Brian Murphy, Rob Vertessy, Fletcher Townsend, Jim Morris, Craig. Beverly, Geoff ... Fred DeClosey from DECC and Colin Steele from NSW Premiers Depart-.
WATER RESOURCES RESEARCH, VOL. 43, W08415, doi:10.1029/2006WR005016, 2007
Partitioning the effects of pine plantations and climate variability on runoff from a large catchment in southeastern Australia Narendra Kumar Tuteja,1,2 Jai Vaze,3,2 Jin Teng,3,2 and Martin Mutendeudzi4 Received 5 March 2006; revised 14 February 2007; accepted 28 March 2007; published 15 August 2007.
[1] Effects of substantial increase in area of pine plantations from 1960 to 2000 on runoff
in a large catchment in southeastern Australia are quantified. Reliable land use maps were prepared for 1960–1979, 1980–1989, and 1990–2000 conditions from various data sources. Land use changes in the subcatchments have occurred at varying rates (16 to 28%) with pines replacing pasture and native woody vegetation. On the basis of long-term trends in rainfall-runoff relationships, flow duration curves, and history of land use changes, it is shown that there is strong evidence of reduction in runoff over a wide range. Modeling methodology using a lumped catchment-scale rainfall-runoff model (SMAR) and landscape-scale ecohydrological models (CLASS U3M-1D, CLASS PGM, and 3PG+) was implemented in a catchment framework to partition the effects of land use change and climate variability during 1960–2000. Runoff reductions from land use change in the range 22–52 mm/yr are estimated for different subcatchments. Annual yield impact per 10% of the catchment forested (AYI/10%) from catchment-scale modeling is estimated to vary between 14.3 to 19.2 mm/yr for the subcatchments. AYI/10% from landscape-scale modeling is estimated to vary between 12.8 to 21.3 mm/yr. Citation: Tuteja, N. K., J. Vaze, J. Teng, and M. Mutendeudzi (2007), Partitioning the effects of pine plantations and climate variability on runoff from a large catchment in southeastern Australia, Water Resour. Res., 43, W08415, doi:10.1029/2006WR005016.
1. Introduction [2] The potential of tree plantations to compete for limited water resources has been investigated by a number of researchers across the world [e.g., Dye, 1987; Roberts and Rosier, 1993; Farley et al., 2005]. Substantial evidence exists across the globe that afforestation with either pines or eucalypts can reduce streamflow [e.g., Vertessy, 1999; Zhang et al., 2001; Jackson et al., 2005]. The total water use of plantations depends on the climate factors affecting climate demand, the amount and seasonal distribution of rainfall, the hydrological factors affecting supply and the response of trees to climate, hydrology and plantation management. In Australia, if annual rainfall is less than 1000 mm, plantations will use most if not all rainfall and there will likely be little or no drainage [White et al., 2000]. The crop factor or relative transpiration (k), a ratio of actual water use to potential evaporation [Priestley and Taylor, 1972], provides a means of comparing water use with groundcover/leaf area index (LAI) of different vegetation. Maximum k value of 1 for irrigated E. globulus plantations (LAI > 6; Battaglia and Sands [1997]), 0.8 for E. maculata and P. radiata (LAI 3– 4; White et al. [2000]), 0.55 to 0.7 for mixed Eucalyptus tree belt (LAI 0.25; Dunin et al. 1 Scientific Services Division, NSW Department of Environment and Climate Change (DECC), Queanbeyan, NSW, Australia. 2 Also at eWater Cooperative Research Centre, Canberra, Australia. 3 NSW Department of Water and Energy (DWE), Queanbeyan, NSW, Australia. 4 Bureau of Rural Sciences (BRS), Canberra, ACT, Australia.
Copyright 2007 by the American Geophysical Union. 0043-1397/07/2006WR005016
[1999], White et al. [2000]) have been observed indicating high water use by trees. [3] In temperate environments the evapotranspiration of a mature grassland is significantly less than the forested land [Holmes and Sinclair, 1986; Zhang et al., 2001]. Simple but reliable models are available that can predict the impact of afforestation on mean annual runoff. According to water use impact curves of Holmes and Sinclair [1986], a fully forested eucalypt catchment would evaporate 40, 90, 215, 240, and 250 mm more per year than a fully grassed catchment with mean annual rainfalls of 600, 800, 1300, 1500, and 1800 mm, respectively. Zhang et al. [2001] tested a simple process-based model on 250 catchments around the world and found similar results up to 1200 mm mean annual rainfall. Zhang et al. [2001] estimated higher evapotranspiration rates up to 1400 mm at wet locations compared to 1150 mm predicted by Holmes and Sinclair [1986]. Both Holmes and Sinclair [1986] and Zhang et al. [2001] provide generalized relationships and do not consider species differences within the forest and grasslands, partial land use change and heterogeneity within the catchment. [4] Farley et al. [2005] in a global synthesis of the plantation impacts on water yield found that annual runoff reduced on average by about 44% (±3%) and 31% (±2%) when grasslands and shrublands were afforested, respectively. Farley et al. [2005] indicated that in regions where natural runoff is less than 10% of mean annual rainfall, afforestation should result in complete loss of runoff; where natural runoff is 30% of mean annual rainfall, it will likely be reduced by half when trees are planted. Stednick [1996] summarized forestry impacts from 95 deforestation studies in the US and showed that annual yield impacts at 10% of
W08415
1 of 18
W08415
TUTEJA ET AL.: EFFECTS OF PINE PLANTATIONS AND CLIMATE ON RUNOFF
land use change (AYI/10%) varied in the range of 9 – 61 mm with an overall average of 25mm. Stednick [1996] concluded that about 15 to 50% threshold of land use change must occur in a catchment before water yield change can be detected, supporting the findings of Bosch and Hewlett [1982]. Results from afforestation studies are not directly interchangeable with those from deforestation studies although they indicate the likely range of impacts on streamflow. [5] Our knowledge of the effects of forestry on streamflow in Australia is based on a few paired catchment treatment experiments conducted across high rainfall locations (>900mm/yr) [Vertessy, 1999; Vertessy et al., 2003; Bosch and Hewlett, 1982; Cornish, 1989; Dye, 1996; Huxman et al., 2005; Ruprecht and Stoneman, 1993]. Most of these are site specific paired small catchment studies affected by local climate, geomorphic and hydrogeological conditions where the catchments are either cleared, forested, regrowth or forest reconversion experiments [Best et al., 2003]. Few existing studies report that catchments afforested with pine show larger reductions in streamflow compared to those with eucalypt forests [e.g., Dunin and Mackay, 1982; Feller, 1981]. These differences between pines and eucalypts are opposite to those reported in South Africa [Bosch, 1979; Smith and Scott, 1992] and in the global review by Farley et al. [2005]. Vertessy [1999] point out that in Australia, pine plantations have been compared to native eucalypt forests, rather than to eucalypt plantations, as in South Africa and elsewhere. [6] Generalised relationships of tree water use provide a good initial assessment of the impacts of land use change at mean annual timescale. However, these approaches do not account for differences in species, partial land use change, heterogeneity in soils and climate within the catchment. Although they have been applied to large catchments, their concepts have been developed and tested from afforestation experiments in small catchments (
