Leaf life-span in relation to leaf, plant, and stand characteristics among diverse ecosystems

Abstract

Variation in leaf life-span has long been considered of ecological significance. Despite this, quantitative evaluation of the relationships between leaf life-span and other plant and ecosystem characteristics has been rare. In this paper we ask whether leaf lifespan is related to other leaf, plant, and stand traits of species from diverse ecosystems and biomes. We also examine the interaction between leaf, plant, and stand traits and their relation to productivity and ecological patterns. Among all species, both mass- (Amass) and area-based (Aarea) maximum net photosynthesis decreased with- increasing leaf life-span, but the relationship was stronger on a mass (P < .001, r2 = 0.70) than an area (P < .05, r2 = 0.24) basis. Similarly, mass-based leaf nitrogen (leafNmass) decreased (P < .001, r2 = 0.52) with leaf life-span, but area-based leaf N (leaf Narea) did not (P > .25, r2 = 0.01). Specific leaf area (SLA, leaf area/leaf dry mass) and leaf diffusive conductance also decreased with increasing leaf life-span. Decreasing Amass with increasing leaf life-span results from the impact of decreasing Nmass and SLA on Amass· Variation in leaf traits as a function of leaflife-span was similar for broad-leaved and needle-leaved subsets of the data. These leaf-scale data from several biomes were compared to a data set from a single biome, Amazonia. For several leaf traits (e.g., SLA, Nmass• and Amass) the quantitative relationship with leaf life-span was similar in the two independent data sets, suggesting that these are fundamental relations applicable to all species. Amass was a linear function of Nmass (P < .001, r2 = 0.74) with a regression similar to previous analyses, while Aarea was not significantly related to Narea· These results suggest that the photosynthesis-leafN relationship among species should be considered universal when expressed on a mass, but not on a leaf arna, basis. Relative growth rates (RGR) and leaf area ratio (LAR, the whole-plant ratio of leaf area to total dry mass) of seedlings decreased with increasing leaflife-span (P < .001, r2 = 0.61 and 0.89, respectively). LAR was positively related to both RGR and Amass (r2 = 0.68 and 0.84, respectively), and Amass and RGR were also positively related (r2 = 0.55). Absolute height growth rates of young trees decreased with increasing leaf life-span (P < .001, r2 = 0.72) and increased with Amass (P < .001, r2 = 0.78). It appears that a suite of traits including short leaf life-span and high leafNmas"' SLA, LAR, and Amass interactively contribute to high growth rates in open-grown individuals. These traits interact similarly at the stand level, but stands differ from individuals in one key trait. In closed-canopy forests, species with longer lived foliage (and low LAR as seedlings) have greater foliage mass per unit ground area (P < .001, r2 = 0.74) and a greater proportion of total mass in foliage. The above ground production efficiency (ANPP /foliar biomass) of forest stands decreased markedly with increasing leaf life-span or total foliage mass (P < .001, r2 = 0.78 and 0.72, respectively), probably as a result of decreasing Amass• Nmam and SLA, all of which were positively related with production efficiency and negatively related to total foliage mass. However, high foliage mass of species with extended leaf life-spans appears to compensate for low production per unit foliage, since above ground net primary production (ANPP, in megagrams per hectare per year) of forest stands was not related to leaf life-span. Extended leaf life-span also appears to compensate for lower potential production per unit leaf N per unit time, with the result that stand-level N use efficiency is weakly positively related to leaf life-span. We hypothesize that co-variation among species in leaflife-span, SLA, leafNmam Amass• and growth rate reflects a set of mutually supporting traits that interact to determine plant behavior and production, and provide a useful conceptual link between processes at short term leaf scales and longer term whole plant and stand-level scales. Although this paper has focused on leaf life-span, this trait is so closely interrelated with several others that this cohort of leaf traits should be viewed as causally interrelated. Generality in the relationships between leaf life-span and other plant traits across diverse communities and ecosystems suggests that they are universal in nature and thus can provide a quantitative link and/ or common currency for ecological comparisons among diverse systems.