Abiotic or environmental stresses such as drought, heat, salinity and cold are major impediments to plant survival and productivity in many parts of the world. Plants respond to abiotic stress conditions through diverse biochemical and physiological processes such as accumulation of osmolytes and proteins, reduction in stomatal conductance, increase in photorespiration and general reduction in growth rate. Osmotic adjustment is one of the common mechanisms of plant response to abiotic stress signals. Water availability is the limiting factor common to all abiotic stresses. As the water potential of the soil water decreases, plants accumulate solutes to reduce the osmotic potential and to maintain water uptake . Several inorganic solutes such as K+, Na+, Cl- and organic solutes such as total soluble sugars, proline, glycine betaine and mannitol are involved in osmotic adjustment. Stress conditions also lead to accumulation of harmful reactive oxygen species (ROS) such as hydroxyl radicals, singlet oxygen, hydrogen peroxide (H2O2) and super oxide (O2-). Antioxidant enzymes such as superoxide dismutase (SOD), ascorbate peroxidase (ASX) and catalase help protect plants cells from the harmful effects of ROS . Expression of several detoxification enzymes was shown to increase under stress conditions .
Several studies of transcriptional responses to abiotic stress using microarrays [4–8] have identified stress inducible genes that often belong to one of two groups, based on the functions of their gene products. Genes belonging to group 1 are mainly involved in water transport (aquaporins), cellular membrane protection and integrity under stress conditions (proline, glycine betaine, mannitol), scavenging of free oxygen radicals (SOD, catalase), and protecting macromolecules (late embryogenesis abundant proteins - LEA, chaperons). The second group consists of genes that encode regulatory proteins (transcription factors, protein kinases, protein phospatases and calmoduluin binding proteins) and proteins involved in signal transduction [9, 10]. Stress induced transcription factors are classified into two classes, ABA dependent and ABA independent. The ABA dependent transcription factors include MYC/MYB and ABA responsive element binding/ABA binding factor (AREB/ABF) and the ABA independent transcription factors are dehydration responsive element binding/C-repeat binding factors (DREB/CBF) belonging to the ethylene responsive factor/APETALA2 (ERF/AP2) family [9, 11]. The other transcription factors responding to abiotic stress conditions are basic-domain leucine-zipper (bZIP), WRKY binding  and NACs .
While microarray analyses are useful in revealing genes that are responsive to different conditions, identification of allelic variants from genes showing differential expression may enable their application in breeding by marker-assisted selection. Recent developments in sequencing technology are making it possible to combine gene discovery with identification of allelic variation. Transcriptome sequencing or RNA sequencing (RNA-seq) is an approach for quantifying transcripts, in which RNA samples are converted to cDNA and sequenced, typically using high throughput methods. The resulting reads are then mapped against a reference genome sequence or assembled de novo to produce genome-scale transcriptome maps consisting of the structure and abundance of each gene . The abundance of each transcript is determined by counting the number of sequences mapped to the corresponding gene thus providing digital estimate of gene expression. The main advantages of RNA-seq over microarray analysis are a). As RNA-seq is based on counting sequences, cross hybridisation problems associated with microarrays are avoided b). RNA-seq has high dynamic range of detection i.e. very low and very high abundance transcripts can be detected with RNA-seq while microarrays lack sensitivity to detect genes expressed at either high or low levels. Using this technique Zenoni et al. detected several genes expressed during berry development in Vitis vinifera. Similarly several protein coding genes related to xylem formation were identified in an Eucalyptus plantation tree using RNA-seq . RNA-seq is also useful for identifying and estimating transcript abundances from alternatively spliced variants . By sequencing several individuals from different populations it is also possible to identify single nucleotide polymorphisms (SNPs) from genes showing differential expression.
In addition transcriptome sequencing can also be used to study the evolutionary selection patterns of genes by estimating nonsynonymous to synonymous substitution (Ka/Ks) ratios. Novaes et al. have shown that most of the genes are under purifying selection by sequencing RNA from different tissues bulked from several individual trees in E. grandis. Combining gene discovery with analysis of selection signatures may provide insights into natural selection patterns under drought stress.
Eucalyptus camaldulensis is one of the most widely planted tree species in the world , and is grown extensively in plantations for pulp production in the tropics of South and South East Asia [21, 22]. Water availability is the most important factor determining the establishment and composition of tree species in the dry tropics . The seedling stage is the critical period for survival and establishment of trees . In this study we analysed the physiological responses of seedlings of three E. camaldulensis populations subjected to water stress. RNA extracted from leaves of these seedlings was used in RNA-seq analysis to study gene expression patterns under well watered and water stressed conditions. The main objectives of this study are to identify genes differentially expressed under control and stress conditions, to identify allelic variants from these genes and to study the evolutionary signatures of selection.