Grantee Research Project Results
2011 Progress Report: Assessing Trade-Offs among Ecosystem Services in a Payment-for-Water Services Program on Florida Ranchlands
EPA Grant Number: R834567Title: Assessing Trade-Offs among Ecosystem Services in a Payment-for-Water Services Program on Florida Ranchlands
Investigators: Swain, Hilary M , Fauth, John E. , Bohlen, Patrick J , Jenkins, David , Kiker, Gregory , Quintana-Ascencio, Pedro , Shukla, Sanjay
Institution: Archbold Biological Station , University of Florida , University of Central Florida
Current Institution: Archbold Biological Station , University of Central Florida , University of Florida
EPA Project Officer: Packard, Benjamin H
Project Period: January 1, 2010 through December 31, 2012 (Extended to January 31, 2013)
Project Period Covered by this Report: January 1, 2011 through December 31,2011
Project Amount: $498,835
RFA: Enhancing Ecosystem Services From Agricultural Lands: Management, Quantification, And Developing Decision Support Tools (2009) RFA Text | Recipients Lists
Research Category: Ecological Indicators/Assessment/Restoration , Aquatic Ecosystems
Progress Summary:
In 2011, we continued to sample fifteen wetlands on four ranches with water management alternatives (WMAs) to investigate the relationships among hydrology, biodiversity, ecosystem services and ecosystem stressors. The four ranches were Alderman-Deloney Ranch, Buck Island Ranch, Williamson Ranch, and Peleaz Ranch. Detailed descriptions of these four ranches can be found in the Quality Assurance Project Plan, pages 65-73and page 84-85. In January 2011, we installed surveyed staff gauges in wetlands (ap1, ap2, ap3, wp2, wp3). We also installed several groundwater wells in wetlands (127, 246, and 17) on Buck Island Ranch. These instruments bolster our ability to characterize water availability and calibrate hydrological models. Approximately every two weeks wetland water levels were manually checked at staff gauges.
We sampled wetland species abundance and richness five times in 2011 at the beginning (April 9-11), middle (July 29-30, Aug 19-21), and end of the wet season (Oct 14-16, Nov 18-20). We are continuing to sample the 3-6 wetlands holding water in 2012 (Jan 3-4, Feb 17-18). This was more sampling then we expected but improves our ability to detect temporal trends in abundance of several taxa. The wetlands represent a range of hydrologic condition and hydroperiods, allowing us to test the relationships among ecosystem services, stressors, and hydrologic conditions.
We sampled plants, vertebrates, and invertebrates at the same time to determine interactions among biota and in response to hydrologic gradients. During sampling, we also recorded water depth and wetted perimeter of wetlands. Hydrologic data collected by the Florida Ranchlands Environmental Services Project (www.fresp.org) continued to be collected in 2010 and 2011.
Our project Quality Assurance Plan was reviewed and approved by EPA.
Plant Sampling Status
Plant data for 2010 was presented in the last report. Species-area curves generated in 2010 indicated that our sampling was close to reaching equilibrium in most wetlands.
In 2011, we sampled 459 plots across the 15 wetlands. We found 142 vascular plant species (of which 12 are classified as morpho-species) across all wetlands. The five most abundant species were 1. Panicum hemitomon, 2. Juncus effusus, 3. Polygonum punctatum, 4. Bacopa monnieri, and 5. Rhychospora inundata. Plant richness per wetland ranged from 14 to 66 species (Table 1). This was higher than 2010 due to a greater number of plots sampled. Add to table biomass weights. Biomass of plants was collected at the end of the growing season in 2010 and 2011 and these samples are currently being processed for forage quality (crude protein and total digestible nutrients).
We conducted preliminary analysis with regression to determine how diversity, non-native cover, and forage cover related to one measure of water availability – water depth. In the future, we plan on also exploring other measures of water availability. Across all points, we found that plant species richness decreased with increasing water depth. This was also true in 5 out of 15 wetlands. In three wetlands, the best fit curve was quadratic and showed a weakly concave relationship. At a water depth of approximately 30 cm, richness was lowest but slightly increased at deeper depths (Table 2). This suggests that our hypothesis that diversity of plants would peak at intermediate water availability, at least in terms of water depth, was not supported These preliminary analyses suggest that increased water depths associated with the FRESP project have the potential to negatively affect plant diversity. Of the wetlands that had significant relationships with water depth, all are diverse, wet prairies with short hydroperiods are as evidenced by the negative relationships. Only one wet prairie did not show a significant relationship versus eight non-diverse wetlands with no significant relationships.
We found fourteen non-native plants. Non-native cover and occurrence was heterogeneous making patterns difficult to find. In the four wetlands with significant relationships with water depth, there were contrasting patterns. In BIR 220, non-native cover decreased with water depth (R2=0.25) while in BIR 225, non-native cover increased with water depth (R2 = 0.55), and the other had a distinct concave relationship (BIR246 R2=0.30; Table 3). The most abundant non-native in these wetlands was West Indian Marsh grass (Hymenachne amplexicaulus) but two wetlands (BIR220 and BIR246) were grazed while BIR225 was not. In the ungrazed wetland, H. amplexicaulus dominated the center zone where it grows best. In the grazed wetlands, the ecotone was heavily grazed, leading to higher cover of non-natives on the edges. This pattern suggests that dispersal limitation and herbivory may be more important than water depth in determining non-native cover in these wetlands.
Patterns in forage availability were also idiosyncratic and depended on species. This includes both non-native and native forage species. Across all sampling points, wetland forage grasses had a concave relationship with water depth, being highest on the edges and in the center of wetlands. One wetland at Palaez ranch was surrounded by the planted non-native forage species, Hermarthria altissima. This plant can survive in wet, shallow conditions and we found that its cover decreased with increasing water depth. Several wetlands contained West Indian Marsh grass (Hymenachne amplexicaulus) and in these wetlands, forage cover had a concave relationship with water depth. Forage cover was higher on the edges of wetlands, decreased at intermediate depths and increased again in deeper conditions when H. amplexicaulus was present. This pattern could be due to grazing behavior; we observed cattle grazing mainly along wetland edges. Hymenachne amplexicaulus is an invasive wetland species with negative effects on invertebrate and plant diversity and should not be planted for forage. Cover of maidencane (Panicum hemitomon), a native grass beneficial as forage, had no relationship with water depth and was abundant over the entire gradient in most wetlands. Our hypothesis that upland forage cover declines with increasing water availability defined by water depth was supported. However, forage cover did not decrease at deeper depths when H. amplexicaulus was present.
Table 1. Descriptions of wetland plant assemblages across the 15 wetlands from four different ranches in 2010 and 2011.
Ranch |
Pond ID |
Total size (ha) |
Land-use type |
# sampling plots in 2010 |
# sampling plots in 2011 |
Native Plant Richness (2010) |
Native Plant Richness (2011) |
Exotic Plant Species |
Exotic Plant Species |
Mean plant species/m2 |
Alderman |
ALDp1 |
4.45 |
Semi-native |
18 |
72 |
23 |
63 |
0 |
5 |
6.3 |
Alderman |
ALDp2 |
1.28 |
Semi-native |
15 |
36 |
13 |
41 |
0 |
0 |
4.9 |
Alderman |
ALDp3 |
4.56 |
Semi-native |
36 |
69 |
21 |
40 |
0 |
2 |
5.25 |
Buck Island |
BIR105 |
0.78 |
Improved |
18 |
15 |
29 |
35 |
3 |
6 |
6.67 |
Buck Island |
BIR 127 |
0.39 |
Improved |
18 |
18 |
15 |
20 |
2 |
2 |
5.33 |
Buck Island |
BIR 17 |
0.66 |
Improved |
18 |
24 |
14 |
26 |
2 |
4 |
6.14 |
Buck Island |
BIR 218 |
0.74 |
Semi-native |
18 |
15 |
22 |
32 |
0 |
0 |
4.42 |
Buck Island |
BIR 220 |
0.24 |
Semi-native |
18 |
9 |
36 |
41 |
3 |
3 |
8.70 |
Buck Island |
BIR 225 |
0.63 |
Semi-native |
18 |
3 |
19 |
14 |
4 |
3 |
4.38 |
Buck Island |
BIR 246 |
0.37 |
Semi-native |
18 |
12 |
29 |
40 |
3 |
4 |
7.27 |
Buck Island |
BIR 310 |
0.57 |
Improved |
24 |
6 |
23 |
14 |
5 |
2 |
5.77 |
Peleaz |
PALp1 |
12.9 |
Improved |
30 |
66 |
21 |
43 |
4 |
5 |
6.05 |
Peleaz |
PALp4 |
7.5 |
Improved |
9 |
12 |
7 |
7 |
5 |
3 |
1.90 |
Williamson |
WILp2 |
31.06 |
Basin wetland |
39 |
57 |
43 |
47 |
3 |
4 |
5.91 |
Williamson |
WILp3 |
6.55 |
Basin wetland |
38 |
45 |
45 |
32 |
3 |
2 |
5.41 |
Table 2. Relationship between plant richness, non-native cover, and forage cover and water depth at different spatial scales. Relationships determined from linear and polynomial regression with α = 0.05 for regressors. Only statistically significant relationships are shown.
Grouping |
All sample points across 15 wetlands |
Within individual wetlands, only 8 of 15 had significant relationships |
Within four ranches |
Plant Richness |
Decreasing, (R2=0.07) |
ALDp2= decreasing (R2=0.09) |
WIL = concave (R2=0.32) |
|
|
ALDp3= decreasing (R2=0.19) |
BIR= decreasing (R2=0.13) |
|
|
BIR218 = decreasing (R2=0.21) |
ALD= decreasing (R2=0.08) |
|
|
BIR220 = decreasing (R2=0.61) |
|
|
|
BIR246= decreasing (R2=0.40) |
|
|
|
PALp1= concave (R2=0.14) |
|
|
|
WP2= concave (R2=0.28) |
|
|
|
WP3= concave (R2=0.36) |
|
Non-native Cover |
Concave, (R2=0.04) |
BIR225= concave (R2=0.55) |
BIR= concave (R2=0.07) |
|
|
BIR246= concave (R2=0.31) |
PAL=concave (R2=0.26) |
|
|
BIR220= concave (R2=0.25) |
WIL=concave (R2=0.05) |
|
|
PALp1= concave (R2=0.14) |
|
Forage Cover (wetland grasses) |
Concave, (R2=0.01) |
BIR127= decreasing (R2=0.12) |
ALD = Increasing, R2=0.02 |
|
|
BIR220= concave (R2=0.48) |
WIL= convex R2=0.04 |
|
|
BIR225= concave (R2=0.33) |
PAL= concave R2=0.07 |
|
|
BIR246= concave (R2=0.30) |
|
|
|
WILp2=convex (R2=0.08) |
|
Individual forage responses (non-natives) |
|
|
|
Hermarthria altissima |
Decreasing (R2=0.31) |
|
|
Hymenachne amplexicaulus (invasive) |
Concave(R2=0.64) |
|
BIR Increasing (R2=0.63) |
Vertebrate Sampling Status
We used a dropbox sampler (see Attachment A in Quality Assurance Plan page 40) to estimate the species composition, abundance, and size of insects, crayfish, fishes and amphibians. Dropbox samples were taken adjacent to every plant sample and we repeatedly dipnetted the dropbox until two consective sweeps were empty or we completed a total of 20 sweeps. This procedure recovered >99% of all individuals within the dropbox.
From 2010 through the January 2012, sampling session, we captured 1,338 insects larger than ~0.8 cm, including odonates, hemipterans, and beetles; 215 crayfish (Procambarus sp.); 1,403 fishes belonging to 8 species; and 357 amphibians belonging to 10 species. All individuals were native species except 5 brown hoplo (Hoplosternon littorale), which are an invasive South American fish. Therefore, our initial prediction that increasing water availability would increase the abundance and species richness of invasive vertebrates could not be tested since non-natives were too rare. The wetlands we sampled were isolated enough from sources of invasive species (e.g., Cuban treefrogwalking catfish) that colonization was limited, even during the massive flooding event at Alderman Ranch in 2011.
Abundances of all four taxa were significantly positively correlated at the level of individual point samples (0.11 < Spearman’s rho < 0.32, all p < 0.006). Correlations remained positive or did not differ significantly from zero when evaluated at the level of wetlands or ranches. Positive correlations in abundance among taxa known to compete or prey on one another suggests that an external driver like water availability or resource quality is more important to overall patterns than interspecific interactions.
Total abundance of all species captured with the dropbox sampler peaked at water depths of 10-35 cm. Where there were significant relationships this pattern generally remained whether the relationship was evaluated at the level of individual sample points, among wetlands, or among ranches (Table 3). Wetlands at Williamson Ranch were shallower than those at Alderman Ranch and were on the ascending portion of the abundance vs. depth curve. Where there were significant relationships taxa exhibited the same relationship; abundance peaked at water levels near 20 cm or increased with increasing water depth. Abundance decreased monotonically with increasing water depth in just three cases, all at the level of individual wetlands on different ranches (Table 3): insects at one Buck Island wetland, fishes at one wetland on Alderman Ranch, and amphibians in the deepest wetland in our study, which was on Pelaez Ranch. These preliminary analyses suggest that water depths observed with the FRESP project generally had little or a positive effect on abundance of macro-invertebrates, fishes, and amphibians (Table 3).
Table 3. Relationship between abundance of wetland macroinvertebrates, fishes, and amphibians and water depth at different spatial scales. Relationships determined from linear and polynomial regression with α = 0.05 for regressors. Only statistically significant relationships are shown.
Taxon |
All sample points |
Within individual wetlands |
Within ranches |
All |
convex |
ALD3 = convex |
ALD = convex |
|
|
BIR218 = increasing |
WIL = increasing |
|
|
WIL2 = increasing |
|
|
|
WIL3 = increasing |
|
Insects |
convex |
ALD2 = convex |
ALD = convex |
|
|
BIR218 = increasing |
BIR = convex |
|
|
BIR225 = decreasing |
WIL = convex |
|
|
WIL2 = convex |
|
Crayfish |
convex |
ALD3 = convex |
ALD = convex |
|
|
BIR17 = convex |
BIR = convex |
Fishes |
convex |
ALD3 = decreasing |
WIL = increasing |
|
|
WIL2 = increasing |
WIL = increasing |
Amphibians |
none |
BIR218 = increasing |
WIL = increasing |
|
|
PAL1 = decreasing |
|
|
|
WIL2 = convex |
|
Invertebrate Sampling Status
In the last year, we completed sorting invertebrate samples for 2010 and entered these into the project database. We have not conducted any analyses on these data in relation to water depth yet. In 2010, we found 12 mosquito species and abundance ranged from 2 individuals to 13,523 individuals found among ponds (Table 4). The most abundant species was Culex declarator (12,314 individuals) followed by Culex nigripalpus (1,260 individuals).
In 2011, we sampled the same plots as the plant and vertebrate teams (see complete sampling methods in Attachment A in Quality Assurance Plan page 39-40). We swept a 1 m x 1 m area using a smaller mesh net to target small invertebrates. We preserved all samples with ethanol and are currently sorting the samples collected in 2011 in the lab.
Table 4. Mosquito Richness and Abundance among sampled ponds for 2010.
Pond ID |
# of Plots Sampled |
Mosquito Richness |
Mosquito Abundance |
ap1 |
3 |
3 |
7 |
ap2 |
4 |
4 |
16 |
ap3 |
10 |
4 |
45 |
bp105 |
8 |
6 |
84 |
bp127 |
6 |
1 |
2 |
bp17 |
16 |
5 |
77 |
bp218 |
6 |
3 |
11 |
bp220 |
10 |
4 |
18 |
bp225 |
7 |
5 |
10 |
bp246 |
6 |
2 |
2 |
bp310 |
7 |
2 |
15 |
p1 |
23 |
7 |
125 |
p4 |
1 |
dry |
dry |
wp2 |
12 |
6 |
409 |
wp3 |
10 |
6 |
13,523 |
Hydrological Data
The calibration and evaluation of MIKE-SHE/MIKE11 (hydrological model) for the Buck Island Ranch has been completed for 120 m grid size. Although this grid size is larger than desired, shorter grid size resulted in higher run time. Efforts were made to set up the model for 10, 20, and 30 m grid. However, the model run times became unusually longer ranging from more than a week for 10 m grid to 17 hours for 30 m grid. The calibration and evaluation target for the model was the groundwater depth measured within the WMA areas and surface flows from the outlet of watershed 35. Work is underway to compare the predicted groundwater and surface water depth with the manual staff gage data for a wetland that was selected for ecological measurements. Statistical techniques for evaluating model performance such as Nash-Sutcliffe coefficient and index of agreement coefficients were used to evaluate the model predictions. Model validation was conducted using another part of the hydrologic data from the ranches. Statistical results from model evaluation showed that calibration and validation were satisfactory (N-S coefficients > 0.5). Several WMA scenarios including the baseline condition (no boards for water retention), FRESP WMA (higher boards for 2008-2011), NEPES WMA (higher than FRESP boards), and other board heights have been formulated and will be evaluated in 2012. Spatially distributed water levels and flow from the outflow structures for different scenarios will be combined with the ecological relationships to predict the hydrological and ecological responses. We will also conduct preliminary analysis integrating ecological and hydrological data.
The MIKE-SHE/MIKE11 modeling for Buck Island Ranch site has been challenging due to multiple sub-watersheds, complexity in identifying the watershed boundary for such a flat landscape with complex network of ditches throughout the ranch, accurate representation of surface and groundwater fluxes across the watershed boundary, and complex hydrology of the flatwoods region with highly interactive surface water and groundwater systems. We anticipate the analyses of scenarios to be completed by August 2012 for the BIR ranch. This is longer than we anticipated and was mainly due to the complexity in setting up the model, presence of numerous calibration parameters, and longer run times. The MIKE-SHE/MIKE11 model setup for the Pelaez ranch has begun and it is anticipated that the exploratory model runs will be started by May 2012. Although the original plan was to conduct modeling for two ranches, resources permitting, we may be able to conduct modeling for the third site, Williamson Ranch. We have started to explore the development of hydroecological relationships to relate the model predictions to changes in ecological variables such as fish and frog populations and vegetation.
For a detailed description of hydrological data collected at the four ranches, please see (Attachment A in Quality Assurance Plan page 51-62 and 63-85). We installed staff gauges at all wetlands in the project and these staff gauges will be visited biweekly. The water level at the wetlands on Peleaz Ranch is being monitored with data loggers which record a level every 15 minutes. There are groundwater monitoring wells at one wetland on Williamson Ranch and one at Alderman-Deloney Ranch. There is also a groundwater well in or near each of eight wetlands at Buck Island Ranch.
Establishing Hydro-ecological Relationships
We are conducting analyses at three different spatial scales; the sampling point, wetland, and ranch. The ranch is the largest spatial scale to conduct analyses, but hydrological data is more complex. We have conducted preliminary analyses for plants, invertebrates, and vertebrates using water depth as a proxy for water availability. Future analyses will incorporate other hydrological variables (Table 5).
Table 5. Hydrological variables we plan to use to define water availability at three different spatial scales. Different variables may be important for different organisms. A model selection (AIC) regression analysis will be used to select models with most explanatory power. Boxes highlighted in gray are hypothesized to be the most important parameters.
Ranch |
Wetland |
Point |
|
Volume |
x |
x |
|
Inundated Area |
x |
x |
|
Surface area: Volume |
x |
x |
|
Stage:Volume |
|
x |
|
Depth |
|
|
x |
Relative depth |
|
x |
x |
Hydraulic gradient |
x |
x |
|
Fractal dimension of the wetland |
x |
x |
|
External connectivity |
|
x |
|
Δ depth (index of flashiness) |
x |
5 wetlands |
|
Time since filling |
x |
x |
x |
Days wet/yr |
x |
x |
x |
Date |
x |
x |
x |
Preceding length dry |
x |
x |
x |
Rainfall |
x |
|
|
Drought |
x |
|
|
Temperature |
x |
|
|
Management type |
|
x |
|
Table 6. Summary of Project Status 2011.
Expenditures to date
Institution |
Expenditures to date |
University of Central Florida |
$132,051 |
Archbold Biological Station |
$51,935 |
University of Florida |
$32,114 |
Quality Assurance
We have taken the steps to ensure quality data generation for both hydrological and ecological data. We have developed an approved Quality Assurance Plan and Standard Operating Procedures for all aspects of the project and provided the EPA with a copy of this plan.
Future Activities:
A major goal for 2012 is to conduct the analyses integrating hydrologic and ecological data. In our preliminary analyses thus far, we have only used water depth as a proxy for water availability. We have defined several other hydrological parameters that may be important to relating ecosystem services and stressors to water availability (Table 2). We will employ Akaike Information Criterion (AIC) model selection to determine the most important hydrological variables to different organisms. We expect that different variables will be important for different organisms. Relationships between ecological and hydrological data will be used to inform a decision support system for trade-off evaluation of multiple ecosystem services. During this year we will work together to formulate the decision support tool beginning in summer 2012. This will include determining the target audience for the tool (government agencies/ private landowners) and evaluating several criteria, alternatives, and scenarios to analyze.
Our group is organizing two sessions on trade-offs in wetland ecosystem services in working landscapes at INTECOL International Wetlands Conference on June 3-8, 2012 in Orlando, FL. One session will cover conceptual frameworks and the other session is on biodiversity and hydrologic services. Co-PIs from our project will present in each of the sessions including a suite of other researchers from the USDA, USGS, Canada, England, and The Netherlands. We also submitted an abstract to Ecological Society of America and the 5th National Grazing Lands Conference this year. We are waiting to hear from EPA when they might hold the PI meeting.
Journal Articles on this Report : 2 Displayed | Download in RIS Format
Other project views: | All 2 publications | 2 publications in selected types | All 2 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Boughton E, Quintana-Ascencio P, Jenkins D, Bohlen P, Fauth J, Engel A, Shulka S, Kiker G, Hendricks G, Swain H. Trade-offs and synergies in a payment-for-ecosystem services program on ranchlands in the Everglades headwaters. ECOSPHERE 2019;10(5). |
R834567 (2011) |
Exit Exit |
|
Guo Y, Boughton E, Bohlman S, Bernacchi C, Bohlen P, Boughton R, Delucia E, Fauth J, Gomez-Casanovas N, Jenkins D, Lollis G, Miller R, Quintana-Ascencio P, Sonnier G, Sparks J, Swain H, Qiu J. Grassland intensification effects cascade to alter multifunctionality of wetlands within metaecosystems. NATURE COMMUNICATIONS 2023;14(1):8267 |
R834567 (2011) |
Exit |
Progress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.