Forest Investment Account (FIA) - Forest Science Program
FIA Project M075035

    Measurement and Modelling of Mountain Pine Beetle Impacts on the Annual Forest Water Balance
Project lead: Carlyle-Moses, Darryl
Contributing Authors: Burles, K.A.; Carlyle-Moses, Darryl E.
Imprint: Kamloops, BC : Thompson Rivers University, 2007
Subject: Forest Investment Account (FIA), Dendroctonus Ponderosae, British Columbia
Series: Forest Investment Account (FIA) - Forest Science Program
Aerial overview surveys indicated that in 2004 mountain pine beetle (MPB) infestations in the Southern Interior Region (SIR) of British Columbia covered approximately 4.2 million ha. A 1.5 fold expansion is expected in 2005. Stand characteristics of affected SIR watersheds have been or will be significantly altered due to the natural impact and extent of MPB incursion and the corresponding forest management operations. The post-beetle landscape of these watersheds will be an assemblage of cover types including, for example, clear-cut areas and stands of unaffected and affected juvenile lodgepole pine, selective cut mature mixed species, and mature lodgepole pine stands that have suffered varying degrees of MPB related mortality and defoliation. In addition, regeneration of harvested areas will result in a landscape comprised of a mosaic of juvenile stands of varying ages. Forests play a vital role in the terrestrial hydrologic cycle by partitioning water into different stores and fluxes such as canopy interception loss, snow melt, transpiration, and soil moisture storage. Thus, changes to the composition of forests as a result of MPB and associated harvest activities will have an impact on the magnitudes of hydrologic variables at different spatiotemporal scales. Although generalizations can be made regarding the impact insect infestations and harvesting practices may have on the hydrology of the SIR landscape, many such impacts are likely to be specific to the affected tree species, the management response, and the climatology, pedology and geology of the area. In addition to location specific variability, little is known about the hydrological effects of partial versus complete stand mortality or of the time to reach hydrologic recovery in these areas once regeneration begins. Thus, a detailed study that is specific to different MPB and related management scenarios is required if the impact on the hydrology of the SIR is to be determined. The proposed research will examine how different stands affected by MBP infestations, as well as management practices, such as clear-cutting, understory retention, and juvenile stand management affect the stand water balance. The proposed research addresses three of the research priorities outlined by Hélie et al. (2005): i) MPB impacts on soil moisture storage (Priority 1); ii) the impacts on precipitation/interception loss (Priority 4), and iii) the impacts on evapotranspiration loss (Priority 5). The proposed research will also support the calibration of forest hydrology models that predicatively link changes in hydrologic processes. The research site to be used during this study will be located at Mayson Lake on the Thompson Plateau north of Kamloops where widespread stand losses are anticipated over the next few years and salvage harvesting is underway. It is also the site of a long-term snow hydrology research program. Evaporation from forested environments is comprised of canopy interception loss, transpiration, litter layer interception loss, and direct evaporation from the soil matrix. Canopy interception loss, IC, the interception and subsequent sublimation and evaporation of precipitation by vegetation canopies, represents an important and sometimes the dominant component of the water balance of vegetated environments accounting for approximately 25 – 30 % of gross precipitation input to mature, undisturbed coniferous forest communities (Dunne and Leopold, 1978; Thurow et al., 1987; Carlyle-Moses, 2004; Winkler et al., 2005). Rain and snow that is not intercepted and subsequently lost from the canopy is partitioned into one of two understory precipitation (PU) inputs: throughfall (TF) and stemflow (SF). Throughfall is the portion of PU that reaches the forest floor by passing directly through gaps in the canopy or as canopy drip, while SF represents the portion of PU that flows down the boles of trees. Although TF typically accounts for 90 % or more of PU in coniferous environments, including lodgepole pine stands (see Brabender, 2005), SF may represent an important point input since many trees funnel water from their relatively large canopies where it is concentrated at their relatively small bases. Given the quantitative importance of IC and PU fluxes a number of models have been developed to estimate their magnitudes. Commonly reported regression models are limited in their spatiotemporal transferability since stand characteristics and meteorological variables controlling the rate of evaporation of intercepted rainfall are often ignored. The most commonly used conceptual models are the Rutter (1971, 1975) model and the analytical version of this model proposed by Gash (1979) and subsequently modified by Gash et al. (1995) and Valente et al. (1997). Although good agreement between observed and simulated IC and PU fluxes have been found at the plot-scale using the Rutter and Gash models in a variety of environments (e.g., Loustau et al., 1992; Carlyle-Moses and Price, 1999), these models, like their regression model counterparts, are limited in terms of their spatiotemporal transferability since they treat the canopy as a two-dimensional 'big-leaf’. The modified versions of the Rutter and Gash models scale parameters such as the storage canopy (S) and the during-rainfall evaporation rate from a saturated canopy (E) to the fraction of canopy cover, but no other stand characteristic. Thus, under the current Rutter and Gash model structures spatiotemporal extrapolation would generate similar model parameters, and thus IC and PU estimates, in stands of comparable canopy cover fraction even though other stand characteristics such as leaf area index (LAI), stand volume, and tree density may differ appreciably. Soil-vegetation-atmosphere transfer (SVAT) and watershed hydrology models often use the Rutter or Gash approaches in their interception modules to simulate IC and PU fluxes. These models, including the distributed hydrologic model of Wigmosta and Lettenmaier (1994), often scale IC, S and or E using non-linear functions of LAI without any theoretical justification. Few process-based field studies have examined the relationship between LAI and IC, and the research that has been completed suggests that LAI may not be the correct variable to use for scaling canopy water balance parameters. For example, direct methods of determining the per unit area storage capacities of leaves, needles, and branches have found that the storage capacities of the woody components of a stand are often an order of magnitude greater than the capacities of their leaf and needle counterparts (e.g., Llorens and Gallart, 2000). The results of recent measurement and modelling studies (e.g., Degushi et al., 2005; Wattenbach et al., 2005) also suggest that LAI is not the dominant factor controlling the magnitude of IC in other environments. Since PU represents the hydrologic input that determines the magnitude of terrestrial hydrologic stores and fluxes such as soil moisture, transpiration, and groundwater recharge, proper simulation of PU, and thus IC, is needed if good agreement between observed and simulated hydrologic stores and fluxes is to be achieved. Improvements to existing SVAT and watershed models can be realized if process-based research reveals more appropriate scaling relationships to be used to estimate the canopy water balance at different spatiotemporal scales so that the transition can be made from highly calibrated hydrologic models to more physically-based models that provide desired results for the right reasons. This process-based approach will help hydrologists and watershed managers move beyond the plot-scale and assess watershed and regional response to stand modification as a consequence of MPB and other types of forest disturbance. Improvements to existing models should, however, not result in the models becoming overly complex. Instead, the dominant processes and relationships between canopy water balance components and stand and meteorological variables in differing environments need to be identified and modelling approaches developed that concentrate on these dominant mechanisms. This approach, termed the Dominant Processes Concept (DPC) by Blöschl (2001), should focus on stand and meteorological information that is readily available so that the results of DPC studies will not be purely academic, but rather have a significant practical value.


Executive Summary (0.1Mb)
Abstract - Preliminary Investigation of ... (25Kb)
Abstract - Water Storage Capacities ... (75Kb)
Abstract - Magnitude and Variability ... (23Kb)

To view PDF documents you need Adobe Acrobat Reader, available free from the Adobe Web Site.

Updated August 16, 2010 

Search for other  FIA reports or other Ministry of Forests and Range publications.

Please direct questions or comments regarding publications to