Leaf structure vs. nutrient relationships vary with soil conditions in temperate shrubs and trees

Abstract

Often there are significant positive interspecific relationships between leaf area per unit dry mass (SLA) and foliar phosphorus and nitrogen concentrations ([P] and [N]). Most of these studies have been conducted on moderately acidic soils, and little is known of the generality of these relations as potentially affected by soil characteristics. We investigated foliage mineral composition in relation to leaf structure in a wooded meadow on calcareous alkaline soil, in a bog on strongly acidic soil, and in a flood plain on moderately acidic soil. Foliar nutrient contents and fertilization experiments indicated that foliage physiological activity was co-limited by both P and N availabilities in the wooded meadow, by P in the bog, and by N in the flood plain. In the wooded meadow and in the bog, there were positive relationships between SLA and P concentration ([P]), and no relationship between SLA and nitrogen concentration [N]. Given that the fraction of support tissues generally increases with decreasing SLA, the requirement for mineral nutrients is lower at low SLA. Thus, these contrasting relations between mineral nutrients and SLA suggest that P was distributed in a more “optimal” manner among the leaves with varying structure than N in P-limited communities. In the flood plain, SLA was positively related to both [P] and [N], possibly manifesting a strategy to cope with N limitations by enhancing N turnover, and accordingly, greater P requirement for nucleic acid formation in N-limited soils. Total variation in foliar structural and chemical characteristics was similar in all sites, and was mainly determined by variation among the species. Part of this variability was explained by life form and plant size. [P] was higher in trees than in shrubs, and [P] and P/N ratio increased with increasing total plant height, indicating that P nutrition was improved relative to N nutrition with increasing plant size. Since the capture of less mobile soil elements such as P is dependent on extensive root systems, but not that of readily mobile and temporarily variable elements such as N, this correlation was attributed to more extensive root systems in larger plants. Our study indicates that foliar structure vs. [N] and [P] relations may be separately regulated, but also that the generality of leaf structure vs. nutrient content relations may vary depending on soil conditions.

Introduction

The concentrations of foliar N ([N], Sobrado and Medina, 1980, Reich and Walters, 1994, Schulze et al., 1994, Niinemets et al., 2002) and P ([P], Sobrado and Medina, 1980, Bongers and Popma, 1990, Reich and Walters, 1994, Niinemets et al., 2001) often increase with increasing specific leaf area (SLA, leaf area per dry mass). Such relationships may partly be explained by a larger fraction of support structures in thicker and denser leaves that possess a lower SLA, and accordingly by a reduced requirement for mineral nutrients for construction of foliage with low SLA Sobrado and Medina, 1980, Reich et al., 1992, Niinemets et al., 1998, Niinemets et al., 1999. Since the nutrient vs. structure relations are frequently observed, it has been suggested that they represent fundamental convergence in leaf formation and functioning (Reich et al., 1999). Provided such relationships occur for all combinations of environmental factors, they would have paramount significance in regional and global scaling of stand leaf area and foliage carbon gain capacities.

In some situations, SLA vs. nutrient relationships are poor, disagreeing with their fundamental nature. For instance, the strength of SLA vs. [N] relations may be different in communities of varying water availability (Wright et al., 2001). Such differences may partly be explained by modifications in support investments due to changes in water availability (Niinemets, 2001). However, the hypothesis of the scaling of leaf nutrient concentrations with SLA due to alterations in fractional support investments does not explain why in some instances there are strong relationships of SLA with both [N] and [P], but in other cases only with one of these nutrients. Depending on nutrient supply from the soil, mineral nutrient contents may vary largely at a common SLA (e.g., Thompson et al., 1992). This probably arises from the circumstance that not all the foliage P and N are associated with cell metabolic functions. In fact, the nutrients in excess may be stored in the vacuoles Rebeille et al., 1983, Leigh and Storey, 1991. This may further indicate that more severely limiting nutrients present in only functional cell compartments have the strongest relationships with SLA, while the concentrations of less limiting elements may more strongly fluctuate due to the presence of the non-metabolic storage.

Nitrogen is generally considered as the most limiting nutrient in earth ecosystems, but there is increasing evidence that phosphorus may be the primary limiting element in many communities on acidic soils. Severe P limitations have been observed in tropical forests on soils rich in iron and aluminum forming insoluble compounds with phosphate Vitousek and Sanford, 1986, Reich et al., 1995, in temperate fens and bogs, where the availability of P is limited by slow turnover of organic P compounds Wilson and Fitter, 1984, Wassen et al., 1995, Niinemets et al., 2001, in Mediterranean-type ecosystems (Niinemets et al., 1999) and in temperate forest communities (Prescott et al., 1992) where both P immobilization by soil micro-organisms and P occlusion on soil minerals may play a role. Phosphate salts with Ca and Mg are also poorly soluble at higher soil pH (Lindsay and Vlek, 1977), and accordingly, P availability may be low in alkaline calcareous soils as well (e.g., Clement and Gessel, 1985). However, little attention has been paid to this potential P limitation resulting from the low mobility of P in alkaline soils. So far, the leaf structure/nutrient relationships have been studied only in acidic soils, but gaining insight into the generality of these relations requires more advanced understanding of the potential variability of these relations in different soils.

To test the hypothesis that the correlations between foliar nutrient concentrations and SLA are dependent on the mineral nutrient supply from the soil, we studied foliar morphology and stoichiometry of mineral nutrients in widespread perennials in a wooded meadow on calcareous soils, in a bog on very acidic soils and in a flood plain on moderately acidic soils. Primary production of herbaceous vegetation has been previously shown to respond primarily to P rather than to N in the wooded meadow (Pork, 1979) and in the bog Paavilainen, 1980, Finér, 1992, while the flood plain vegetation is mainly limited by N (Pork, 1963). Thus, we hypothesized (1) that P is present in limiting concentrations in the foliage in the bog and wooded meadow, but not in the flood plain, that [P] and [N] are positively correlated, and that other mineral elements are in a satisfactory supply. We further suggested (2) that there is a stronger positive relationship between [P] and SLA, than between SLA and other mineral nutrients in the P-limited ecosystems.

Within the same community, foliar nutrient concentrations vary largely among different species and among different individuals of the same species despite similar soil conditions Ricklefs and Matthew, 1982, McJannet et al., 1995, but the plant and species characteristics responsible for this variability are still poorly understood. Although such species and plant effects may be superimposed by general constraints on variation of leaf structure and nutrient contents, they may significantly alter the strength and predictive ability of these relationships. Total height of sampled woody plants varied by two orders of magnitude across the species in the current study. Since the plant ability to capture less mobile soil nutrients such as phosphate strongly depends on extensive root systems exploring large soil volumes, but that of more mobile mineral nutrients such as nitrate does necessarily not (Sands and Mulligan, 1990), we suggested that (3) foliar concentrations of less-mobile soil mineral nutrients are greater in larger woody plants, and that (4) this pattern is more evident in the case of more limiting nutrients.