value was found in H26 (11.64 µmol/g) and the lowest value was observed in H3 (4.42 µmol/g). In cross M 13.12 x G II 19.5, the highest value was found in hybrid H34 (9.68 µmol/g) whereas the lowest value was observed in H35 (6.73 µmol/g). In cross M13.12 x G VI 55 the highest value was observed in H48 (9.87 µmol/g) and the H45 (5.29 µmol/g) recorded the lowest glycine betaine value.
In cross G I 5.9 x M 13.12, high GB values were observed in H52 (8.10 µmol/g) while H55 (5.36 µmol/g) had the lowest value.In cross G I 5.9 x G II 19.5, the highest value was observed in H63 (9.52 µmol/g) to as low as 6.45 µmol/g in H70.In cross G I 5.9 x G VI 55, the highest value was observed in hybrid H71 (9.56 µmol/g). Lowest values were observed in H73 (7.07 µmol/g).In cross G II 19.5 x M 13.12 , highest GB value was observed in H88 (10.08 µmol/g) while the lowest value was observed in H80 (6.78 µmol/g).In the cross between G II 19.5 x G I 5.9, H103 (13.79 µmol/g) recorded the highest value while the lowest value was observed in H99 (7.20 µmol/g).The cross G II 19.5 x G VI 55 had two hybrids, the hybrid H107 had high GB value of 9.34 µmol/g and the other hybrid, H108 which is the susceptible one, had 7.18 µmol/g of GB. In cross G VI 55 x G I 5.9 , H112 was having the highest GB content of about 10.61 µmol/g while the lowest value was observed in H109 (6.93 µmol/g).In the cross G VI 55 x G II 19.5 , the highest value was observed in hybrid H119 (9.23 µmol/g) while the lowest value was observed in H116 (4.42 µmol/g).

Logistic regression analysis
The positive and comparable value of odds ratio Exp (B) and positive correlation indicated that proline content and NRA content had a positive correlation with drought tolerance and also these characters expressed a signi cant value of less than 0.05 which is the constant indicating the 95 per cent accuracy with the results.
Based on the Exp (B) value from regression model, expressed percentage for drought tolerance over the base population was calculated and it was found that if selection is based on proline, new population formed from the base population will express 51 per cent improvement regarding the tolerance. In case of NRA, new population will show 87.48 per cent improvement over base population ( Table 4). The biochemical phenes and their association with drought tolerance in cocoa is depicted in Fig 1. Hence, these two biochemical parameters can be used as biochemical marker for screening drought tolerant hybrids of cocoa in the future breeding programmes. vigor and stress imposition were done. In cocoa initial vigour is directly correlated to yield [12].
The evaluation and screening of crops under drought stress were generally done on the basis of biochemical parameters. Proline is an important amino acid found in proteins. It has a signi cant contribution in drought tolerance. Even within a single cross, highly tolerant as well as susceptible hybrids were found. This is because parents are heterozygous in nature [13]. The analysis of proline clearly indicated that resistant plants were high in proline content compared to susceptible ones. A clear-cut difference was observed between highly tolerant and susceptible genotypes indicating that proline plays an important role in drought tolerance of cocoa. High levels of proline enabled the plant to maintain low water potentials. Apart from acting as an osmolyte for osmotic adjustment, proline contributes to stabilizing sub-cellular structures (eg., membranes and proteins), scavenging free radicals and buffering cellular redox potential under stress conditions [14].
Another important biochemical parameter contributes to drought stress is nitrate reductase activity. Water de cit induces an abrupt reduction in the uptake and nitrate ux rates from roots to leaves, preventing the mechanisms of NR protein synthesis induction and NR activity [15]. The NR activity decline during water stress is mainly attributed to low NO3 -absorption and availability resulting from water uptake deprivation [16]. The present results indicated that the tolerant hybrids had high resistance to drought stress and were able to regulate the nitrate reduction activity even with less water. Usually, drought stress reduces the enzyme activity and that is the reason the amount of reductase enzyme was low in hybrids whereas in the control, it was more. NRase is closely associated with plant growth and development [17]. It is generally accepted that drought stress has a negative impact on plant's photosynthetic activity, N concentrations, free amino acids or soluble protein contents accompanied with a decline of nitrate reductase activity in many plant species, such as maize [15], potato [18], winter wheat [19], etc. The plants subjected to water stress produces less amount of total protein which results in a decrease in the synthesis of nitrate reductase activity caused by low nitrate ux [20].
Reactive Oxygen Species (ROS) accumulation during stress greatly depends on the balance between ROS production and ROS scavenging mechanism [21]. When plants are subjected to any kind of stress, the cells have an increased production of reactive oxygen species (ROS) which in normal cases, is removed from time to time. Under stress, these become high in number and results in oxidative damage. These are removed by anti-oxidant systems which form the rst line of defence which is superoxide dismutase.
ROS-scavenging mechanisms were shown to have an important role in protecting plants against osmotic stresses [22].
In the experiment when plants were subjected to analysis, the tolerant and highly tolerant hybrids showed more amount of superoxide dismutase enzyme as compared to the susceptible hybrids and the control which was kept under fully irrigated condition had the least amount of SOD in it. This indicated that SOD will get accumulated under drought stress conditions and forms a defence system against the stress [23]. Similar studies on development of water stress in Curtilobum solanum and Solanum tuberosum due to over production of SOD in chloroplasts were reported by [24] supported our ndings.
Many plants accumulate compounds, termed compatible solutes, to cope with stress conditions. One of the most extensively studied compatible solutes is glycine betaine [25] Not only GB acts as an osmoregulator, but also stabilizes the structures and activities of enzymes and protein complexes, and maintains the integrity of membranes against the damaging effects of stress [26].
Genes associated with glycine betaine synthesis in higher plants and microbes have been transferred into plants which do not accumulate glycine betaine, such as Arabidopsis thaliana [27] , Brassica napus [28] Persimmon [29] and rice [30]. The metabolic engineering of glycine betaine biosynthesis in these plants improved the tolerance of transgenic plants to salt, drought and extreme temperature stresses [31].
Hence, these biochemical parameters were reliable source to detect drought stress in cocoa. However, to nd out the relationship between the dependent variables (the biochemical parameters) and the independent variable (the number of leaves retained, a binary regression analysis was carried out.The results of present study indicated that the genotypes selected could e ciently be used as contrasting parents in different hybridization breeding programs intended for genetic improvement of cocoa for drought tolerance.

Conclusion
The parameters proline and glycine betaine represented osmolyte group whereas nitrate reductase and superoxide dismutase represented enzyme group. In all the crosses, the content of proline was high in highly tolerant and tolerant hybrids as compared to susceptible hybrids. The control which was fully irrigated condition was having the least amount of proline. The glycine betaine also followed the same trend, as tolerant and highly tolerant hybrids were having more amount of glycine betaine as compared to susceptible hybrids and the control indicating that these two osmolytes accumulated only during water stress. In case of superoxide dismutase, the highly tolerant and tolerant hybrids were having high amount of superoxide dismutase as compared to the susceptible hybrids whereas the control was having the least amount of superoxide dismutase enzyme. In case of nitrate reductase activity, the highly tolerant and tolerant hybrids were having high amount of enzyme as compared to the susceptible hybrids. The control kept at fully irrigated condition was having the highest amount of nitrate reductase enzyme. The hybrids having high amount of nitrate reductase were more tolerant because generally, this enzyme reduces under drought stress. Regression analysis indicated that 87.46 per cent of improvement over the base population can be achieved in case of nitrate reductase and 51 per cent improvement in case of proline if these two are used as biochemical markers. Hence, these two characters can be used as reliable parameters in drought breeding programmes in the future. Declarations