Parsons, Scott Anthony (2010) Decomposition, nutrient cycling and climate change in Australian tropical rainforests. PhD thesis, James Cook University.

Preview

PDF (Thesis) - Submitted Version Download (6MB) | Preview

Abstract

Knowledge of the mechanisms that dictate the composition and dynamics of ecosystems is essential for understanding the natural world. It is important to consolidate our understanding of ecosystem processes due to the need to understand and adapt to global anthropogenic climate change. The processes of decomposition and nutrient cycling that occur on the soil surface in forested environments both sustain ecosystems and have a substantial influence on the biosphere at many scales. This encapsulates the processes of plant litterfall, litter decomposition and nutrient release. Litter decomposition and nutrient cycling, although controlled by climate, vegetation and soil communities, are highly spatially and temporally variable. Our lack of understanding of these processes limits our ability to understand and adapt to climate change. The Australian Wet Tropics bioregion of north Queensland is an interesting natural environment in which to investigate the drivers and controls on litter decomposition and nutrient cycling. The region contains a range of tropical rainforest types on a variety of soils and is subject to varied climatic conditions. The risks of adverse effects from climate change are also high for this region, potentially leading to substantial losses of biodiversity and rare plant communities.

Decomposition and nutrient cycling were studied in this region at 21 locations (from near sea level to around 1300 m elevation), covering most of the climate range and soil conditions of the region. The aim was to understand the patterns and controls on these processes and how these controls may deviate from their current states under climate change scenarios. The approach determined spatial and temporal patterns in: leaf litter decomposition rates and nutrient dynamics in leaf litter using the litterbag technique for ~ 420 days; litterfall nutrients and chemical quality (i.e. chemical potential to decompose) of litterfall; litterfall rates and seasonality (in-conjunction with another PhD student); the amount of litter on the soil surface (litter standing crop, LSC) and the seasonality and turnover/duration of litter on the soil surface. Models explaining litter dynamics were then applied to climate change predictions specific to the region to determine the sensitivity of litter processes to climate.

Plant litter chemical quality is a highly significant driver of decomposition and nutrient cycling processes; however, standard methods for determining litter quality indices are often arduous and limited in their ability to explain ecological phenomena. Near infrared spectrometry (NIRS) has the potential to extend standard methods for chemical quantification. NIRS was used here to quantify the concentrations of nutrients, carbon fractions (total carbon and lignocellulose portions) and plant secondary compounds in litterfall, and leaf litter that underwent decomposition. NIRS also accurately predicted litter decomposition rates based on their initial NIR spectral composition (i.e. organic chemical composition) to determine litter decomposability, and was successfully used to model chemical changes in the material during decomposition.

The first exponential decomposition rate constants of in situ leaf litter (litter characteristic of each site) ranged from 0.33 y⁻¹ (upland microphyll fern forest on granite) to 2.15 (complex mesophyll vine forest on basalt). Decomposition rates were explained well by climate, soil and litter quality, for litter collected in situ: average leaf wetness in the dry season (LWDS, moisture condensation) and the initial P content of the leaves (r² = 0.78, p < 0.001, n = 17), or LWDS and initial C (r² = 0.75, p < 0.001, n = 17); control treatment (a standard leaf litter, no litter quality effect): rainfall seasonality (% dry season days with 0 mm rainfall), soil P, and mean annual temperature (r² = 0.78, p < 0.001, n = 12). Nutrients were not mineralised for periods of more than 12 months. Increased temperatures and moisture (especially in the dry season) improved lignocellulose and C mineralisation.

Litterfall leaf litter quality (24 months worth of sampling from 40 study plots at 20 sites) was driven by the combination of soil fertility (nutrient contents), climate (phenolics and C) and species/disturbance (lignocellulose components). Trends in litter quality indicated a negative feedback on soil and nutrient cycling processes in more stressed environments characterised by higher rainfall seasonality and lower soil fertility. Also, short term climate changes were determinants of litter chemical quality, with NIRS predicted decomposability lower, and total phenolic contents higher, in the dry season.

Two year average litterfall rates ranged from 4.89 to 11.29 t ha⁻¹ y⁻¹ (n = 40 plots). No environmental variable could explain litterfall rates but calculations were hindered by the secondary status of the vegetation, particularly resulting from damage caused by Cyclone Larry at many sites in March 2006. Seasonality of litterfall (vector algebra index) was linearly related to mean annual temperature and soil nitrogen. The temperature effect was partially explained by dry season moisture, however the trend was for higher seasonality in the wetter/cooler uplands (n = 29). LSC was determined in the field by a volumetric method developed especially for this study. LSC values ranged from 3.70 t ha⁻¹ to 10.94 t ha⁻¹ (n = 36 plots), and were explained by litter quality (NIRS decomposability) and the composition of litterfall, along with soil Na, mean annual temperature and leaf litter C content. LSC turnover quotients ranged from 0.57 to 2.81, and were controlled by similar variables to mean annual LSC. Seasonality in LSC was linearly related to soil Na. Local variability was high for mean annual LSC, with around 35% of the full regional variation contained within single 1 km transects. Climate change scenarios suggest temperature increases and decreases in dry season rainfall, with associated uncertainty, and dependent on emission scenario (mean for the full range of SRES, WRE450 and WRE550). The predicted changes in climate related to increases in the climate decomposition index (potential for climate driven decomposition, determined from min/max temperatures and monthly rainfall totals/seasonality) of +5.2 to +20.5% from current conditions (average from 40 study plots). Predicted changes in leaf decomposition rate and leaf lignin mineralisation rate (from control litterbag study), and full litter layer turnover rate were determined at 10 year time steps until 2080. For 2080 relative to present day, leaf decay rate showed large uncertainty: -7.46 to +8.15%; lignin mineralisation increased: -0.32 to +3.39%; and litter turnover increased: +5.9 to +24.2. The uncertainty in the leaf decay models were driven by uncertainty in the changes in dry season rainfall. The data suggests increasing decomposition rates from current conditions for poorer (chemical) quality material, such as whole litter standing crop and leaf lignin, compared to less recalcitrant material such as leaf litter. The magnitude of change is predicted to be greater at upland sites than lowland sites due to the non-linear relationship between temperature and the climate decomposition index, and on poorer nutrient soils due to the increasing effect of temperature on the decomposition of low chemical quality litter.

The extent and direction of change in these forests will depend not only on the direct effects of temperature and dry season rainfall and subsequent alterations in soil level litter processes, but also changes in primary productivity, including the timing and seasonality of litter inputs, and climate-driven succession of vegetation and plant traits such as litter chemical quality. These changes in the landscape feed back to global biogeochemical cycles in complex ways. Increases in litter decay, as mostly predicted to occur from this work, may act to further accelerate global warming. However, the direction of climate change driven changes in primary productivity, vegetation communities and plant litter quality, are essential in determining outcomes.

Item ID:	29564
Item Type:	Thesis (PhD)
Keywords:	climate change; tropical rainforests; nutrient cycling; decomposition; litterfall; leaf litter quality; biogeochemical cycles; Wet Tropics; North Queensland
Related URLs:	http://eprints.jcu.edu.au/19629/ http://eprints.jcu.edu.au/19630/ http://eprints.jcu.edu.au/5618/
Copyright Information:	Copyright © 2010 Scott Anthony Parsons
Additional Information:	Publications arising from this thesis are available from the Related URLs field. The publications are: Chapter 2. Parsons, Scott A., Lawler, Ivan R., Congdon, Robert A., and Williams, Stephen E. (2011) Rainforest litter quality and chemical controls on leaf decomposition with near-infrared spectrometry. Journal of Plant Nutrition and Soil Science, 174 (5). pp. 710-720. Chapter 4. Parsons, Scott A., Congdon, Robert A., Storlie, Collin J., Shoo, Luke P., and Williams, Stephen E. (2012) Regional patterns and controls of leaf decomposition in Australian tropical rainforests. Austral Ecology, 37 (7). pp. 845-854. Chapter 5. Parsons, Scott, Shoo, Luke P., and Williams, Stephen (2009) Volume measurements for quicker determination of forest litter standing crop. Journal of Tropical Ecology, 25 (6). pp. 665-669.
Date Deposited:	11 Oct 2013 01:23
FoR Codes:	05 ENVIRONMENTAL SCIENCES > 0501 Ecological Applications > 050101 Ecological Impacts of Climate Change @ 33% 05 ENVIRONMENTAL SCIENCES > 0501 Ecological Applications > 050102 Ecosystem Function @ 34% 06 BIOLOGICAL SCIENCES > 0602 Ecology > 060208 Terrestrial Ecology @ 33%
SEO Codes:	82 PLANT PRODUCTION AND PLANT PRIMARY PRODUCTS > 8201 Forestry > 820104 Native Forests @ 40% 97 EXPANDING KNOWLEDGE > 970106 Expanding Knowledge in the Biological Sciences @ 40% 97 EXPANDING KNOWLEDGE > 970103 Expanding Knowledge in the Chemical Sciences @ 20%
Downloads:	Total: 599 Last 12 Months: 1
	More Statistics