Soil fluxes of mineral elements and dissolved organic matter following manipulation of leaf litter input in a Taiwan Chamaecyparis forest

Abstract

Forest ecosystems in Taiwan are frequently influenced by typhoons that cause large amounts of litter input to the soil. The rapid decomposition of such litter under the high precipitation and temperature conditions may trigger nutrient losses via seepage. Our goal was to investigate the effects of exceptionally high inputs of Chamaecyparis obtusa var. formosana leaves to the soil on the solute fluxes and concentrations of mineral elements, dissolved organic carbon (DOC) and dissolved organic nitrogen (DON) in a Lithic Leptosol. We simulated the effect of a typhoon by adding the annual aboveground litterfall (4600 kg ha⁻¹) and about 3 times the annual litterfall (13,900 kg ha⁻¹) as fresh leaves in a single event to small manipulation plots. All plots were also subjected to the natural litterfall. Soil solution was collected 4 months before and 15 months following the litter manipulation in 14 days intervals underneath the forest floor with free draining lysimeters and by ceramic suction cups at 60 cm depth. The 3-fold litter addition caused increased fluxes of K, Mg, Ca, and NH₄ in forest floor percolates as compared to the 1-fold litter treatment. DOC and DON fluxes also increased, but this was only statistically significant for DON. DON was the dominant form of N both in forest floor percolates and in seepage. The 3-fold litter manipulation increased the seepage fluxes of K, DON, and DOC relative to the 1-fold treatment with largest differences recorded for K, reaching 12.4 kg ha⁻¹ 15 month⁻¹. The fluxes of DOC with forest floor percolates and with seepage in both treatments exceeded published values by far. Our results suggest that the decomposition of large amounts of fresh leaf litter may trigger K losses from the ecosystem via seepage, whereas the probability for additional N and DOC losses is moderate. For Ca and Mg, additional losses seem to be rather unlikely.

Introduction

Forest ecosystems of Taiwan are subjected to high precipitation rates, with large annual variations depending on the number and strength of typhoons. On average typhoons hit Taiwan 3.7 times a year (Wu and Kuo, 1999). In those ecosystems that are affected by high wind speeds, drastically increased litter inputs to the soil were reported for Taiwan forests (Horng et al., 1995, Lin et al., 2003) and elsewhere (Ostertag et al., 2003). The litter derived from typhoon events represents to a large extent fresh leaves with relatively high nutrient concentrations in comparison to regular litter (Xu et al., 2004). The expected high rates of nutrient release from the decomposition of fresh leaves might exceed the capacity of the vegetation, microorganisms, and the soil to retain the nutrients in the ecosystem. As a result, nutrient losses by seepage are likely, especially under high precipitation conditions. The losses of mineral elements with runoff from a forested catchment in Puerto Rico, influenced by a hurricane, were increased for K and to a minor extend also for NH₄ and NO₃ (Schaefer et al., 2000). Increased gaseous N losses from the soil were also reported as a consequence of a hurricane (Erickson and Ayala, 2004).

The release of nutrient cations from decomposing litter is often dramatically faster than mass loss, especially in case of K, and to lesser degree also for Mg (Xu et al., 2004, Hirobe et al., 2004, Laskowski et al., 1995). In a decomposition study with Chamaecyparis obtusa var. formosana leaves, about 80% of the K in the leaves was released by leaching during the first 2 months of field exposure (Rees et al., 2006). The release of Mg was also much faster than the mass loss of the decomposing litter. On the other hand, the N content of the decomposing litter increased, suggesting a low risk of mineral N losses by seepage.

The leaching of dissolved organic carbon (DOC) and nitrogen (DON) from decomposing litter might be an important contribution to mass loss, especially under high precipitation conditions, since positive relations between DOC and DON fluxes with forest floor leachates and the amount of precipitation were reported for temperate forest ecosystems (Michalzik et al., 2001). Under low N inputs by atmospheric deposition, the fluxes of DON represent important contributions to the N cycle. Some studies found DON as the dominating form of N in soil solution (Van Breemen, 2002, Qualls et al., 2002) and also the dominating loss of N from the ecosystem via seepage and runoff (Perakis and Hedin, 2002). The fluxes of DOC and DON in soils depend largely on tree species and other environmental conditions, but still it is very difficult to predict those from boundary conditions alone (Qualls, 2000, Kalbitz et al., 2000).

While there are numerous field studies available dealing with solute organic matter fluxes in soils of the temperate climate zone, only few studies were conducted in subtropical and tropical systems under high precipitation and temperature regimes. The relevance of DOC in the C cycle under these conditions was already emphasized (e.g.: Liu and Sheu, 2003, Neff et al., 2000, McDowell, 1998), but only very few studies have focussed on DON: Wilcke et al. (2001) found DON as the dominant form of N in litter leachates in a mountain forest in Ecuador, while – on the contrary – the dominance of mineral N was reported in solute fluxes in a wet tropical forest ecosystem in Costa Rica (Schwendenmann and Veldkamp, 2005). However, the flux of DON in litter percolates at the Costa Rica site was considerable (13 kg N ha⁻¹ year⁻¹).

The goal of our study was to investigate the effects of exceptionally high litter inputs on the fluxes of mineral elements, DOC, and DON in soil solution in order to improve our understanding of potential typhoon effects on forest ecosystems. Given high precipitation rates and temperatures at the study site, our hypothesis was that exceptionally high leaf litter inputs cause losses of nutrients with seepage. The study is based on the experimental manipulation of litter input, on measurements of decomposition rates in litter bags and on soil solution analysis. Here, we report on concentrations and solute fluxes in forest floor percolates and soil solutions at 60 cm depth.