Ana səhifə

Predicting the effects of sea level rise and salinity changes on west coast tidal marsh plant and avian communities


Yüklə 7.88 Mb.
səhifə6/8
tarix24.06.2016
ölçüsü7.88 Mb.
1   2   3   4   5   6   7   8

Greenhouse experiment


To assess the salinity and inundation tolerance of tidal marsh dominant plants, we will conduct a common garden experiment with the same six plant species for which productivity measurements were taken. In February 2008, we will collect rhizomes of B. maritimus, S. acutus, S. americanus, S. foliosa, and T. angustifolia from a subset of sites and transport them to a greenhouse where they will be rinsed of native sediment, weighed, planted in sandy soil, and watered with freshwater for 3 months in order to minimize any negative effects of transplanting. S. pacifica will be started from seed since it is not as easily propagated from roots. Fertilizer will be applied once at the beginning of the acclimation period to aid in growth. After the plants have acclimated, ten pots of each species will placed into each of five 8 x 4 x 2 ft plywood containers lined with a double layer of plastic tarp. Two inundation regimes representing high and low marsh conditions will be implemented. These regimes will be executed by placing five replicates of each species on cinder blocks of equal height and the remaining replicates on the bottom of the container. Plants placed on the bottom will be inundated for longer periods than the plants placed on cinder blocks. Each container will be connected to its own reservoir via tubing and water will be pumped in ever 12 hours to mimic diurnal tides. The cinder block height and the water level in the container will be calibrated to result in designated tidal depths. The tidal amplitude will remain fixed throughout the experiment. Each container will receive a salinity treatment with a target level of 0, 4, 8, 16, or 32 ppt NaCl. Diluted sea water will be used and the concentration of salt will be increased by 4 ppt each week from the beginning of the experiment until the target water salinity is obtained. Weekly soil salinity measurements will be taken in a pot containing only soil and measured using a refractometer. A variety of plant characters will be measured weekly, including height, number of leaves, ramets, and inflorescences, among others, and the amount of senescence will be documented. The plants will be monitored for 6 months to evaluate effects of increased salinity and prolonged inundation on growth. At the end of the experiment, all surviving plants will be rinsed of all sediment, weighed, and dried to a constant weight to determine biomass.

Results of the greenhouse experiment will be used in conjunction with plant distribution data to identify salinity and inundation tolerances for each study species, and predict likely shifts in plant distributions under shifts in salinity and inundation regimes associated with climate change. These data will be used to fine tune model predictions under different climate change scenarios.


Spatial Modeling


Plant and avian survey data described above will be used to develop spatial models of species distribution, abundance, diversity, and productivity. Data collected by this proposal’s collaborators and institutions will be supplemented with publicly available species occurrence data (plants and birds) for the distribution modeling component. Plant distribution, abundance, diversity, and productivity predictions modeled from climate parameters will be used as inputs to vertebrate distribution and abundance models. Other intermediate inputs such as channel density will also be modeled from climate parameters.

Species Occurrence Inputs


Species distribution models will be constructed for tidal marsh plant and avian species, using two primary data sources: (1) vegetation sampling plots and bird surveys conducted by the principal investigators and their organizations (see Figure 2 for locations); and (2) public database records for special status species occurrence, including the California Natural Diversity Database (CNDDB; http://www.dfg.ca.gov/whdab/html/cnddb.html) and Jepson Herbarium on-line database (http://ucjeps.berkeley.edu/db/smasch/).

Species to be modeled include:



  • Plants: tidal marsh dominant species, special status tidal marsh species, and invasive species (Table 2)

  • Vertebrates: tidal marsh specialist avian species (Table 3)

In addition to species distributions, we will develop models of abundance for common avian species (see Table 3), and models for plant abundance (percent cover), productivity, and species diversity.

Environmental Inputs


The spatial resolution of our models will be tied to the resolution of available digital elevation models (DEMs), which will be used as a basis for future SLR and tidal inundation scenarios: 10-m x 10-m pixels from the national elevation dataset (NED). Other coarser data layers will be sampled down to this resolution. Our SLR scenarios will encompass a range of predictions based on several emissions scenarios from the upcoming Intergovernmental Panel on Climate Change (IPCC) Assessment 4 (AR4) simulations, incorporating thermal expansion as well as melting of glaciers and ice caps, and adjusted for California (Cayan et al. 2005) (Table 4).

Future marsh elevation predictions will be based on current topography and will not include geomorphic change, unless such predictions become available for San Francisco Bay. Future values for marsh relative elevation will be based on current elevation values, predicted SLR, and rates of marsh accretion (Figure 4). Because there is uncertainty as to how much future marsh accretion may occur, we will use two different estimates of marsh accretion: one which is indicative of current conditions (based on sampling by the BREACH team in north San Francisco Bay and Callaway in other Bay locations as well as published values in Patrick and DeLaune 1990 and other sources) and the potential for maximum marsh accretion based on a survey of other marsh systems (Table 4; see the following for a range of accretion estimates: Stevenson et al. 1986; Reed 1995; Callaway et al. 1996).

In addition to future elevation, we will model tidal inundation, using continuous water level data from NOAA, various municipalities, and restoration projects, including IRWM sites. We will develop tidal inundation graphs for each tide gauge location and calculate total monthly and maximum daily tidal inundation during the growing season (June/July), as well as tidal range. Inundation metrics will be interpolated across the subtidal and intertidal portions of the Bay-Delta, and adjusted for each SLR scenario to estimate future inundation metrics.

For estimates of future salinity, we will rely on predictions being generated by the CalFed-funded CASCaDE project (http://sfbay.wr.usgs.gov/cascade/), an extension of previous California climate modeling work conducted by the principal investigators (Knowles and Cayan 2002; Dettinger et al. 2004; Knowles et al. 2006). Using GCMs scaled-down for California, temperature and precipitation predictions were converted to monthly estimates of snowmelt runoff and stream flow, which were used to generate salinity predictions under various scenarios. These predictions will be available from the CASCaDE team and will be used in conjunction with tidal marsh salinity measurements collected by the IRWM project and the South Bay Salt Pond Restoration Project, to extend salinity predictions into the tidal marsh zone.

Finally, current land use from NOAA’s 2000 C-CAP dataset, will be included in models for vertebrate species, whose distributions and abundances are known to be limited by the composition and configuration of surrounding uplands (Shriver et al. 2004; Spautz et al. 2006). From this and other land use data, we will identify barriers to shoreward migration and use them as a mask for future distributions.

Species Modeling Approach


Depending on the type of data that are available for each species/metric, we will use variations on two different distribution modeling approaches:

  • Presence-only data: Maximum entropy (MaxEnt) models

  • Presence/absence data: Generalized linear models (GLM) or Generalized additive models (GAM) with a binary distribution

  • Abundance data: GLMs or GAMs with a Poisson or negative binomial distribution

  • Species diversity / productivity data: GAMs with a Gaussian distribution.

Data will be combined across multiple years to produce a single set of points for each metric of interest. Each data point will be weighted by sampling effort, the importance of which will be explicitly evaluated in the modeling process. Relationships between species metrics and environmental inputs will be used to develop models that predict current distributions (and abundance, etc.), as well as potential future distributions under various climate change scenarios (Table 4). Model predictions will consist of spatial data layers covering potentially tidal habitats and immediately adjacent uplands within the Bay-Delta system (see Figure 3). Each model will be built using 75% of the dataset, and evaluated using the other 25%, to obtain indicators of model predictive ability; this will be repeated three more times so that all data are included in one test dataset. Functional relationships will be evaluated and used to further evaluate the performance of each model. Examples of preliminary current and future predictions for a special status tidal marsh plant species, based on coarse environmental inputs, are shown in Figure 4.

Predicted distributions of dominant and invasive (but not special-status) tidal marsh plant species, as well as predicted primary productivity and species diversity, will be used as inputs to the vertebrate models (with non-tidal areas masked out for model development). Tidal marsh channel density will also be modeled based on environmental inputs and used as an input to vertebrate models.

Shifts in distribution, abundance, productivity, and species diversity, will be assessed and compared under each scenario, and the key environmental drivers will be identified For vertebrate species, we will evaluate the contribution of physical (salinity, elevation, inundation, channel density) factors compared with biotic factors (plant species composition and diversity) to better understand the mechanisms influencing their distribution and abundance. Co-occurring species will also be evaluated in terms of the similarity of future distributions, providing an indication of maintenance or disruption of future community integrity. Finally, for each species/metric, we will evaluate areas of highest potential loss and gain, across emissions scenarios, accretion rate assumptions and dispersal assumptions, providing insight for conservation and restoration priorities.

1   2   3   4   5   6   7   8


Verilənlər bazası müəlliflik hüququ ilə müdafiə olunur ©atelim.com 2016
rəhbərliyinə müraciət