Screening of Sugarcane Varieties for Tolerance to Water Deficiency Using Containers

,


Introduction
Te Zimbabwe's sugarcane industry is being threatened by the adverse efects of water defcits and rainfall variability caused by climate change. Marin et al. [1] warned that climate change is projected to have adverse efects on rainfall patterns, resulting in droughts and foods across the planet. It is predicted that in Southern Africa, climate change will cause more frequent droughts [2]. In Zimbabwe, frequent droughts are likely to curtail the production of many crops including sugarcane. Several dams including Mutirikwi-Tokwe; Manjirenji-Siya; and Manyuchi, supply irrigation water to the sugarcane estates [3]. However, poor and erratic rainfall due to climate change may result in poor replenishment of these dams, leading to a shortage in irrigation water [4].
One of the mitigation and adaptation strategies against water defcit due to droughts caused by climate change in sugarcane production in Zimbabwe involves planting drought tolerant varieties [5]. Te degree of tolerance to water-defcit varies among sugarcane genotypes [6,7]. However, in Zimbabwe, selection criteria for sugarcane genotypes released do not include water-defcit tolerance. Te selection of varieties in Zimbabwe is mainly based on high cane and sugar yields, tolerance to smut and ratoon stunting disease, and ability to resist sugarcane stalk borer (Eldana saccharina), among others [8]. Screening of sugarcane varieties for drought tolerance needs to be added to the selection criteria of sugarcane genotypes.
Screening sugarcane genotypes for water-defcit tolerance in the feld may be costly due to substantial requirements such as seed cane, water, fertilisers, herbicides and other inputs. Also, controlling irrigation and other sources of water in the feld is difcult. Instead of screening the sugarcane varieties in the feld, selection can be done in containers. Screening sugarcane varieties in containers enables the researcher to control variables such as irrigation, nutrition, and growing medium. Using containers in screening sugarcane varieties is also likely to reduce costs since the area required is small, which means fewer inputs for growing the plants.
According to Ferreira et al. [9] the most important stages of sugarcane sensitive to water-defcit are tillering and stem elongation phases. Terefore, the screening of sugarcane varieties for tolerance to water-defcit can be done in containers in the frst few months of growth. Te aim of the study was to screen sugarcane varieties commercially grown in Zimbabwe for their tolerance to water-defcit.

Materials and Methods
Te study was conducted at the Zimbabwe Sugar Association Experiment Station (ZSAES) located in the South Eastern Lowveld of Zimbabwe (21°01′S: 28°38′N; 430 m above sea level). Te area receives an average rainfall of 625 mm per annum, much of which falls in summer (October to March). Te mean air temperature ranges from 16°C in June and July to 26°C in October to January.

Experimental Design and Experimental Procedure.
A completely randomised design (CRD) was used consisting of a factorial treatment combination of 14 varieties and two levels of water application in which the treatment combinations were each replicated three times. Te 14 sugarcane varieties tested in the experiment were ZN1, ZN2, ZN3, ZN4, ZN5, ZN6, ZN7, ZN8, ZN9, ZN10, CP72-1312, NCo376, N14, and CP72-2086. Te two levels of irrigation tested were well-watered with 160 ml of water (100% of feld capacity) and water-stressed with 48 ml of water (30% of feld capacity). Te feld capacity was determined using the water budget method [8].
Sugarcane setts were cut at the base from 12-months old varieties using sterilised cane knives. Knives were sterilised by immersing for 5 minutes in diluted propan-2-ol or isopropanol (JEYES fuid) (0.5 L JEYES fuid per 10 L of water). Tree internodes from the base and top of the stalk were discarded. Before planting, one-eyed setts were cut from the remaining stalk and dipped for 5 minutes in triadimenol (Shavit) (1 ml Shavit/1000 ml of water) to prevent ratoon stunting disease. Tree sterilised single-eyed setts of each variety were planted 40 mm deep close to the centre of a container (54 cm diameter and 90 cm depth) in a flter cake + pine bark (1 : 1 v/v) medium on 20 September 2018. Te emerged plants were thinned at 14 days after planting (DAP) to leave one plant per container. A blend fertiliser Triple 16 (16% N, 16% P 2 O 5 and 16% K 2 O) was applied at a rate of 937.5 mg/l to the surface of the media in containers at 14 DAP and, thereafter, fortnightly until 70 DAP. Te plants were irrigated with one litre per container per day for the frst 14 DAP until germination. Weeds were removed by hand pulling. Regent (Fipronil) insecticide was sprayed at the base of the containers and around the experimental site to control termites and aphids. White grubs were controlled at planting using Carbrayl 85 WP (1-naphthol N-methylcarbamate) at the rate of 200 g Carbrayl in 200 L of water which was applied every fortnight.

Measurements and Statistical
Analysis. Stem height (cm) of the primary tiller was measured using a metre rule, from the base of the plant at the surface of the medium to the apex of the plant at 150 DAP. Tiller number in each container was taken at 150 DAP. SPAD index was measured on topmost leaf with visible dewlap (TVD leaf ) on the primary tiller at 150 DAP using chlorophyll metre (model Minolta SPAD-502, Minolta Co., Osaka, Japan) [10]. Leaf temperature, photosynthetic rate, transpiration, stomatal conductance, and vapour pressure defcit were measured on the TVD leaf of primary tillers between 0700 and 1000 hours using a portable photosynthetic system (CIRAS 3 model, PP systems, Amesbury, United States of America) at 150 DAP. Relative water content was determined on the TVD leaf of primary tiller at 149 DAP [11]. All tillers per pot were harvested by cutting at the base and separated into leaves, stems, and trash. Te roots were removed from the pots and the growth medium carefully washed of them. All plant parts were placed in a forced-air oven at 105°C for 72 hours. Te total dry mater was determined as the sum of dried roots, leaves, stems, and trash. Te data were subjected to Fisher's Analysis of Variance (ANOVA) using Genstat Version 14 th edition software (VSN international Ltd, Hemel Hempstead, United Kingdom). Treatment means were separated using Least Signifcant Diference and Standard Error Diference tests at 5% level [12].

Number of Sugarcane Tillers per Variety.
Tere was no interaction (p � 0.281) between sugarcane variety and water application rate on number of tillers. Te test varieties difered signifcantly in tillering. Variety ZN2 produced more tillers than the other varieties tested ( Figure 2). Sugarcane varieties ZN3, ZN1, ZN4, ZN5, ZN6, ZN7, ZN8, ZN9, and N14 had fewer tillers when compared to the rest.
Sugarcane plants grown under water-stressed conditions produced more tillers than the well-watered plants ( Figure 3).

SPAD Index of Primary Tiller Leaves with TVDs.
Tere was an interaction (p � 0.411) between the varieties and the water application rate on leaf SPAD index was not signifcant. Variety ZN3 had the highest leaf SPAD index but was not signifcantly diferent from that for ZN2, ZN5, and NCo376 ( Figure 4). Variety ZN4 had the lowest leaf SPAD index but was not signifcantly diferent from that of ZN1 and ZN10 (Figure 4).

Vapour Pressure Defcit (kPa) of the Primary Tiller TVD
Leaf. Tere was interaction between sugarcane variety and water application rate on vapour pressure defcit of the primary tiller TVD leaf. Te leaf vapour pressure defcit was not signifcantly afected by water application rate in 10 of the 14 test varieties (ZN1, ZN2, ZN3, ZN4, ZN5, ZN6, ZN7, ZN9, CP72-1312, and CP72-2086) ( Figure 5). In contrast, the leaf vapour pressure defcit increased markedly in three (ZN8, ZN10 and N14) of the remaining four varieties in the water stressed plants ( Figure 5) and decreased in the fourth variety (NCo376) ( Figure 5).

Relative Water Content, Photosynthetic Rate, and Temperature of the Primary Tiller TVD Leaf.
Tere was no interaction between variety and water application rate on relative water content, photosynthetic rate, and leaf temperature. Of all the varieties tested, only ZN2 and ZN10 had relative water content of <80% (Table 1), although it was insignifcant when compared to other. Well-watered plants had 4.9% more relative water content than water-stressed Sugarcane varieties  International Journal of Agronomy plants. Similarly, the photosynthetic rate of well-watered plants was higher than that of water-stressed plants by 46.8% (Table 1).
In contrast, transpiration rate decreased in the waterstressed treatment in the varieties ZN5, ZN8, ZN10 and N14 ( Figure 7). Relationships between transpiration rate and vapour pressure defcit (Figure 8(a)) as well as between stomatal conductance and vapour pressure defcit (Figure 8(b)) were inversely linear. However, the relationship between transpiration rate and vapour pressure defcit was a linear positive (Figure 8(c)).
On average, varieties ZN3, ZN10, ZN6, and ZN10 had the longest internodes (Table 2). In contrast, ZN2, ZN4, ZN5, ZN7, and CP72-2086 had the shortest average internode length (  internode, the lengths of internodes of well-watered plants increased more than that of water-stressed plants ( Table 2). Tere was a 75% decrease in the average lengths of internodes of water-stressed plants ( Table 2). Vapour Pressure Deficit (kPa) Figure 5: Vapour pressure defcit in the primary tiller TVD leaf among sugarcane varieties under well-watered or water stressed treatments.  (Table 3). Well-watered had more green leaves and green leaf area per container than water-stressed plants ( Table 3). Tere were no diferences in green leaf area among varieties tested.

Total Plant and Root Dry Matter and Shoot: Root Ratio.
Varieties ZN2, ZN5, and N14 had similar and greater root dry matter than the other varieties tested (

Discussion
Varieties ZN1, ZN3, ZN5, CP72-1312, NCo376 and CP72-2086 were comparatively taller under water-defcit stress than ZN2, ZN6, ZN7, ZN8, ZN9, ZN10 and N14 as a result of longer internodes lengths. Tis suggested that stem growth of varieties ZN1, ZN3, ZN5, CP72-1312, NCo376 and CP72-2086 was more tolerant of water stress than was the case with varieties ZN2, ZN6, ZN7, ZN8, ZN9, ZN10 and N14. Perhaps, may be explained by genetic diferences that existed between sugarcane varieties in extraction of water, which infuences stem elongation [9]. Water-defcit reduces stem elongation of   plants by decreasing cell elongation because of poor cell turgor pressure [13]. Water-defcit stress may also cause a decline in cell division, resulting in poor stem growth [13][14][15]. Te lack of interaction of variety and water application rate on number of tillers corroborates the observation by Ryes et al. [16] that tiller production during formative stages of sugarcane was not diferent among genotypes. Tillering is a complex physiological process which is afected by a wide array of factors that include environmental, endogenous, biotic and their interactions [17]. Tiller number in grasses is controlled by quantitative trait loci that have additive and not dominant efects [18]. For example, quantitative trait loci that afects tillering identifed at early stages of plant growth of rice was undetectable at maturation stage [19]. Although there was no interaction between variety and water application rate, the main efects were signifcant. Variety ZN2 had more tillers than other varieties. Ryes et al. [16] also reported diferences in tiller count among ten sugarcane genotypes. Tis suggested the presence of genetic variation among sugarcane genotypes on tiller production. Te present study also showed more tillering of plants under water-defcit stress than well-watered plants during the early stages. Water defcit-stress can promote tillering as a way of compensating for the reduced assimilation production during drought [20].
Te absence of any interaction between water application rate and variety on leaf SPAD index, suggests that this parameter is not useful for screening varieties for tolerance to water-defcit stress, especially in the formative stage of sugarcane growth. Tis was contrary to the fndings by Silva et al. [6,10] who recommended, that the leaf SPAD index could be used to screen sugarcane varieties for their tolerance to water defcit-stress. Te present results showed varietal diferences in leaf SPAD indices, and indication of genetic infuence on this parameter, and could perhaps be linked to diferences in nitrogen extraction. Tere is a strong correlation between nitrogen uptake and leaf SPAD index of sugarcane plants [21][22][23].
Varieties ZN8, ZN10, and N14 had higher leaf vapour pressure defcits than the other varieties under waterdefcit stress. Higher vapour pressure defcit in plants causes the leaves to close their stomata under drought, and thus facilitating conserving water [12]. Tis assertion was confrmed by negative relationship between vapour pressure defcit and stomatal conductance (Figure 8(a)) or transpiration rate (Figure 8(c)). Te stomatal conductance and transpiration rates of ZN8, ZN10, and N14 were signifcantly lower in water stressed plants (Figures 6 and  7). Tus varieties ZN8, ZN10, and N14 conserved more water than the other genotypes tested under drought [15]. Varieties ZN8, ZN10, and N14 were among varieties with higher total plant dry matter. Sugarcane varieties with higher dry matter and low stomatal conductance were reported to conserve water under drought conditions and  8 International Journal of Agronomy to be tolerant to water defcit stress by avoidance of dehydration [9,24]. Variety NCo376 had lower leaf vapour pressure defcit than other varieties under water-defcit stress ( Figure 5). In addition, the stomatal conductance of variety NCo376 was higher in water stressed plants ( Figure 6). NCo376 variety was among varieties with higher total plant dry mass (Table 4). Tese results suggested that NCo376 has a greater dehydration tolerance mechanism which allowed it to accumulate dry matter under water-defcit stress than other varieties tested [25]. Sugarcane varieties with dehydration tolerance mechanism under water-defcit stress are important for growing during drought seasons [9]. ZN8, ZN10, N14, and NCo376 were among varieties with the highest shoot: root ratio (Table 4). Tis confrms the tolerance of ZN8, ZN10, N14, and NCo376 to water-defcit stress when compared to the other genotypes. Under water defcit stress, plants enhance their root system as a tactic to extract more water, therefore, reduce shoot: root ratio [26]. Under drought conditions, plants re-allocate assimilates from shoot growth to root growth and this increases root length [27]. Lemoine et al. [28] reported that mild water defcit stress restricts shoot growth with little efect on root growth. In addition, ZN8, ZN10, N14, and NCo376 were among varieties with more green leaves than other varieties (Table 3). Tis supported the assertion that water defcit stress did not afect the shoot growth of ZN8, ZN10, N14, and NCo376.
Tere were neither interactions between water application rate and variety nor signifcant diferences among varieties on relative water content, photosynthetic rate, and leaf temperature (Table 1). Tus, relative water content, photosynthetic rate, and leaf temperature were parameters not suitable for use in screening sugarcane varieties to water defcit. Tis was in contrast to results by Silva et al. [29] and Marchiori et al. [30], who reported that sugarcane varieties could be screened for their tolerance to water defcit using relative water content, photosynthetic rate and leaf temperature. Although there were no signifcant diferences between varieties, the relative water contents of ZN2 and ZN10 were <80% (Table 1), indicating high sensitivity to water-defcit stress [31]. Water-defcit stressed plants had a reduced photosynthetic rate relative to well-watered plants ( Table 1). Oskabe et al. [32] reported that during drought, plant cells accumulate abscisic acid in the guard cells triggering stomatal closure which ultimately reduce photosynthesis. Other results of water defcit reducing photosynthesis have been reported in plants [33] and in sugarcane [9].
Tere were neither interactions between variety and water application rate nor diferences among varieties on green leaf area (Table 3). In contrast, Castro-Nava et al. [34] noted signifcant diferences between sugarcane genotypes in terms of green leaf area in their response to water-defcit stress, and this was conspicuous as the plant aged. Te green leaf area of sugarcane in this study was determined at 150 DAP, which might have been too early to note the diferences between varieties in their response to water-defcit stress.

Conclusions
Based on stem height, stomatal conductance, vapour pressure defcit, transpiration rate, and dry matter parameters measured in the present study, sugarcane varieties that are recommended to cane farmers in Zimbabwe when faced with drought are NCo376, ZN1, ZN8, ZN10, and N14. Further research in the screening of genotypes should include the use of molecular fngerprinting techniques to increase precision in identifying drought tolerant varieties. Tere is also need to include all released and unreleased genotypes in the screening of varieties to identify tolerant varieties for use in future breeding programs.

Data Availability
Te data used to support the fndings of this study are available from the corresponding author upon reasonable request.

Conflicts of Interest
Te authors declare that they have no conficts of interest.