Variation in leaf xeromorphism in the desert palm genus Washingtonia (Arecaceae)

Highlights

• The larger Sonoran Desert exhibits an environmental gradient where two species of Washingtonia palms occur.

• Palms at northern, arid sites exhibit evident xeromorphic traits like amphistomaty and isolaterality.

• Palms at more tropical sites have leaves that do not show xeromorphic traits.

• Leaf xeromorphism constitutes a syndrome in Washingtonia palms.

• The evolution of leaf xeromorphism is correlated with increasing environmental aridity.

Abstract

Washingtonia palms occur scattered in oases and canyons of the larger Sonoran Desert from lat. 24° to lat. 34° N. Northern oases have an arid temperate climate while those in the south experience seasonally dry, tropical conditions. A marked latitudinal cline in morphological characters has been described within the genus. We hypothesized that aridity-adapted leaf traits in Washingtonia palms, such as amphistomaty and isolaterality, could exhibit a predictable trend along the 1300-km-long gradient, allowing us to test whether the evolution of leaf xeromorphism is correlated with environmental aridity.

We took measurements of anatomical and functional traits of leaves in 16 oases spanning the whole distributional range of the genus in Mexico and the US. Using regression models, we examined the response of specific leaf area, leaf greenness, ẟ¹³C, stomatal size, and stomatal density to latitude, site, and landform. We used Principal Components Analysis and correlation matrices to identify bioclimatic variables that might be playing a role in the observed latitudinal patterns.

Palms at northern sites have waxy, glaucous, thick, leaves and exhibit xeromorphic traits such as amphistomaty and isolaterality. Palms at more tropical sites have glossy green, thin leaves that do not show xeromorphic traits.

Leaf xeromorphism, a morphological and physiological syndrome in Washingtonia palms, increases with aridity.

Introduction

One of the goals of community ecology is to find consistent trait-environment linkages that explain the occurrence of a group of species at a particular site (Díaz et al., 2004). Understanding how plant traits vary among species will ultimately have consequences at the community and ecosystem levels (Reich et al., 1999; Wright et al., 2004; Cornwell and Ackerly, 2009; Albert et al., 2010; Shipley et al., 2016; Funk et al., 2016). Leaves, in particular, play an extremely important role as they allow light interception and CO₂ uptake necessary for photosynthesis (Lambers et al., 2008). Leaf thickness, stomatal traits, gas exchange regime, and water use efficiency are important features of leaves that have a significant effect on the physiology of plants (Hill et al., 2014; Boer et al., 2016) and community composition (Hetherington and Woodward, 2003).

Most angiosperms possess leaves with a bifacial or dorsiventral structure. The upper (adaxial) side normally harbors a layer of palisade tissue, formed by a chloroplast-rich parenchyma of tightly packed columnar cells under the upper epidermis. Between the palisade tissue and the lower (abaxial) epidermis there is a spongy mesophyll, with cells widely separated from each other so that the circulation of CO₂ entering the leaf through the abaxial stomata and diffusing on to the palisade tissue above, is enhanced. In short, most angiosperms show some level of functional specialization in their leaf sides, the upper surface being specialized in the capture of light, and the lower one being specialized in the exchange of gases with the surrounding atmosphere (Smith et al., 1997).

In arid and other high-light environments such as coastal marshes, however, it is common to observe plants that have lost the dorsiventral specialization that distinguishes the adaxial from the abaxial side of their leaves, showing instead isolateral leaves with palisade tissue on both sides. Isolaterality in dryland plants is often accompanied by amphistomaty (the presence of stomata in roughly equal density in both sides of the leaf), as well as by a vertical orientation of the leaf laminae and increased leaf thickness (Smith et al., 1998).

Stomata play a crucial role for gas exchange across leaves (Hetherington and Woodward, 2003; Hill et al., 2014; Raven, 2014; Wang et al., 2015; Carlson et al., 2016; Boer et al., 2016). Stomatal density and size, two traits that have important consequences for the conductance to H₂O and CO₂ (Carlson et al., 2016), are influenced by different environmental factors such as temperature (Hill et al., 2014), humidity, CO₂ (Boer et al., 2016), and water availability (Fraser et al., 2009). Despite the importance that stomatal distribution has for stomatal conductance and, ultimately, photosynthesis, relatively few studies have evaluated the consequences of amphistomaty (Gindel, 1969; Clay and Quinn, 1978; Parkhurst, 1978; Mott et al., 1982; Sundberg, 1985; Taylor et al., 2011; Bucher et al., 2016).

Amphistomaty and isolaterality are very strikingly visible in many dominant desert plants such as jojoba (Simmondsia chinensis) in the Sonoran Desert or the different species of creosote bushes (Larrea) in both North and South American deserts (Pyykkö, 1966; Gibson, 1996). Indeed, many studies suggest that the xeromorphic leaf anatomy might be dominant in most drylands. Almost a century ago, Wood (1932) noted that while only 5% of plants in British woodlands are amphistomatic, 88% of plants in the sclerophyll forest in Victoria, Australia (34 out of 39 species) and a full 100% of 28 species sampled from the arid scrubland of Koonamore, Australia, were amphistomatic, concluding that amphistomaty “is possibly correlated with increasing aridity”. Hull and Bleckmann (1977) found that the stomatal density on the adaxial surface of leaflets of Prosopis tamarugo in growing chambers increased with lower relative air humidity, making the plants more amphistomatic. Similarly, Mott, Gibson, and O'Leary (1982) reported that, from a list of 119 dominant dryland species from floras of California, Arizona, and Mexico, 50% of the thinner-leaved species (100–200 μm) and over 80% of thick-leaved plants (500 μm) were amphistomatic. Sundberg (1986) found that amphistomatic leaves are more common in plants with succulent leaves and green stems. Gibson (1996) reported that, from a list of 301 globally distributed desert species, 278 showed isolateral leaf mesophylls. Finally, Arambarri et al. (2011) reported that, from a sample of 32 species of shrubs from the Dry Chaco forest in Argentina, “more than half presented a xeromorphic, or isolateral, [leaf mesophyll]”.

In short, the xeromorphic syndrome (isolateral, amphistomatic leaves, often thickened and vertically oriented) seems to be the dominant leaf morphology in desert plants. This might seem contradictory with existing theory, as some authors contend that hypostomatous leaves are better adapted to dry conditions than amphistomatous ones (Willmer and Fricker, 1996). This belief has been accepted, in part, because of the few existing studies that report stomatal density for both leaf surfaces. In those cases where the relationship between stomatal density and aridity or water stress has been studied, some unexpected results have emerged. In some desert species, and contrary to what is accepted, stomatal density has been reported to increase with decreasing water availability (Penfound, 1931; Evenari, 1962; Buttery et al., 1993; Smith et al., 1998). Some authors have suggested that variation in stomatal density in desert plants depends on the carbon fixation pathway (Sundberg, 1985; Herrera and Cuberos, 1990) with non-succulent desert plants having higher stomatal density than succulent plants with CAM metabolism. Others have found that amphistomaty is related to leaf thickness, where thick leaves tend to be more amphistomatic than thin ones (Parkhurst, 1978).

Most of the studies discussed above imply the comparison of plants from arid against plants from non-arid environments, which often have different taxonomic lineages. Thus, differences in plant morphology can be attributed to both environmental influence and taxonomically fixed traits, and separating the effect of the environment from the effect of taxonomic inheritance is often a complex task (Harvey and Purvis, 1991). A simpler approach to understand changes in leaf traits, like stomatal density, across different environments is to study the variation in a single species or in closely related species of a genus throughout their distribution range. This approach —examining variability within the same taxon— has recently been encouraged by several researchers studying plant functional traits (Cornwell and Ackerly, 2009; Jung et al., 2010; Albert et al., 2010; Burns and Strauss, 2012; Shipley et al., 2016).

The larger Sonoran Desert, including the Baja California peninsula, exhibits a dramatic environmental gradient characterized by an arid temperate climate dominated by winter rains in the north and a dry tropical climate with summer rains in the south (Vogl and McHargue, 1966; Garcillán and Ezcurra, 2003). Latitudinally, it extends some 1300 km, from the Cape Region at latitude 23°N to the Lower Colorado Valley and the Mojave, at 36°. The genus Washingtonia, constituted by tall phreatophytic palms with C₃ photosynthesis, is found in canyons and oases along the whole gradient, from the tropical tip of the peninsula to the hot, arid ranges of southern California and Arizona (Fig. 1). Although the genus has been historically considered as comprised of two species (W. filifera in the north and W. robusta in the south), there is really a latitudinal cline within the genus, with continuous variation in morphological characters such as stem diameter, petiole width, or crown size, all of which change gradually from the tropics to the mid-latitude deserts (Villanueva-Almanza et al., 2018). Continuous morphological variation with latitude supports the hypothesis posed by some authors (e.g., Nabhan, 1986) that the group is composed of one single taxon and that W. robusta is really a variant of W. filifera.

Based on these results, we hypothesized that leaf functional traits in Washingtonia palms, and very especially those related to the bifacial or isolateral anatomy of the leaves, could also exhibit a predictable trend along the N–S gradient and allow us to test the hypothesis that the evolution of leaf xeromorphism is correlated with increasing environmental aridity. Additionally, gradually changing leaf morphology along the N–S latitudinal gradient would add support to the hypothesis that what has been considered two distinct species of Washingtonia is, in fact, a single variable taxon with leaf adaptations resulting from different environmental preferences. Therefore, the research questions of our study were: (1) How does stomatal density in both leaf surfaces of Washingtonia palms, as well as other leaf traits, change along the latitudinal gradient? and (2) are there any environmental variables that might explain observed variations in leaf morphology?