Xylem is a complex tissue with multiple functions, including water, nutrient and hormone transport, plant support and reserve storage. Xylem anatomy has to adapt to the contrasting demands of these functions (Carlquist, 2001; Chave et al., 2009), and this involves trade-offs in terms of invested resources and of room devoted to the different tissue elements (Poorter et al., 2010). Despite a phylogenetic control of the xylem traits (Sperry, 2003; Willson et al., 2008), interspecific differences in xylem anatomy and function have been shown to be related to environmental factors such as climate (Myer, 1922; Lev-Yadun & Aloni, 1995; Martínez-Cabrera et al., 2009) and to contrasting plant life histories (Poorter et al., 2010). Xylem anatomy also shows intraspecific variability (Esteban et al., 2010; von Arx et al., 2012) and interannual variation within an individual along the different ontogenetic stages (Domec & Gartner, 2002).
It is thus not surprising that the quantitative analysis of xylem anatomical characteristics has become a promising field in dendrochronology (see Fonti et al., 2010). Most of the existing studies have focused on the analysis of conductive elements, both in conifers (e.g. Kirdyanov et al., 2003; Panyushkina et al., 2003; Olano et al., 2012; Xu et al., 2012) and angiosperms (e.g. Woodcock, 1989; García-González & Fonti, 2006; Fonti et al., 2007). This focus on the conductive function is probably related not only to its robust climate signal, but also to the increasing understanding of the relationships between the structures and functioning of conductive elements (Hacke & Sperry, 2001; Pittermann et al., 2006). By contrast, analysis of inter-year variability in xylem parenchyma traits and their relationships with environmental variation has remained largely unexplored (but see Lev-Yadun, 1998; von Arx et al., 2012), probably because a direct connection with the environmental and physiological mechanisms that induce parenchyma formation is not so evident.
The paucity of studies analysing xylem parenchyma contrasts with the key functions that are driven by this tissue. Ray parenchyma is the primary means of radial transport and as such is also involved in communication between xylem and phloem and the plant body in general, wound response, stem biomechanics, heartwood formation, water accumulation and reserve storage (van Bel, 1990; Gartner et al., 2000; Burgert & Eckstein, 2001; Pruyn et al., 2002; Spicer & Holbrook, 2007). The reserve storage function of the ray parenchyma may satisfy different needs, such as providing a sink for photoassimilates (Myer, 1922; Gartner et al., 2000) and establishing a pool from which to produce chemical defence compounds and energy for stem resprouting (Climent et al., 1998). Moreover, an interconnection between carbon starvation and plant mortality after stressful conditions such as drought has been highlighted (McDowell et al., 2008; McDowell & Sevanto, 2010). Recent research suggests a potential role of xylem parenchyma in the refilling of conductive elements after the occurrence of embolism (Holtta et al., 2006; Salleo et al., 2009; Nardini et al., 2011); a strong link is therefore seen to exist between xylem parenchyma and the adequate functionality of the hydraulic system (McDowell, 2011; McDowell et al., 2011).
Reserve levels in parenchyma show strong seasonal variation, apparently with no associated anatomical changes (Sauter & Neumann, 1994; Cruz & Moreno, 2001; Palacio et al., 2007). However, the amount of ray parenchyma within a tree ring may be a good estimator of its potential maximum carbohydrate storage capacity (Spicer & Holbrook, 2007). Assessing the amount of xylem parenchyma on a year-by-year basis may provide information on the variation in a plant's stem resources storage capacity through its life. Annual sequences of the amount of ray parenchyma from several individuals from a population can be averaged into a chronology that could potentially be related to time series of environmental factors, in order to assess the external control of parenchyma formation. The aim of this study was to explore the potential of ray parenchyma as a new environmental proxy. Specifically, we wanted to evaluate the relationships between different parenchyma-based chronologies and climatic conditions, and to compare these relationships with those obtained from tree-ring width, the most widely used proxy. To that end, we constructed yearly ring-width and ray parenchyma chronologies for Spanish juniper (Juniperus thurifera) growing under continental Mediterranean conditions, and related them to climatic time series from the study area along a 40-yr period. Our specific questions were: which climatic factors influence the amount of ray parenchyma?; Have the ray parenchyma-based chronologies a climatic signal independent from ring width?; and Has this signal the potential to be used as a climate proxy?