The relationship between climate and leaf shape in the Agave cerulata complex.

Abstract

Agave adaptation to aridity is examined, comparing trends among phylads, clines in Deserticolae taxa of Baja California, and variation in A. deserti along an elevational gradient. Agave physiology is reviewed and recent evolutionary scenarios are discussed. Tentative hypotheses predict characteristics of Agave leaves and rosettes in arid climates. A study of bioclimatology in the Vizcaino Region of Baja California follows, aimed at defining aspects of a subtropical arid climate that are relevant to a plant. The Vizcaino Region is described in terms of its physiography, vegetation physiognomy and floristics. Ombrothermic diagrams and juxtaposed graphs relate temperature and rainfall at stations throughout the region. Temperature regimes are compared using the 5th, 25th, 75th, and 95th percentiles of their respective distributions of monthly means, and subregional groups with similar regimes are defined. Conditions when soil moisture is available are examined by segregating 'wet' months with rainfall totals of 5 mm or more. Thermal distributions of wet months are compared with respect to their shapes and to temperatures delimiting the central two-thirds of wet months. Patterns in mean annual precipitation are presented. Variables are derived to estimate stress imposed on plants by drought. Median annual precipitation deficit, defined as the difference between total annual potential evapotranspiration and total annual rainfall, expresses the stress typically experienced by plants. The 90th percentile of potential evapotranspiration of dry intervals estimates the severity of droughts that longer-lived perennials survive. Each derived climate variable shows a different geographic pattern. Variance in leaf measurements from collection sites throughout the range of Agave cerulata is studied with a principal components analysis. Leaf characters associated with the major components of variance together with leaf volume/surface ratios are used as dependent variables in multiple regression equations with climate variables, which are estimated for each collection site. As the precipitation deficit increases, leaf size generally decreases, but where the longest droughts are most stressful, leaves tend to be larger. Leaf volume/surface is coupled with leaf size. When size variables are included in regressions, higher volume/surface is associated with more extreme cold, cooler summers, and warmer winters. Warmer summers, higher precipitation deficits, and more warm-season storms are correlated with higher length/width ratios. Larger basal surfaces occur in conjunction with higher precipitation deficits, warmer temperatures, and less warm-season rainfall.