ANR Proposal. Effect of forest management on water yields and other ecosystem services in Sierra Nevada forests

Transcription

1 ANR Proposal Effect of forest management on water yields and other ecosystem services in Sierra Nevada forests 1

2 Project Narrative Context. The Sierra Nevada harbors globally distinctive forest resources that deliver a wide variety of benefits to the citizens of California and elsewhere. These benefits derived from natural ecosystems also called ecosystem services include recreation-, biodiversity-, conservation-, water-, and forest-product-related services. These ecosystem services often pose competing aims relative to forest management, but there are few mechanisms to evaluate the tradeoffs and complements related to different strategies. Water is arguably the highest-value ecosystem service associated with the conifer forests of California s Sierra Nevada. Yet the provision of this essential service is vulnerable to changes in the energy and water balance associated with climate warming. To date, we have observed more precipitation falling as rain versus snow, earlier snowmelt, and greater summer water deficits. Such climate forcing will impact the water balance for the foreseeable future. However there is the potential to manage the water balance in forest ecosystems. The dominant vegetation (i.e., trees) is highly productive, forms dense canopies, and consequently, uses a great deal of water. There is a strong positive correlation between annual net primary productivity (the ultimate measure of the photosynthetic capacity of the ecosystem) and evapotranspiration (the primary cause of water loss). Any manipulation that reduces the productivity (i.e., removes trees) reduces evapotranspiration, shifts the balance of energy driving snowmelt, and thus may affect soil-water storage and streamflow. Water from the Sierra Nevada provides both hydropower and water supply to downstream users. Reducing and restructuring the forest vegetation density can also mitigate the negative impacts of wildfires as well as accomplishing important forest-restoration objectives. Project design. It is proposed to undertake a three-part, multi-year and multi-disciplinary research and assessment project that addresses issues related to climate warming, vegetation manipulation, and the forest water cycle. The three components of the work plan are: i) measurements at sites of opportunity where fire or thinning treatments have taken place, are taking place, or are proposed, ii) meta-analysis and modeling using available data to interpret these results, and iii) evaluation of multiple ecosystem services and how multiple service providers (land and resource owners/ managers) can effectively interact with service consumers (downstream and downhill residents). This project grew in part out of the Sierra Nevada Watershed Ecosystem Enhancement Project (SWEEP), which was launched to examine the potential to optimize snowpack retention through forest management while also quantifying the economic value of snowpack retention, water storage, and flow attenuation. The proposition underlying SWEEP is that water users that benefit from these services might be willing to pay upstream landowners to provide these services, providing powerful financial incentives to landowners to invest in beneficial management practices. The Sierra Nevada Conservancy and the Bella Vista Fund underwrote SWEEP, which was done in collaboration with the Environmental Defense Fund. All three organizations will be key stakeholders in the current project, and are expected to remain active advocates for a subsequent implementation phase involving on-the-ground management actions that follow from the research proposed here. The broader area of interest is the American R. basin, chosen for two reasons: i) an ability to leverage ongoing research, and ii) a committed and active group of stakeholders who recognize the 2

3 need for this research. SWEEP focused on the American largely because of this second reason, as did other complementary ongoing research. 1 A basic premise for this project, that forest ecosystems can be managed to meet water resource priorities, builds on the well-established link between water and forests (Bosch and Hewlett, 1982; Brown et al., 2005; Hornbeck et al., 1993). Despite general reviews, based on historical data, that augmentations to water from forest management may be undetectable ((NRC 2008; Sedell et al., 2008; Kattelman et al., 1983), our view is that the knowledge base for applying these general conclusions to the Sierra Nevada is very weak. The intersection of climate change, modern forest management, advances in measurement technology, strong economic incentives, and growing stakeholder interest provide very strong incentives for the proposed strategic research (CADWR 2010; CADWR 2008; Bales et al., 2006; Kapnick and Hall, 2010; Miller et al., 2009; Rauman and Soulard 2007; Stewart 1996; Vicuna 2006; Westerling et al., 2006). Activities and outcomes. The main expected outcome from this research is the filling of key knowledge gaps regarding the potential to enhance multiple ecosystem services, particularly those related to water, through forest management. Following is a brief description of the three components of the work plan noted above, with further details found in the Project Plan section. The field measurement program will develop data and information on end-member ecosystems that will inform subsequent analysis and modeling of how water-related ecosystem services respond to climate change and management actions. The particular end members of interest are recently burned areas, and high-density forests that are candidates for significant thinning; intermediatedensity forests are addressed by current measurement programs (see map in Project Plan). Measurements will focus mainly on snowpack accumulation and melting, soil moisture, runoff, evapotranspiration (ET) and leaf area index (LAI). Ancillary measurements on meteorology, snowcover, soils and forest structure are available from other research programs. The meta-analysis and modeling will use available data from SNAMP, SWEEP and other research programs such as the NSF-supported Southern Sierra Critical Zone Observatory and the Blodgett Experimental Forest to integrate data and develop scenarios for experimental forest management. It will also involve a synthesis of information on ecosystem services around hydropower, water supply, carbon sequestration, fuels management, and wildlife. This information will determine the scenarios to be examined in the modeling and analysis. It has been our finding and position throughout the SNAMP and SWEEP studies that this level of detailed modeling must be supported by a focused measurement program that is specific to the Sierra Nevada lands of interest. It must also be informed by forward-looking management approaches rather than historical practices or constraints; and must use models that are sufficiently detailed to explicitly link the ecosystem and water processes using the rich data sets now available. The evaluation of ecosystem services will involve synthesis of information and working closely with key stakeholders. Any decision to pursue a forest management approach to affect snow melt and water yield will involve tradeoffs between the costs of these operations and the maintenance of these conditions with potential benefits to power generation, water supply, fire risks, and wood supply. 1 These include the Sierra Nevada Adaptive Management Project (SNAMP), initiated through a partnership involving the U.S. Forest Service, the California Resources Agency, and UC. SNAMP is expected to continue through The California Department of Water Resources, in partnership with local agencies and UC, has initiated building of a prototype of a new water information system for California in the American R. basin, referred to as H2O2.0. 3

4 Relevance to RFP. This project addresses critical knowledge gaps surrounding the provision of ecosystem services from the Sierra Nevada within the context of sustainable natural ecosystems, with a particular emphasis on water as a high-value ecosystem service. This project is central to two initiatives identified in ANR s Strategic Vision 2025, which identifies scientific, technological, social, and economic demands facing California where ANR can have an impact. Of the five initiatives in the strategic vision, this proposal addresses the critical intersection of two: i) sustainable natural ecosystems and ii) water. The intersection comes about because water not only constitutes a high-economic-value ecosystem service in California s mountain forests, but also is central to most other ecosystem services in this and other California ecosystems. Evaluation of other ecosystem services must consider the central role of water-vegetation interactions in managed forests in order to undertake any meaningful evaluation of tradeoffs, thresholds, and feedbacks associated with forest health, carbon sequestration, fuels management and other services. This project aims to provide both a template for how to achieve synthesis around managing for multiple services, for how management actions affect tradeoffs for water-related and other services, and for enhancing decision-making through better information. The multidisciplinary project team plans to develop integrated products through wellcoordinated outcomes. For example the water-cycle results will be linked to silvicultural outcomes, and the silviculture assessments will be informed by water-cycle responses to climate and vegetation changes. The project maximizes well-established ANR strengths in forest management and forest health, while leveraging other non-anr UC strengths in mountain hydrology and water resources. This project is the next step in leveraging significant funds through state, federal, local and NGO sources for both research and implementation projects. State and foundation monies supported the planning grant used to develop this project. It is expected that further funding will come through future bond funds. This project is the critical next step in preparing for that. SB2x (water bond), which may still go before the voters, would provide an opportunity for funds. It is also planned to continue discussions already initiated with other Sierra Nevada stakeholders, including state and federal agencies, water and power providers, NGOs, and other private-sector interests. Though no ongoing or planned research directly addresses the objectives of this proposed project, there are synergistic efforts underway with which the team can share data and leverage findings. As described in the work plan, the team is committed to an integrated, collaborative approach that engages with ANR research and extension scientists, and other relevant university and agency personnel, to put this project into a broader context. The state critically needs to adopt new approaches to water and forest management that consider the changing climate and other pressures. For UC, this means addressing important knowledge gaps and engaging with resource managers to integrate new knowledge into practice. The research under this project is of strategic benefit to California. The project team will both use our established channels of communication and develop sustainable means for bi-directional communication to bring project results to stakeholders and support science-based decision-making. Qualifications of project team. The project team is led by two senior ANR faculty members (O Hara & Battles) plus draws on an extension specialist (Stewart) and a natural resources advisor (Kocher). Water expertise is provided by a non-anr UC faculty member (Bales), who established a well-regarded research program on Sierra Nevada hydrology, biogeochemistry and resourcemanagement issues over 25 years ago. O Hara works on leaf area and forest productivity relationships and has pioneered efforts to guide silviculture using a LAI allocation approach. Battles is a leader in ecosystem ecology, particularly in the Sierra Nevada, and of carbon dynamics in forests. Stewart specializes in carbon markets and forest management and valuing forest ecosystem services. Kocher works on water quality, forestry and wildlife issues and coordinates public involvement 4

5 This team has worked together on two projects: i) the planning phase that stimulated this proposal (O Hara, Battles, Bales, Stewart), and ii) a parallel adaptive-management project focused on impacts of fuels management in the Sierra Nevada (Battles, Bales, Kocher). These two projects will continue to be linked by sharing fieldsites and data, and leveraging this efficiency for additional funding. 5

6 Project Narrative Literature Cited Birdsall, S.S., and J. Florin Outline of American Geography. US State Department. Bosch, J. M., and J. D. Hewlett A review of catchment studies to determine the effect of vegetative changes on water yield and evapotranspiration. J. Hydro. 55:3-23. Brown, AE, Zhang, L, McMahon, TA, Western, AW, and RA Vertessy (2005). A review of paired catchment studies for determining changes in water yield resulting from alterations in vegetation. J. Hydro. 310: Department of Water Resources Managing an Uncertain Future: Climate Change Adaptation Strategies for California's Water. Sacramento, CA. Oct. 34 p. Department of Water Resources Climate change characterization and analysis in California water resource planning studies. Final Report. CA Natural Resources Agency. Hornbeck, J.W., M.B. Adams, E.S. Corbett, E.S. Verry, J.A. Lynch Long-term impacts of forest treatments on water yield: a summary for northeastern USA. J. Hydro. 150: Kapnick, S. and A. Hall Observed climate snowpack relationships in California and their implications for the future. J. Climate 23: Kattelmann, R. C., N.H. Berg, and J. Rector The potential for increasing streamflow from Sierra Nevada watersheds. Water Res. Bull. 19(3): Miller, N. L., J. Jin, and others An analysis of simulated California climate using multiple dynamical and statistical techniques. CA Energy Commission: CEC F. National Research Council Hydrologic Effects of a Changing Forest Landscape. Committee on Hydrologic Impacts of Forest Management. National Academies Press. Washington, D.C. Raumann, C.G., and C.E. Soulard Land-cover trends of the Sierra Nevada Ecoregion, U.S. Geological Survey Scientific Investigations Report Sedell, J., M. Sharpe, D. D. Apple, M. Copenhagen, and M. Furniss Water and the Forest Service. USDA For. Serv. Policy Analysis. Washington, DC. Stewart, W.C Economic Assessment of the Ecosystem. In Sierra Nevada Ecosystem Project: Final Report to Congress, vol. II, Assessments and Scientific Basis for Management Options. Davis; UC Center for Water and Wildland Resources. Vicuna, S Predictions of climate change impacts on California water resources using CALSIMII: A technical note. CA Energy Commission: CEC SF. SF.PDF Westerling, A. L., H. G. Hidalgo, D. R. Cayan, and T. W. Swetnam Warming and earlier spring increase western US forest wildfire activity. Science 313:

7 Project Plan Specific aims. This research aims to build the information infrastructure for sustainable ecosystem management of Sierra Nevada forests, based on valuation and tradeoffs among complementary and competing ecosystem services. In particular, this research emphasizes high-value water-related services, which have the potential to finance forest management for multiple benefits. The three main components of the project plan, their aims, and expected outcomes follow. 1. Field measurements. The aim is to collect the observational data needed to support the development of forest treatment/management scenarios through modeling and analysis. Expected outcomes are key improvements to the knowledge base of snow accumulation and melt, soil moisture, ET, stream discharge, LAI and the linkage of these attributes to climate, soils and physiographic features of the landscape. A further outcome will be a measurement framework to quantitatively enumerate ecosystem services to support potential future public and private investments. 2. Meta-analysis and modeling. The aim is to integrate data through analysis and modeling of how the forest vegetation and water cycle will respond to climate change and management actions. Expected outcomes are scientific findings around ET, water-use efficiency, water deficit, water storage, and water yield as a function of climate, vegetation and physiographic characteristics of the Sierra Nevada landscape, and a predictive ability that is relevant for forest management. A further outcome will be the design for a watershed treatment experiment designed to optimize water-based ecosystem services. 3. Evaluation of ecosystem services. The aim is to use the American R. area as an informative case study for measuring and valuing water-based ecosystem services, assessing the impact and importance of those services to local stakeholders, and determining competition between services. Expected outcomes include approaches and metrics for valuing and assessing ecosystem services. A further outcome is stakeholder participation in subsequent pilot projects to manipulate forests to optimize for water storage and yield. In addition to these strategic aims, several scientific objectives will be addressed that contribute to our understanding of not only the Sierra Nevada forests, but forests in other semi-arid landscapes. 1. Determine rates of ET in Sierran mixed-conifer/true fir forests. It has recently become apparent that ET estimates based on ecological modeling versus those based on field measurements of hydrologic water balance differ by 50% or more. It is also apparent that the ET and net primary productivity (NPP) values for Sierran forests do not correspond to the wellestablished near-linear relationships between NPP and ET. 2. Determine water use efficiency of trees and shrubs in Sierran mixed-conifer and true fir forests. The carbon gained per unit of water transpired (i.e., water use efficiency) is an overarching silvicultural question. We need to learn how to protect species diversity and maintain a healthy forest given climate change, while also minimizing the use of in-forest water use and maximizing water yield. Essential to this task is quantifying differences in water use efficiency among the common forest species. 3. Determine the potential for forest management to delay snowmelt in Sierran forests. Storage of water in the snowpack attenuates streamflow. Delaying runoff later into the summer drought season has value to multiple ecosystem services. While it is well known that changes in the magnitude and distribution of LAI influence the forest energy balance and thus snowmelt, prescriptions specific to the heterogeneous landscape of the Sierra Nevada and their impact have yet to be quantitatively assessed with hydrologic modeling. 4. Determine potential economic tradeoffs of forest management treatments to affect water yield and ecosystem services. Forest management interventions designed to increase late season runoff and related environmental benefits will need to be substantial enough to produce measurable and predictable changes to justify the development and enforcement of contracts for the additional services. In addition to estimating the on-site costs and benefits, the contractual 7

8 arrangements between the producers and willing buyers of variable quantities of goods and services will be characterized. 5. Involve stakeholders in decision-making regarding forest management and watershed effects. Over the past 2-4 years, stakeholders concerned with Sierra Nevada ecosystem services (primarily water- and wildlife-related services(, have become sensitized to the potential synergies and conflicts between these services, and the potentially large economic benefits to be gained from better information to support decision-making. For example, the investigators on this project have collectively given talks and attended meetings to address these issues an average of at least twice monthly and often several times a month over the past 2-4 years. It is expected that this project, which builds on the planning done under the SWEEP project, will spawn an on-the-ground set of forest treatments. One output of our work will be a site-specific plan, supported by stakeholders, to carry out the thinning and associated assessments at the scale of headwater catchments (~1 km 2 ). This project will provide the needed bridge between the SWEEP planning support and the expected distributed support from state and local sources for the treatment and follow-up research. Science background and preliminary studies. Paired-catchment studies provide historical evidence regarding the effects of vegetation management on forest hydrology (NRC 2008). During the past 60 years, literally hundreds of studies have been conducted worldwide, with results summarized in a series of reviews (e.g., Bosch and Hewlett 1982, Hornbeck et al. 1993, Stednick 1996, Brown et al. 2005). However no paired-watershed study has been conducted in the conifer forests that dominate the west slope of the Sierra Nevada. Nor is the existing information sufficient to support process-based modeling of the causes of observed watershed responses. Perhaps the most significant consideration is that past studies generally imposed a treatment once and then allowed the forest to regrow (Hornbeck et al. 1997). Reports of effects on water yield are typically based on the first five years following treatment (Brown et al. 2005). However the recovery of forest vegetation, ET and runoff to preharvest levels can be rapid unless regrowth is suppressed (Hornbeck et al. 1997). It is this rapid forest recovery that limits the practicality of managing the forest for water yield. Repeated entries into a watershed to maintain the vegetation structure takes time and money, an expense that is not always warranted by the value of the water (Hornbeck et al. 1997). Moreover, vegetation suppression has consequences to key ecosystem attributes such as productivity, nutrient cycling, diversity, wildlife habitat, and sediment transport. Another important filter is to select studies done in similar forest types and comparable climates (Peel et al. 2010). The snow-dominated forests of the Sierra Nevada support temperate, evergreen, needle-leaved trees. They are productive forests that maintain large pools of live biomass with high leaf area. They also experience mild snowy winters and hot dry summers characteristic of Mediterranean regions. Based on generalizations from reviews of paired catchment studies, the Sierra Nevada conifer forest has ecological attributes that suggest a high potential for water-yield gains. First, forest catchments dominated by conifers consistently show greater per capita gains in water yield (mm of water yield/fraction of forest cover removed) than any other forest type. For example, Bosch and Hewlett (1982) found per capita water yield in temperate conifer forests was on average 60% greater than in temperate deciduous forests. As shown by Zhang et al. (2001), changes in water yields depend on the amount of precipitation (ppt). In extremely dry ecosystems (< 500 mm yr -1 ppt) and extremely wet ecosystems (> 1500 mm yr -1 ppt) there is a limited ability to affect water yield by manipulating the vegetation. In the high Sierra, total annual precipitation ranges from a low of about 600 mm in the south to a high of over 1800 mm in the north (Bales et al. 2006). Thus in terms of input, the Sierra Nevada spans a range where there is a near-linear increase in water yield with reductions in forest cover (Zhang et al. 2001). Finally, in snow-dominated systems there is clear 8

9 seasonality in the water-yield response to treatments, with the greatest absolute increases observed during snowmelt but the greatest proportional increases observed during the dry summer months (Brown et al. 2005). The duality in response suggests that upstream forest management can help fill downstream reservoirs in the spring as well as support baseflow during the summer. Thus from qualitative arguments, it seems that there is the potential to manage for water in the Sierra Nevada. Further, using a simple quantitative method developed by Zhang et al. (2001) that is based on precipitation and vegetation cover, we estimated that a 30% reduction in forest cover (e.g. from 90% to 60%) for a Sierra Nevada watershed would increase water yield by 9%.Given the limitations of the data on which Zhang et al. (2001) had for developing their correlations, we feel that this number could be higher. The seasonal snowpack in the Sierra Nevada is a critical component of this water balance. At lower elevations, the snowpack melts shortly after being deposited, but at higher elevations the snowpack typically accumulates from December until March or April, and then melts from April through May-July. Thus the lag between precipitation and discharge depends on elevation, and is about two months in catchments where only about 50% of the precipitation fell as snow (Hunsaker et al. submitted). In general, snow melts out about 20 days later for each 300-m increase in elevation (Bales et al. 2010). Stream discharge from these catchments varies from 10% of precipitation in a more rain-dominated catchment in a dry year, to over 60% in a snow-dominated catchment in a wet year. ET accounts for most of the precipitation not leaving the catchments as discharge. At least in the lower-elevation catchments, trees transpire year round, drawing water from both soil and deeper regolith during the dry summer when ET is highest (Bales et al. submitted). Forest treatments and climate warming will affect both the lag between precipitation and runoff, and the water yield. These changes will be mediated through shifts in the timing of snowmelt and the amount of water evaporated, transpired or sublimated. Snowmelt is driven by temperature and vapor density gradients within the snowcover caused by heat exchange at the snow surface and at the snow soil interface (Marks et al. 1999). Forest cover influences energy exchange to snow, effectively decoupling the above-canopy and sub-canopy atmospheres and suppressing turbulent-energy fluxes (Link and Marks 1999). Thus the energy balance on sub-canopy snow is dominated by radiation. The canopy modifies shortwave irradiance through shading and longwave irradiance via thermal emissions (Link et al. 2004). Forest cover may also affect sub-canopy shortwave radiation by altering snow surface albedo through deposition of tree litter (e.g. twigs, leaves) on snow (Melloh et al. 2002). While the tools are readily available to calculate forest energy balances, the necessary data are not. Tree cover does significantly reduce snow accumulation. In a boreal forest in Canada, there was 30-50% less snow under the canopy compared to nearby clearings (Gelfan et al., 2004). Sublimation of canopy snow has been shown to be a primary factor controlling forest snow losses (Parviainen and Pomeroy, 2000). The energy balance may be different in the Sierra Nevada given the warm winters characteristic of Mediterranean climates. Data from two water years at our sites in the southern Sierra Nevada show significant canopy effects on snow accumulation (Bales et al., submitted). Propagating treatment effects far enough downstream to be measured and therefore to be meaningful for end users requires a large portion of the watershed to be treated. Applying this idea to Coon Creek, a 16.7 km 2 watershed in Wyoming, Troendle et al. (2001) found that removing 24% of the vegetation led to a significant water yield increase of 7.6-cm. Yield increases have been shown to be minimal or negligible during years with drier than normal precipitation and maximized during wet years. 9

10 The hydrologic mechanism for increasing water yield by removing vegetation is attributed to a reduction in ET, both evaporation of intercepted precipitation and transpiration of water by vegetation. How ET losses are balanced between these two processes greatly depend on climate. As Storck et al. (2002) noted, models of energy balance for under-canopy snow accumulation and ablation are well tested for cold regions, but not for more maritime climates like the Sierra Nevada. Within the Sierra Nevada, the Central Sierra Snow Lab, Onion Creek Experimental Forest, Yuba Pass, and Swain Experimental Forest have all reported efforts to study forest treatment impacts on snow accumulation. The Central Sierra Snow Lab found the lowest increases in snow accumulation from selective cutting of red fir. A reduction in crown cover from 90% to 50% resulted in a 5% increase in SWE (Anderson 1974). The highest percentage increase (~50% SWE) resulted from strip cuts. The highest absolute increase (48 cm) resulted from a wall-and-step forest structure designed to maximize snow Ecological versus hydrologic water deficit The partitioning of snowmelt and rain into ET versus discharge depends in part on whether or not there is a water deficit. Index values of water deficit have been widely used to estimate tree recruitment and mortality, based on standard assumptions of short grass vegetation, up to 2 m soil depth and monthly average precipitation and temperature values (e.g. Lutz et al., 2010). Hydrologic data based on measurements from a slightly higher-elevation mixed-conifer forest show a lack of deficit, with ET coming from subsurface storage (soil plus deeper regolith) during the period when supply is near zero (Bales et al., submitted). Together, these observations point to three knowledge gaps that bear investigation in order to design a strategy for forest treatments to maintain or enhance water yield and runoff timing. 1. Water deficit should be defined spatially, or at least the elevational dependence of water deficit defined, based on the actual vegetation, water supply, and water storage estimates. Water deficit should be further investigated as a function of LAI, using hydrologic modeling constrained by LIDAR, satellite, and field measurements. In the absence of complete data, water deficit indices should be developed. 2. LAI should be estimated across basins under consideration for treatments, to identify areas with the greatest potential for LAI reduction and identify magnitudes of reduction that may be achieved. Currently available canopy density values should be verified. 3. For candidate areas for forest treatment, an integrated program of measurements and modeling should be used to better constrain the water balance, including the magnitude and timing of snowmelt and streamflow. accumulation and delay melt (Anderson 1956). Results from all other types of forest harvesting block cutting, commercial selection, selective cutting and clearcutting increased SWE in the treated areas between 14 and 34%. Evidence that these effects on snow accumulation can be long lasting are described by McGurk and Berg (1987), who revisited strip cuts at Yuba Pass 20 years after harvest and found sustained increases in SWE of 25-45%. While most of the increase in water yield is concentrated around removal of the trees themselves, additional factors may also affect the water balance. Tree species vary in their efficiency of water use (grams of carbon photosynthesized/gram of water transpired). The compositional changes related to harvest and recovery dynamics can contribute to longer-term changes in the water balance. For example, based on the Hornbeck et al. (1997) results, we calculated that 16% of the observed decrease in ET over 15 years following harvest was related to differences in water use efficiency (WUE) in the constituent tree species. Understory management may also be an important factor in modifying transpiration effects on water yield. From their research in Sierran conifer forests, Royce and Barbour (2001) found that understory shrubs may deplete soil moisture faster, consuming more available soil moisture than conifers. Increasing temperatures from climate change may actually lead to a decrease in vegetation water use, as snowmelt occurs earlier and less late summer moisture is available (Tague et al. 2009). We also expect the increases in atmospheric CO 2 concentration to generally increase the WUE of plants but these effects will vary by species and tree size (Knapp and Soule 2011). Additionally, not removing slash post-harvest has shown to hasten snow ablation in the spring (Anderson and Gleason 1960), which also affects snowmelt timing. 10

11 Manipulations of the forest composition and structure through silviculture affect not only snow attenuation, ecosystem-wide water use efficiency, and water yield but also the timing of melt. Studies of forest impacts on snow properties in the Sierra Nevada have taken place since methods were developed to measure snow and water content in the early 1900 s (Church 1933), although never implemented on a large scale (McGurk and Berg, 1987). Hornbeck and Smith (1997), in their water-resources decision model for forest managers in the northeastern U.S., suggest even-aged management as the best method for increasing water yield. Ultimately, the effectiveness of any treatment depends on individual stand tree height, slope, and aspect to obtain the right mixture of opening large enough to accumulate additional snow, with enough shading to block direct solar radiation for prolonging ablation. Treatment options are also limited by social and environment constraints. Leaf area index (LAI), or the ratio of cumulative foliage surface area projected downward per unit of ground beneath the canopy, provides a variable with the potential to be an integrative tool for a variety of forest management applications. LAI is a key driver affecting a variety of ecosystem processes such as light interception, photosynthesis and hydrologic processes such as canopy interception and ET. Forest management, by affecting canopy cover and canopy density, alters the amount and arrangement of LAI thereby affecting these processes. An example of a management application based on LAI is the Multiaged Stocking Assessment Model (MASAM) that uses a LAI-allocation approach to control stocking and design stand structures in multiaged stands (O Hara and Valappil 1999). This model is proposed for development in the Sierra Nevada (North et al. 2009). By allocating LAI to design stand structures, managers control these ecosystem processes. The MASAM tool can then be used to design and assess treatments in this study. We calculated LAI dynamics in managed forests at the UC Blodgett Forest and in the unmanaged Onion Creek Experimental Forest. At Blodgett, we reconstructed LAI based on historical inventory data for stands that had received various silvicultural treatments in the last 30 years. The highest total LAI reached 12 by the last inventory in Compartments with no management showed increasing LAI to the present. LAI development on these stands cut 80 to 100 years ago is therefore still increasing. In compartments receiving harvest treatments, maximum LAI approached 9 with differences due to the intensity of the treatment and the time since last entry. All compartments included multiple measurements prior to any harvest activity and showed increasing LAI until the harvest treatment. Compartments that received harvest treatments had LAI recovery rates after treatment that ranged from approximately 1.5 to 2.0 LAI per decade. The unharvested compartments had rates of increase in LAI of approximately per decade. This difference was largely due the lower LAI and higher vigor of residual trees in the harvest treatments as compared to the untreated stands which are assumed to be approaching a maximum LAI level. Recovery following timber harvest likely follows a logistic growth form where the increase is slow immediately after harvest and slow late in development as the stand approaches a maximum. In between these extremes, LAI recovery is probably very rapid. From these preliminary LAI estimates, we may assume that a treatment in a similar uncut stand that reduced LAI from 12 to 8 might take approximately 25 years to recover. A treatment that reduced LAI from 12 to 4 might require more than 50 years for recovery. Since Blodgett represents the upper end of a site productivity range, these estimates do not apply to poorer quality sites. Poorer sites may reach lower maximums, and recovery may be slower than on more productive sites. LAI projections were based primarily on leaf area prediction equations from Gersonde (2003) that were developed in mixed-conifer stands in the central Sierra Nevada. We also conducted a survey of seven adjacent watersheds in the upper Onion Creek basin (Fig. 1). These forests are typical of mature, minimally disturbed forests of the northern Sierra Nevada. 11

12 The upper watersheds at Onion Creek were dominated by red fir (49%) and white fir (29%). The forest is dense with an average of 501 trees per ha and mean basal area equal to 88.9 m 2 ha -1. Average canopy cover is 51%, but ranged by watershed from a low 32% to a high of 68%. The average canopy tree height was 18 m but the tallest trees exceeded 32 m in height. The average age of a canopy-sized tree was 100 years. LAIs were estimated using these plot data and the Gersonde (2003) equations for leaf area prediction from individual trees. In terms of LAI, variation between watersheds is also quite high with a range from 5.4 to Some of this variation may be due to differences in species composition disturbance history, or site productivity. The high LAIs in these watersheds may be due to the prevalence of large amounts of shade-tolerant conifers that typically have high LAIs. For example, all the plots had large amounts of red and white fir. Another possibility is that LAI is being over-predicted in these higher elevation stands. An initial step in the proposed research is refining and testing LAI prediction equations for study sites and for all species present. Economic significance. The primary economic benefits from water timing and yield enhancements come from hydropower and water supply, with high-elevation hydropower being potentially the more significant. Between 1950 and 1970, the reservoir capacity of hydroelectric systems in the Sierra Nevada doubled as most high-value sites were utilized. Most of these projects have just gone through their 50-year FERC renewal or are in the process of renewing their licenses. Increased interest in improving in-stream habitats, producing carbon-free energy, and better overall management of water in California are bringing greater attention to innovation across the whole hydrologic system. An intriguing area of inquiry is whether the costs of altering forest-stand ET, snow retention, and fire risks could be partially or totally covered by the economic value derived from increased water flow through existing turbines. In addition, capturing the complementary value of the reduced fire risk associated with forest stands dominated by large-diameter trees could complement the economic value of the additional water runoff. The institutional challenges of getting beneficiaries to voluntarily pay for new services will require innovative contracts that accurately account for the episodic nature of the delivery of the varied services, and credible validation of benefits. The hydroelectric value of runoff was roughly equal to the diversion value of water from the Sierra Nevada in the 1990s (Stewart 1996). The economic value of incremental hydroelectric power depends on the spot price (higher on hot summer days when all air conditioners are running) and whether it qualifies for a renewable-energy price premium that is given to some forms of hydroelectric power. The economic value of additional irrigation or residential water depends on the willingness to pay (high value nut crops can pay more than alfalfa for the same amount of water), the potential losses from not having the water available at a certain time, and the cost of alternative supplies. While current water and power markets are controlled by long-term contracts and pricing, it is possible that marginal pricing will be introduced that will allow innovators to capitalize on linking newly created supplies with potential buyers. Identifying the watersheds with a high potential for increased supplies as well as downstream users with a high willingness to pay for additional supplies can break out of the historical problem where average interventions valued at average prices did not warrant any new programs. It is the watersheds with estimated values for additional runoff in the upper quartile where greatest attention should be focused. Many of the high value watersheds along the crest of the Sierra Nevada are legally zoned as wilderness as reserves. However, the American R. and San Joaquin r. stand out as areas where the potential for additional revenue is significant and where a considerable fraction of the watersheds are managed for multiple 12

13 benefits by private and public entities. Focusing research and implementation on specific, high-potential areas rather than on a general policy for all lands managed by government agencies should increase the probability of producing relevant results in the next decade. Design and methods. The proposed research is designed around the three components noted above: i) field measurements, ii) data analysis and modeling, and iii) evaluation of ecosystem services. Our approach in each of these three areas follows. Field measurements will complement those made by other programs in progress, and focus on end-member forests (higher elevation, burned and high density) and measurements (ET and LAI) that are largely Figure 1. Location map for field measurements, in Middle and North forks of American R.. Five areas noted by circles and labeled in red. Portion of SNAMP study area shown in lower left. absent from those programs. The proposed higher-elevation catchment for measurements is within the SNAMP study area, in the headwaters of Deep Creek (Figure 1). There is a combined SNAMP/H2O2.0 instrument cluster above the catchment, but no instrumentation within it. The burned catchment for monitoring is in the headwaters of Duncan Creek; although denuded in the 2001 Starfire, has some regrowth of shrubs. The dense-forest catchment is Onion Creek, an experimental forest jointly administered by UC and the USFS. Measurements will follow the protocols already in use at five locations we have previously instrumented ( and will include snow depth using ultrasonic sensors, placed under the canopy as well as at the drip edge and in the open. Stream stage will be measured using a pressure transducer, and ET using sap flow. Methods are described in Bales et al. (2011). Self-logging pressure transducers will be placed in catchment streams. Sap flow will be measured in the dominant species at 10 nodes, with snow-depth and soil moisture placed to sample physiographic variability at 20 nodes. All installations will be on solar, with the nodes in each catchment connected with wireless radios to a base station. Data will be logged at both the nodes and base station. Leaf area will be estimated using individual tree leaf area equations. A summation approach will used to scale-up leaf area measurements from twig to tree after determination of specific leaf area and foliage weight from individual trees. Independent variables for leaf area prediction will include sapwood cross-sectional area, dbh, and crown length. Leaf area equations will be generated for each species and with site quality variables if necessary. These equations will then be used to predict LAI in inventory plots in catchments under study and to calibrate a MASAM model (O Hara and Valappil 1999). 13

14 Hydrologic modeling will use RHESSys, a process-based coupled model of ecohydrologic interactions (Tague and Band, 2004). RHESSys has been successfully used to estimate climaterelated changes in streamflow, snow and vegetation water use and carbon flux for a variety of watersheds in the western US and European Alps (Tague and Grant, 2009; Zierl et al., 2007). We are using it at our previously instrumented catchments in the Sierra Nevada; and those parameterizations will be directly relevant for this project. Our approach to modeling the response of soil water, ET and streamflow to vegetation and climate variation/change will account for complex heterogeneity in snow accumulation/melt, ET, and soil drainage. Assimilation of satellitederived snow products into RHESSys will support this analysis and link simulations of snowmelt with changes in land cover and soil-water storage. We have ground-based, satellite (snowcover) and aircraft (LiDAR ) data from 2 SNAMP catchments and 3 CZO headwater catchments for the metaanalysis (Figure 1). Model parameterization will build on that done for these previously instrumented catchments. The evaluation of ecosystem services will begin with data collection on the actual costs and revenues from the forest management project activities as well as an extrapolation of the potential costs and revenues for treatments applied at a commercial scale of operation. The economic value of the estimated increased late season flow will be calculated from the marginal value of the water as it runs through sequential hydroelectric turbines and is then available for diversion to agricultural or urban water districts. The value of the increased late season in-stream flow will be estimated by the implicit price of contractual obligations in the relevant FERC contracts as well as through sets of structured interviews with water managers and environmental consultants engaged in monitoring instream flows. The project data will then be used to parameterize a more generic model to estimate the benefits and costs on systems with different sized treatment areas, number of turbines, and lengths of streams with improved conditions. Response to RFP criteria 1. Clearly define the strategic initiative addressed. There is a critical knowledge gap surrounding the provision of water as a high-value ecosystem service from the Sierra Nevada within the context of sustainable natural ecosystems. 2. Opportunity to maximize ANR s strengths. The well-established ANR strengths in forest management and outreach will leverage other non-anr UC strengths in mountain hydrology and water resources. 3. Provide well-coordinated outcomes for multidisciplinary projects. The project team plans to develop integrated products. For example the water-cycle results will be linked to silvicultural outcomes, and the silviculture assessments will be informed by water-cycle responses to climate and vegetation changes. 4. Ability to leverage additional funding. State and foundation monies supported the planning grant used to develop this project. It is expected that further state funds would be available for both research and implementation projects through future bond funds. CA Senate Bill 2x (water bond), which may still go before the voters, would provide an opportunity for funds. It is also planned to engage in discussions with other Sierra Nevada stakeholders, including state and federal agencies, water and power providers, NGOs, and other private-sector interests. 5. Build-on the research continuum with science that will benefit California. The state critically needs to adopt new approaches to water and forest management that consider the changing climate and other pressures. For UC and ANR, this means addressing important knowledge gaps and engaging with resource managers to integrate new knowledge into practice. 6. Support science-based decision making and delivery of useful findings to support policy and outreach efforts. The UC faculty will engage the forestry advisors in the counties, work with USFS and CA Resources Agency personnel to ensure the research is focused and is disseminated to end-users. 14

15 7. Use an integrative approach to collaborate with other strategic initiatives. The team is committed to engaging with ANR research and extension scientists to put this project into a broader context. Organizational plan. The PI (O Hara) will have overall administrative and scientific responsibility for the project, with team members having lead responsibility for research related to the science (Battles) and economic (Stewart) aspects of ecosystem services, for hydrology (Bales), for silviculture (O Hara), and for outreach (Kocher). Team members will co-supervise the postdocs and graduate student, to help assure an interdisciplinary approach to the work. The PI will also be responsible to convene frequent team meetings (both remote and in person) to help integrate science and keep the project on track and on schedule. Project Plan Literature Cited (excludes references listed in Project Narrative ) Anderson, HW Managing California s Snow Zone Lands for Water. US Forest Service Research Paper, PSW-6. USDA Forest Service Pacific Southwest Forest and Range Experiment Station; Berkeley, CA: 28p. Anderson, HW and CH Gleason Effects of logging and brush removal on snow water runoff. Int. Assoc. Sci. Hydrol. Publ. 51: Anderson, HW Forest-cover effects on snowpack accumulation and melt, Central Sierra Snow Laboratory. AGU Trans. 35(2): Bales, R., NP Molotch, TH Painter, MD Dettinger, R Rice, and J Dozier Mountain hydrology of the western United States, Water Resour. Res., 42:W08432, DOI: /2005WR Bales, RC, JW Hopmans, AT O Geen, M.Meadows, PC Hartsough, P Kirchner, C Hunsaker, and D Beaudette. submitted. Soil moisture response to snowmelt and rainfall a Sierra Nevada mixed conifer forest. Bales, RC, R Rice, X Meng Spatial distribution of snow water equivalent across the central and southern Sierra Nevada, presented at 2010 Fall Meeting, AGU, San Francisco, Calif., Dec., Abstract C33E Bales, RC, M Conklin, B Kerkz, S Glaser, JW Hopmans, C Hunsaker, M Meadows and PC Hartsough Soil moisture response to snowmelt and rainfall a Sierra Nevada mixed conifer forest. In D Levia, et al. (ed.) Forest Hydrology and Biogeochemistry: Synthesis of Research and Future Directions. Springer-Verlag, Heidelberg, Germany. (in press). Brown, AE, L Zhang, TA McMahon, AW Western, and RA Vertessy A review of paired catchment studies for determining changes in water yield resulting from alterations in vegetation. J. Hydrol. 310: Church, JE Snow surveying: its principles and possibilities. Geog. Rev. 23(4): Gelfan, A., Pomeroy, JW, and L Kuchment Modelling forest cover influences on snow accumulation, sublimation and melt. J. Hydrometeorol. 5: Gersonde, RF Developing a hybrid growth mode for multiaged Sierra Nevada mixed-conifer stands. Ph.D. Dissertation.University of California - Berkeley, CA. Hornbeck, JW and RB Smith A water resources decision model for forest managers. Agric. For. Meteor. 84: Hornbeck, JW, CW Martin, and C Eager Summary of water yield experiments at Hubbard Brook Experimental Forest, New Hampshire. Can. J. For. Res. 27: Hunsaker, C, T Whitaker and RC Bales. Submitted. Snowmelt runoff and water yield along elevation and temperature gradients in California s southern Sierra Nevada. J. Am. Water Resour. Assoc. Knapp, PA and PT Soule Increasing water-use efficiency and age-specific growth responses of old-growth ponderosa pine trees in the Northern Rockies. Global Change Biol.17: