However, more work is required to determine how to better represent coupled microbe–plant nutrient competition and transport limitations. For example, we have previously shown that one can apply a homogenous soil environment assumption and include the substrate diffusivity constraint in the ECA competition parameters. In this approach, the diffusivity constraint can be directly integrated into the substrate affinity parameter. The “effective” KM would be higher than the affinity measured, e.g., in a hydroponic chamber . We hypothesize that our calibrated KM value, which led to an excellent match with the observations , effectively accounted for this extra diffusive constraint on nutrient uptake. A second approach would be to explicitly consider fine-scale soil fertility heterogeneity, explicitly represent nutrient movement , and apply the ECA framework at high resolution throughout the rhizosphere and bulk soil. However, to test, develop, and apply such a model requires fine-scale measurements of soil nutrient concentrations,ebb flow tray microbial activity, and rhizosphere properties and dynamics; model representation of horizontal and vertical root architecture and microbial activity; effective nutrient diffusivities; and potentially, computational resources beyond what is practical in current ESMs.
Yet, there is potential value in this approach if we can produce a reduced order version of the fine-scale model that is reasonable and applicable to ESMs. A third approach, of intermediate complexity, would be to simplify the spatial heterogeneity of root architecture, soil nutrient distributions, and nutrient transport. Roots could be conceptually clustered in the center of the soil column, where nutrients would become depleted and competition between microbes, roots, and abiotic processes would occur. The “radius of influence” concept that defines a root influencing zone could be used to simplify heterogeneity, with CT5 competition applied to this root influencing zone. More model development, large-scale application, and model-data comparisons are needed to justify such an approach. As we argued above, the choice of nutrient competition theory used by ESMs faces a dilemma between necessary model simplification and accurate process representation. Our goal is to rigorously represent nutrient competition in ESMs with a simple framework that is consistent with theory and appropriate observational constraints while not unduly sacrificing accuracy. We conclude that our ECA nutrient competition approach meets this goal, because it is simple enough to apply to climate-scale prediction and is based on reasonable simplifications to the complex nutrient competition mechanisms occurring in terrestrial ecosystems.Over the past two decades, the ecological significance of anadromous Pacific salmon has been well documented in aquatic ecosystems throughout the Pacific coastal ecoregion.
The annual return of salmon to fresh waters and the associated decomposition of post-reproductive carcasses result in the transfer of marine-derived nutrients and biomass to largely oligotrophic receiving ecosystems. Such inputs have been shown to increase primary production , invertebrate diversity and biomass , and fish growth rates . Since rates of primary production are typically low in many coastal salmon-bearing streams, even small nutrient inputs from anadromous salmon may stimulate autotrophic and heterotrophic production and produce cascading effects through the entire aquatic food web. In addition to subsidizing riverine biota, salmon-borne MDN also benefit vegetation within the riparian corridor . Marine nutrients are delivered to the terrestrial environment via deposition of carcasses during flood events, absorption and uptake of dissolved nutrients by riparian vegetation and removal of carcasses from the stream by piscivorous predators and scavengers . Empirical studies have shown that as much as 30% of the foliar nitrogen in terrestrial plants growing adjacent to salmon streams is of marine origin and that growth rates of riparian trees may be significantly enhanced as a result of salmon-derived subsidies . Nitrogen availability, in particular, is often a growth-limiting factor in many temperate forests and annual inputs of marine-derived nitrogen may be critical to the maintenance of riparian health and productivity in Pacific coastal watersheds. Our understanding of the ecological importance of MDN subsidies has been greatly advanced by the application of natural abundance stable isotope analyses.
A biogenic consequence of feeding in the marine environment is that adult anadromous salmon are uniquely enriched with the heavier isotopic forms of many elements relative to terrestrial or freshwater sources of these same elements. When fish senesce and die after spawning, these isotopically heavy nutrients are liberated and ultimately incorporated into aquatic and terrestrial food webs. Our research has determined that the stable nitrogen isotope “fingerprint” of adult anadromous salmon returning to coastal California basins is 15.46 ± 0.27‰ , a value markedly higher than most other natural N sources available to biota in coastal watersheds. This salient isotopic signal allows the application of stable isotope analyses to trace how salmon-borne nutrients are incorporated and cycled by organisms in the receiving watersheds. Researchers interested in the utilization of MDN by riparian trees have generally inferred sequestration and incorporation from foliar δ15N values. Nitrogen is a very minor constituent of wood cellulose and natural abundance levels of δ15N in tree rings have rarely been determined. Poulson et al. were among the first to successfully analyzed δ15N from trees , but analysis required combustion of prohibitively large quantities of material per sample. Since that time, advancements in stable isotope analytical techniques have made it possible to detect δ15N from small samples of material . This permits for non-destructive sampling of live trees via increment cores and provides a novel opportunity to assess the transfer of salmonderived nitrogen into the riparian zones of salmon-bearing watersheds. Reimchen et al. recently reported that wood samples extracted from western hemlock trees in British Columbia yielded clear evidence of SD-nitrogen incorporation with reproducible δ15N values. Intuitively, if spawning salmon represent a significant source of nitrogen to riparian trees in salmon-bearing watersheds,flood and drain tray information on salmon abundance may be recorded in the growth and chemical composition of annual tree rings. By quantifying the nitrogen stable isotope composition of tree xylem it would be possible to explore changes in SD-nitrogen over decadal or sub-decadal time increments and determine whether the nutrient capital of riparian biota has been affected by diminished salmon returns. Moreover, if δ15N can be quantified from annual growth rings, it may be possible to infer changes in salmon abundance over time and reconstruct historical salmon returns for periods and locations where no such information presently exists. Nearly all salmon recovery programs are built upon very uncertain estimates of population sizes prior to European settlement. The development of robust paleoecological methods to determine historical salmon abundance and variability would greatly assist resource managers in identifying and establishing appropriate restoration targets. Our study was carried out in two geographically disparate coastal basins that are widely regarded as important salmonid spawning and rearing habitat in California. The initial method development phase of our research was conducted in the West Branch Mill Creek watershed, a fourth-order tributary of the Smith River in northern California . The WB Mill Creek drains an area of 23.6 km2 and supports five species of anadromous salmonids: coho salmon , fall chinook salmon , chum salmon , steel head trout , and coastal cutthroat trout . The WB Mill Creek watershed has an extensive history of timber harvest dating back to the mid 1800s, with concentrated commercial harvest of coast redwood and Douglas-fir occurring between 1950 and 2000. In 2002, a major portion of the Mill Creek watershed was acquired by Save-the-Redwoods League and subsequently incorporated into the California State Parks system.
Coast redwood trees presently dominate the upland forest while the riparian zone is comprised mainly of second-growth redwood and Douglas-fir, Bigleaf maple , willow , red alder , and California-laurel . The WB Mill Creek was selected for study because standardized annual salmon escapement surveys have been conducted since 1980 providing 23 years of continuous abundance data. The stream at this location is low gradient and the channel is largely unconfined with an extensive floodplain. Additional reach characteristics are described in Waldvogel . Our second study site was Waddell Creek, a small coastal watershed in Santa Cruz County, California . The headwaters of Waddell Creek originate in the Santa Cruz Mountains at an elevation of approximately 600 m and the creek drains to the Pacific Ocean ~27 km north of MonTherey Bay, California. Waddell Creek has two major forks, East and West Waddell, which join to form the main stem approximately 4.8 river km from the ocean. Approximately 84% of the Waddell Creek watershed is located within the boundaries of Big Basin Redwoods State Park. Waddell Creek supports steel head trout and one of the southernmost extant populations of coho salmon . Both anadromous species have access to the main stem of Waddell Creek, the lowermost 2.4 m of the East Fork Waddell and up to 10 km of the West Fork Waddell. High stream gradient and natural barriers restrict coho distribution in the West Fork, however, and they are scarcely reported upstream of the confluence of Henry Creek . For our initial pilot study we collected increment cores from 10 extant riparian Douglasfir trees growing adjacent to WB Mill Creek. All trees were located within 10.0 m of the active stream channel. Core samples were collected on 17 January 2004 from a 250 m section of riparian zone located immediately downstream of the 2.7 km index stream reach used by Waldvogel to derive minimum annual escapement estimates. Small diameter increment core samples were extracted from each tree and prepared for dendrological analysis using standard methods . We concurrently collected a second, large diameter , increment core from four trees for determination of annual nitrogen content and natural abundance stable nitrogen isotope ratios . Diameter at breast height , distance from the active stream channel, and general site characteristics were also recorded for each tree sampled. Increment core samples from Waddell Creek were collected on 16-18 October 2005. Collections were made from two distinct areas within the watershed: a ~750 m length of riparian zone adjacent to the creek where salmon spawning is known to occur , and a ~500 m section of riparian zone located above a natural barrier to salmon migration . We collected paired increment cores from a total of 10 Douglas-fir and 16 coast redwood trees in the Waddell Creek watershed . Distances from the stream channel, DBH, and site characteristics were also determined for each tree sampled. Two coast redwood cores from the upstream control site were later determined to be damaged or unreliable and excluded from our analyses. Small-diameter increment core samples were air dried, mounted, and sanded for analysis of annual growth rings. Prepared cores were converted to digital images and ring widths were measured to the nearest 0.001 mm using an OPTIMAS image analysis system. Each increment core was measured in triplicate and mean ring widths values were used to generate a time series for each tree. Each time series was then detrended using the tree-ring program ARSTAN to remove trends in ring-width due to non-environmental factors such as increasing age and tree size. Detrending was accomplished using a cubic smoothing spline function that preserved ~50% of the variance contained in the measurement series at a wavelength of 32 years. Individual growth index values were derived by dividing the actual ring-width value by the value predicted from ARSTAN regression equations. Chronologies using the growth index values were subsequently combined into a robust growth index series for each sample site. Cross-dating of coast redwood trees from Waddell Creek was not successful due to the presence of anomalous rings, a high degree of ring complacency in some cores, and small sample sizes. Previous dendrochemical research has found than nitrogen may be highly mobile in the xylem of some tree species . Although the degree to which coast redwood and Douglas-fir trees exhibit radial translocation of nitrogen is largely unknown, such mobility could potentially obscure interpretation of nitrogen availability at the time of ring formation . To minimize potentially confounding effects associated with the translocation of nitrogenous products across ring boundaries, increment core samples from Waddell Creek were pretreated to remove soluble forms of nitrogen following the “short-duration” protocol outlined in Sheppard and Thompson . Briefly, tree-ring samples were sequentially Soxhlet extracted for 4 h in a mixture of ethyl alcohol and toluene , 4 h in 100% ethyl alcohol, and 1 h in distilled water .