Abstract
The Partners for Fish and Wildlife Program and Wetlands Reserve Program are U.S. federal programs that provide financial and technical assistance to restore wetland habitats on private property, and are important tools for the conservation and management of waterfowl. This study examined whether these wetland restorations successfully restored one important component of waterfowl habitat, the availability of vegetative forage, at sites in the St. Lawrence River valley of New York. We conducted surveys at 47 restored and 18 reference wetlands to characterize the vegetation assemblage in terms of its value as forage for waterfowl. Results suggest that these public–private partnership wetland restorations develop assemblages of wetland vegetation that are similar to reference wetlands. Vegetation assemblage metrics, including estimates of species richness, the richness of species of food value, the Vegetative Forage Quality Index, and the cover of species of food value, did not differ between restored–reference wetland pairs. However, invasive species were common at sites, and we detected a negative association between the cover of invasive species and the Vegetative Forage Quality Index at both restored and reference wetlands. On the basis of these results, we conclude that Partners for Fish and Wildlife Program and Wetland Reserve Program wetland restorations provide quality forage for breeding and migratory waterfowl in this region, but that the presence of invasive vegetation at sites has the potential to decrease the quality of vegetative forage at sites over time.
Introduction
Public–private partnership wetland restorations are an important tool for the conservation and management of waterfowl in North America (NAWMP 2012). These include restoration initiatives sponsored by diverse interests, including nonprofit organizations such as Ducks Unlimited, local land trusts, or state and local government entities. The U.S. Fish and Wildlife Service's (USFWS) Partners for Fish and Wildlife Program (PFWP) and Farm Bill programs administered by the U.S. Department of Agriculture (USDA) are two U.S. federal programs that sponsor public–private partnership habitat restoration initiatives. Private property owners provide the land, and federal agencies provide financial and technical assistance with the restoration process. Once the restoration is complete, the agency may recommend a management plan or place the area restored under a conservation easement. Under this model, the USFWS, USDA, and Ducks Unlimited have restored an estimated 7 million ha of wetlands across the United States and Canada (Filsinger and Milmoe 2012; USDA NRCS 2014; Ducks Unlimited 2018).
Postrestoration monitoring of public–private partnership wetland restorations indicate that programs are accomplishing their goals in terms of providing habitat for waterfowl and other wetland-associated wildlife, and that restorations are generally comparable with reference wetlands (e.g., Brown and Smith 1998; Rewa 2005, 2007; Gray and Teels 2006; Tapp et al. 2017; Benson et al. 2018). However, one aspect of restoration success that has received limited attention is the quality of restored habitat, such as the availability of cover or food resources for wildlife (King et al. 2006; Evans-Peters et al. 2012; Fleming et al. 2012; Olmstead et al. 2013; Tapp et al. 2017). Wetland vegetation diversity, composition, and structure are among the most important components of waterfowl habitat (Baldassarre and Bolen 2006; Ma et al. 2010). Waterfowl rely on vegetation directly for food by consuming leaves, fruits, seeds, or tubers, and indirectly as habitat for prey species. Vegetation provides refuge from predators, thermal cover, nest sites and nesting material, and habitat for foraging, roosting, and loafing (Martin et al. 1951; Fredrickson and Taylor 1982; Strader and Simpson 2005; Strickland et al. 2009).
Monitoring the availability and quality of food resources at wetland restorations is key to assessing their value as habitat for waterfowl. Food resources are considered to be a limiting factor for waterfowl during all aspects of their life cycle and affect body condition, survivorship, and recruitment (e.g., Delnicki and Reinecke 1986; Kaminski and Gluesing 1987; Heitmeyer 1988; Guillemain et al. 2008; Anteau and Afton 2009). Furthermore, estimates of food availability are important components of the bioenergetics models used to estimate waterfowl carrying capacity (duck-use days) during migratory and wintering periods under the North American Waterfowl Management Plan (Goss-Custard et al. 2002; Petrie et al. 2011; Williams et al. 2014). Several studies have examined the dynamics of waterfowl food resources at wetlands, but most have focused on wetlands on public land that use moist-soil management practices to enhance food production for waterfowl (e.g., Bowyer et al. 2005; Greer et al. 2007; Kross et al. 2008; Hagy and Kaminski 2012a; Hagy et al. 2014). However, wetlands restored via public–private partnership programs are likely not analogous to managed wetlands on public land, as restorations are generally smaller in size and management actions may be less consistent (King et al. 2006; Brasher et al. 2007; Evans-Peters et al. 2012; Tapp et al. 2017; Benson et al. 2018). Fewer studies have characterized the availability of waterfowl food resources at public–private partnership wetland restorations (Brown 1999; Evans-Peters et al. 2012; Fleming et al. 2012, 2015; Olmstead et al. 2013; Tapp and Web 2015; Tapp et al. 2017).
In this study, we characterized the wetland vegetation assemblage in terms of its value as waterfowl forage at public–private partnership wetland restorations in the St. Lawrence River valley (SLRV) of New York. Either the USFWS PFWP or the USDA Wetlands Reserve Program (WRP) completed restorations included in this study. The PFWP was established in 1987 with the objective of restoring habitat on private land for the benefit of federal trust species, including migratory birds (USFWS 2018). The 1990 Farm Bill first established the WRP to provide assistance with wetland restoration on private land with the goal of restoring wetland ecosystem services to agricultural landscapes, including quality wildlife habitat (USDA NRCS 2014). In 2014, the WRP and other USDA easement programs were incorporated into the Agricultural Conservation Easement Program. A primary goal of both PFWP and WRP programs is to restore waterfowl habitat; therefore monitoring the establishment of food resources at sites is an important component of evaluating restoration success (Brown 1999; King et al. 2006; Evans-Peters et al. 2012; Olmstead et al. 2013).
The USFWS has designated the SLRV as a focal region for wetland restoration in the northeastern United States, as it provides important breeding and migratory habitat for several waterfowl (USFWS 2006). A total of 18 species of waterfowl uses the SLRV for breeding, migratory, or wintering habitat, including the greatest density of breeding mallards Anas platyrhynchos (40,000 pairs) in the Atlantic Flyway (NAWMP ACJV 2005). In addition to waterfowl, the SLRV is home to several breeding waterbirds, including several wetland-associated species of greatest conservation need (NYSDEC 2015). This includes the American bittern Botaurus lentiginosus, black-crowned night heron Nycticorax nycticorax, black tern Chlidonias niger, least bittern Ixobrychus exilis, pied-billed grebe Podilymbus podiceps, and sedge wren Cistothorus platensis (Brown and Smith 1998; NAWMP ACJV 2005; McGowan and Corwin 2008; Benson et al. 2018). Thirteen of the 19 peer-reviewed studies that have evaluated vegetative forage for waterfowl have been conducted in the Mississippi Alluvial Valley and Southeast regions (e.g., Bolduc and Afton 2004; Greer et al. 2007; Kross et al. 2008; Foster et al. 2010; Olmstead et al. 2013, Hagy et al. 2014, Fleming et al. 2015; Tapp et al. 2017). Therefore, this study provided an opportunity to assess waterfowl food resources in a region that provides important habitat for waterfowl in the Atlantic Flyway, but is underrepresented in the literature.
Our study had two primary objectives. The first objective was to characterize the vegetation at 47 PFWP and WRP wetland restorations in terms of their value as forage for waterfowl, and to compare these metrics between 18 restored and reference wetland pairs. We hypothesized that restorations would support vegetation assemblages rich in species of waterfowl food value and that assemblage metrics would not differ between restored and reference wetlands. These hypotheses were based on Brown (1999), which indicated that in as little as 3 y, PFWP wetland restorations in the in the SLRV developed vegetation assemblages with greater richness and cover of species of waterfowl food value than found at reference wetlands. Moreover, Benson et al. (2018) detected no difference in waterbird species richness at restored–reference wetland pairs of the same wetlands as were the subject of the present study. Our second objective was to examine the relationship between the cover of invasive species and vegetative forage quality at sites. Invasive species are common at sites, and we hypothesized that they would negatively affect vegetative forage. It is well documented that invasive wetland vegetation has the ability to create dense, monotypic stands and can rapidly outcompete native species, many of which provide quality forage for waterfowl (Zedler and Kercher 2004; Vilá et al. 2011).
Study Site
The U.S. portion of the SLRV encompasses a roughly 600,000-ha expanse of the New York State counties of Jefferson (44°3′0.33″N, 75°57′18.95″W) and St. Lawrence (44°28′10.59″N, 75°9′37.80″W), and is distributed across three ecological regions: the Eastern Ontario Plains, St. Lawrence Plains, and Indian River Lakes region (Reschke 1990). Both the Eastern Ontario and St. Lawrence Plains are characterized as agricultural lowlands of hayfields, dairy farms, and row crops, interspersed with northern mixed hardwood and conifer forests; agricultural land uses represent ∼35% of the total land cover in these regions. The Indian River Lakes region contains numerous glacial lakes, granite outcrops, greater forest cover, and ∼15% of land is dedicated to agricultural uses. Elevation varies from 80 m adjacent to Lake Ontario and the St. Lawrence River to 140 m in the Indian River Lakes region. The climate is cool (annual average temperature is 6.6°C, average minimum winter temperature is −17.5°C, and average maximum summer temperature is 24.9°C at Canton, New York) and seasonal, including long cold winters and short cool summers (growing season is 125 d). Precipitation is high, 94 cm annually, and occurs throughout the year (NOAA NCDC 2010).
A total of 47 PFWP and WRP wetland restorations and 18 reference wetlands was surveyed over the 4-y study period (2009–2011, 2014; Figure 1). This included 15 PFWP sites, 28 WRP sites, and four sites that were shared between programs. Additionally, Ducks Unlimited provided assistance at 18 sites and 31 sites were protected under long-term conservation easements. All wetlands were characterized as either scrub-shrub or emergent palustrine wetlands (Cowardin et al. 1979). Restored wetlands ranged from 0.14 to 7.49 ha (mean ± SD = 5.32 ± 2.26 ha) in size and were restored using a variety of methods, including the excavation of small or large depressions, installation of levees/berms and water control structures, ditch plugging, and tile drain removal. The degree to which restored wetlands were actively managed by landowners varied considerably. A study conducted by Welsh et al. (2018) at a subset (n = 35) of the restorations surveyed in the this study indicated that in terms of active management at sites, six landowners conducted water drawdowns, 10 planted food plots for wildlife, and 23 maintained periodic disturbance of upland vegetation by mowing. We paired restored wetlands with reference sites on the basis of wetland area, proximity, basin morphometry, and landscape context (Brinson and Reinhardt 1996; Figure 2). Reference wetlands ranged from 0.17 to 10.73 ha (7.86 ± 2.69 ha) in size, 16 of 18 sites were located on private property, and all restored–reference pairs were surveyed in the same year. There was evidence of North American beaver Castor canadensis activity at over half of restored wetlands, including 8 of 18 restored–reference wetland pairs. A more detailed description of the study region and wetland site attributes can be found in Benson et al. (2018).
Locations of restored (n = 47) and reference wetlands (n = 18) surveyed in 2009–2011 and 2014. The inset map indicates the location of sites within New York State. The larger map indicates site locations within the St. Lawrence River valley, where restorations are indicated by yellow circles and reference wetlands are indicated by red squares. Aerial imagery is adopted from the Microsoft® BingTM Maps Platform; product screen shots reprinted with permission from Microsoft Corporation.
Locations of restored (n = 47) and reference wetlands (n = 18) surveyed in 2009–2011 and 2014. The inset map indicates the location of sites within New York State. The larger map indicates site locations within the St. Lawrence River valley, where restorations are indicated by yellow circles and reference wetlands are indicated by red squares. Aerial imagery is adopted from the Microsoft® BingTM Maps Platform; product screen shots reprinted with permission from Microsoft Corporation.
Aerial captured in 2011 depicting an example of a restored (A) and reference (B) wetland pair located in the St. Lawrence River valley of New York State. We paired all restorations with reference wetlands on the basis of wetland area, proximity, basin morphometry, and landscape context, and we surveyed all restored-reference pairsduring the same year. Aerial imagery is adopted from the Microsoft® BingTM Maps Platform; product screen shots reprinted with permission from Microsoft Corporation.
Aerial captured in 2011 depicting an example of a restored (A) and reference (B) wetland pair located in the St. Lawrence River valley of New York State. We paired all restorations with reference wetlands on the basis of wetland area, proximity, basin morphometry, and landscape context, and we surveyed all restored-reference pairsduring the same year. Aerial imagery is adopted from the Microsoft® BingTM Maps Platform; product screen shots reprinted with permission from Microsoft Corporation.
Methods
Vegetation surveys
We surveyed sites one time during the study period, in either August or early September, using a standard, validated vegetation survey protocol developed by the Great Lakes Coastal Wetlands Monitoring Program (Albert 2008). This protocol was developed for the region and provides area-based recommendations for the number of quadrats required to adequately characterize a wetland's vegetation. Vegetation was surveyed along three randomly located transects that ran perpendicular to the wetland edge. Each transect consisted of three 1-m2 quadrats positioned on an elevation gradient to capture the transition from submergent, emergent, to upland vegetation zones. The first quadrat encompassed the emergent vegetation zone and was located at the water's edge, delineated by the maximum observed water level during early spring. The second quadrat was located at an elevation of 20 cm below the first in the submergent vegetation zone. The third quadrat was located at an elevation of 20 cm above the first in the upland vegetation zone. A total of 9 m2 of vegetation was surveyed at each site, including three 1-m2 quadrats per vegetation zone (Brown 1999; Albert 2008). We identified all vascular plant species within quadrats and visually estimated the percent cover per species. A voucher specimen of each species was collected per site, pressed, and verified by local botanical expert Anne M. Johnson (Eldblom and Johnson 2010). Plant common and scientific names are consistent with nomenclature in the USDA Plants Database (USDA NRCS 2018).
Data analysis
From data collected in these quadrats, we estimated five vegetation assemblage metrics: total species richness, the richness of species of food value, the Vegetative Forage Quality Index (VFQI; Fleming et al. 2012), proportional cover of species of food value, and the proportional cover of invasive species. To calculate these metrics at the site level, we combined data across all nine quadrats to estimate the total species richness, the richness of species of food value, and VFQI. We averaged data across quadrats to determine the average proportional cover of species of food value and cover of invasive species at sites. To calculate these metrics within vegetation zones, we averaged data across the three quadrats from submergent, emergent, and upland zones.
The VFQI was developed by Fleming et al. (2012) to characterize the quality of the vegetation assemblage as forage for waterfowl and is a modification of the Floristic Quality Index and mean coefficient of conservatism (Taft et al. 2006). We adopted the VFQI with minor adjustments. Instead of using a panel of regional experts to assign forage quality coefficients, we used coefficients reported by Fleming et al. (2012) when possible or used coefficient values (poor = 1, average = 2, excellent = 3) reported by previous studies (e.g., Martin et al. 1951; Payne 1992; Brown 1999; Strader and Simpson 2005; Nelms 2007). Additionally, we used the proportional cover of each species instead of detection frequency. Invasive species were those plants identified as invasive or nuisance species by the New York State Department of Environmental Conservation (NYSDEC 2018).
The average mean difference in vegetation assemblage metrics between restored–reference wetland pairs (n = 18) was analyzed via paired t-tests (or Wilcoxon signed-rank tests [W] when data violated the assumptions of normality or homoscedasticity) for all metrics at the site level and within vegetation zones. When observed mean differences did not differ significantly from zero, some would advocate that a retrospective power analysis be reported (either an “observed power” or “detectable effect size” analysis) to evaluate the likelihood that the null hypothesis (average mean difference = 0.0) would have been rejected were it false. However, we agree with Hoenig and Heisey (2001) that the use of power analysis in this way is incorrect, and that CIs are more informative. An average mean difference near zero and tight CIs provide the most convincing evidence that there is little to no difference between treatment types. Thus, we report the 95% CI for all pair-wise comparisons. To examine the influence of invasive species on forage quality, we used linear regression (R2 or Spearman's rank order correlation coefficient [rs] when data violated the assumptions of normality or homoscedasticity) to examine the relationship between the average proportional cover of invasive species and VFQI at the site level and within vegetation zones. All statistical analyses were completed using the {stats} package in R project for statistical computing (R Core Team 2018).
Results
We detected a total of 94 species representing 34 genera of waterfowl food value at sites, including 80 species at restorations (n = 47 sites) and 71 species at reference sites (n = 18; Table 1). The total species richness of vegetation at sites averaged 28.5 ± SD 7.75 at restorations and 31.06 ± 7.41 at reference wetlands, whereas the average richnesses of species of food value were 13.78 ± 4.15 and 16.74 ± 4.24 at restored and reference wetlands, respectively. Results for all vegetation assemblage metrics at the site level and within vegetation zones can be found in Table 2. Statistical analyses indicate no difference in any of the metrics tested between restored–reference wetland pairs (n = 18 pairs). The most commonly detected species of food value at both restored and reference wetlands were rice cutgrass Leersia oryzoides, sedges of the genus Carex, common duckweed Lemna minor, common rush Juncus effusus, and broadfruit bur-reed Sparganium eurycarpum. A list of all plant species detected at restored and reference wetlands is available as Supplemental Material (Table S1).
Vegetation species of waterfowl food value detected at restored (n = 47) and reference (n = 18) wetlands in the St. Lawrence River valley of New York State during 2009–2011 and 2014. The relative forage quality of each species (poor, average, excellent) is reported along with the proportional occurrence at restored and reference sites.

Summary of vegetation metrics detected at restored (n = 47) and reference (n = 18) wetlands in the St. Lawrence River valley of New York State during 2009–2011 and 2014. Metrics were calculated at the site level and within submergent, emergent, and upland vegetation zones. Metrics included the total species richness (total spp. richness), the Vegetative Forage Quality Index (VFQI), the total richness of species of waterfowl food value (food value spp. richness), the proportional cover of species of food value (cover of food value spp.), and the proportional cover of invasive species (cover of invasive spp.). The range, mean, and SD are reported for each metric at restored and reference wetlands. Results from t-tests, including the average mean difference (AMD) between treatments, 95% CIs, and P-values (P) are reported between restored and reference wetland pairs (n = 18). Wilcoxon signed-rank tests (a) were used to analyze nonparametric data and the AMD in ranks for these analyses are reported.

Nine invasive species were detected across sites, including common frogbit Hydrocharis morsus-ranae, common reed Phragmites australis, Eurasian watermilfoil Myriophyllum spicatum, hybrid cattail Typha × glauca, Morrow's honeysuckle Lonicera morrowii, narrowleaf cattail Typha angustifolia, purple loosestrife Lythrum salicaria, reed canary grass Phalaris arundinacea, and Tatarian honeysuckle Lonicera tatarica. The average proportional cover of invasive species was 0.21 ± 0.04 at restorations and 0.17 ± 0.05 at reference wetlands (Table 2). The most commonly detected species at both restored and reference wetlands were reed canary grass (81% of restored and 72% of reference wetlands), common frogbit (51% of restored and 72% of reference wetlands), and purple loosestrife (21% of restored and 44% of reference wetlands). Only three restorations and one reference site appeared to be free of invasive vegetation. We did not include the richness of invasive species in our analyses because we generally detected only one to two invasive species per site. Statistical analyses indicated a significant negative correlation between the average proportional cover of invasive species and the VFQI at the site level and within submergent and emergent vegetation zones (site level: F1,64 = 4.84, P = 0.03, R2 = 0.07; submergent: rs = −0.57, P < 0.01; emergent: rs = -0.59, P < 0.01; Figure 3A–C). However, there was no relationship among these variables in the upland vegetation zone (rs = −0.04, P = 0.74; Figure 3D). All raw data are available as Supplemental Material (Data S1).
Results of linear regression (R2) or Spearman's rank-order correlation coefficient (rs) analyses examining the relationship between the average proportional cover of invasive species and average Vegetative Forage Quality Index (VFQI) at the site level (A) and within submergent (B), emergent (C), and upland (D) vegetation zones; P-values are indicated by P. Restorations are indicated by solid black circles and references are indicated by open white circles. Analyses at the site level and within vegetation combine estimates of invasive cover and VFQI at both restored (n = 47) and reference wetlands (n = 18) collected during 2009–2011 and 2014.
Results of linear regression (R2) or Spearman's rank-order correlation coefficient (rs) analyses examining the relationship between the average proportional cover of invasive species and average Vegetative Forage Quality Index (VFQI) at the site level (A) and within submergent (B), emergent (C), and upland (D) vegetation zones; P-values are indicated by P. Restorations are indicated by solid black circles and references are indicated by open white circles. Analyses at the site level and within vegetation combine estimates of invasive cover and VFQI at both restored (n = 47) and reference wetlands (n = 18) collected during 2009–2011 and 2014.
Discussion
Our results support the hypothesis that public–private partnership wetland restorations in the SLRV develop vegetation assemblages that include species with food value for waterfowl and that they are similar to reference wetlands in the region. None of the vegetation assemblage metrics we examined differed between restored and reference wetlands. Of the total species detected at sites, on average, half provided some level of food value for waterfowl. Additionally, species of food value also made up an average of half of the vegetative cover within quadrats at both restored and reference wetlands.
Our results are similar to those of other studies that have examined vegetative forage quality at public–private partnership wetland restorations. They also suggest that restorations provide quality forage for waterfowl and that they are generally providing resources that are similar to or exceed those found at reference wetlands (Brown 1999; Evans-Peters et al. 2012; Fleming et al. 2012, 2015; Olmstead et al. 2013; Tapp et al. 2017). A majority of these studies was conducted in the Mississippi Alluvial Valley, and the moist soil impoundments typical of this region are structurally distinct from the wetlands in the SLRV, which makes it difficult to use these studies to interpret our results (Rewa 2005, 2007). However, Brown (1999) surveyed 13 PFWP wetland restorations in the SLRV and found that 3 y postrestoration, sites had higher richness and cover of species of waterfowl food value than reference wetlands. They concluded that restorations provided greater food value because they had greater cover of early successional emergent herbaceous vegetation, whereas reference wetlands generally had greater cover of woody, scrub-shrub vegetation that is characteristic of late successional wetlands.
In this study, the age of restorations ranged from 2 to 25 y old. Similar to Brown (1999), we observed that younger restorations were dominated by early successional vegetation, including many species of grasses and forbs that provide quality forage for water, whereas older sites (10+ y old) generally had well-developed stands of emergent annual and perennial herbaceous vegetation and shrub-scrub vegetation, including willows Salix spp., common buttonbush Cephalanthus occidentalis, and dogwoods Cornus spp. Research suggests that vegetative forage may decrease as wetlands advance from early to late successional vegetative stages. However, many late successional species provide other important components of waterfowl habitat, including refuge from predators, nesting habitat, or thermal cover (Martin et al. 1951; Fredrickson and Taylor 1982; Strader and Simpson 2005; Strickland et al. 2009).
The VFQI was developed by Fleming et al. (2012) to describe the vegetation assemblage in terms of its forage quality for waterfowl on the basis of measures of species richness and abundance at wetlands. This differs from more traditional methods of estimating seed biomass in soil cores. Although the latter techniques provide a more precise estimate of available food, they are time and resource intensive. One of the major benefits of the VFQI is that it can be applied post hoc to the type of visual survey data (e.g., species richness or abundance) that are typically collected when surveying vegetation assemblages. To date, Fleming et al. (2012) is the only other study that has used VFQI to characterize waterfowl forage quality. Although they were also working on public–private partnership wetland restorations (WRP), their study was located in the Mississippi Alluvial Valley and examined how active and passive water drawdowns influenced the development of vegetative forage at sites. They report that the average VFQI ranged from 0.18 to 1.86 across treatments, and that VFQI was greater at sites with early-season water drawdowns than those where water drawdowns occurred during late season or were passively managed. In this study, the average VFQI was 1.37 ± 0.29 SD (range = 0.81–1.75) at restored wetlands in the SLRV. It is encouraging that the quality of vegetative forage at restorations in this region was similar to or exceeded that detected in the Mississippi Alluvial Valley, where moist-soil management is a more common practice.
As with Fleming et al. (2012), many studies of waterfowl vegetative forage have focused on the role of active management in promoting the growth of species of food value. Evidence suggests that when wetlands are actively managed, whether through periodic disturbance or hydrological manipulation, they generally contain a greater abundance of quality forage for waterfowl (e.g., Brasher et al. 2007; Greer et al. 2007; Kross et al. 2008; Hagy and Kaminski 2012b; Olmstead et al. 2013). The potential for hydrologic manipulation to benefit waterfowl populations in the SLRV is also supported by Kaminski et al. (2006). They conducted a study of WRP restorations in the Oneida Lake Plain of New York, a region adjacent to the SLRV, and found that the species richness and abundance of waterfowl was greater at sites that were hydrologically managed through water drawdowns. On the basis of data collected by Welsh et al. (2018), it does not appear that hydrologic management is widely implemented by landowners at restorations in the SLRV. Of the 35 landowners that participated in survey efforts, only six conduct water drawdowns at sites. It is worth noting that not all restorations in this region are restored with water control structures, but many are, and working with landowners to implement hydrologic management would likely further increase the quality of waterfowl forage at sites.
Our results also supported the hypothesis that invasive vegetation would negatively affect the quality of vegetative forage at restored and reference wetlands. We detected a negative association between the average proportional coverage of invasive species and the VFQI at both the site level and in submergent and emergent vegetation zones. However, there was no correlation between these metrics in the upland vegetation zone. This pattern was expected given that the majority of vegetative species of waterfowl food value are obligate and facultative species, and most are not found in the upland vegetation zone. Although we detected no difference in the cover of invasive species at restored–reference wetland pairs, these results are not entirely encouraging. The average proportional coverage of invasives was 0.21 ± 0.04 at restorations and 0.17 ± 0.05 at reference wetlands, and only 3 of 47 restorations and 2 of 18 reference sites appeared to be free of invasive species.
Given the ubiquity of invasive species at sites, we felt it was important to examine its impact on waterfowl food value, and to the best of our knowledge, this is the first study to test this hypothesis. That we found a negative association between the cover of invasive species and VFQI is not surprising, as the impact of invasive species on native ecosystems is well documented. Invasives have the capacity to grow rapidly in population size and outcompete native species. Wetlands in general, and restored wetlands specifically, are particularly susceptible to colonization and rapid establishment of invasive vegetation (e.g., Moore et al. 1999; Mulhouse and Galatowitsch 2003; Zedler and Kercher 2004; Balcombe et al. 2005; Bantilan-Smith et al. 2009). Although some invasive species do provide forage for waterfowl, and as such, were included in our calculations of VFQI, their forage quality is generally poor compared with the native species they replace (Dugger et al. 2007; Nelms 2007; Strickland et al. 2009).
Although the establishment of invasive vegetation seems inevitable at sites in the SLRV, this represents another opportunity to involve landowners in actively managing sites. Ideally, this would include information on identifying common invasives in the region and strategies to avoid their accidental introduction. This knowledge would allow landowners to identify invasives during the early stages of colonization, when populations are smaller and more easily eradicated.
On the basis of our results, we conclude the PFWP and WRP wetland restorations in the SLRV are meeting program-level goals in terms of one important component of waterfowl habitat, the availability of vegetative forage. Vegetation assemblage metrics, including overall species richness and several measures of food value, did not differ between restored and reference wetlands in this region. Given the large number of these and other public–private wetland restorations in the SLRV, wetland restoration programs' aggregate contribution to the availability of waterfowl forage is likely to be substantial. However, invasive species were common at sites, and our results suggest that the presence of invasive vegetation could decrease the quality of vegetative forage over time. Working with landowners to improve the implementation of active management strategies, such as conducting water drawdowns and monitoring the establishment of invasive species, are opportunities for restorations managers to improve the quality of waterfowl habitat in this region.
Supplemental Material
Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding author for the article.
Table S1. All vegetation species detected at restored (n = 47) and reference (n = 18) wetlands in the St. Lawrence River valley of New York during 2009–2011 and 2014. We detected a total of 248 plant species, with 212 species at restorations and 168 species at reference wetlands.
Found at DOI: https://doi.org/10.3996/092018-JFWM-080.S1 (26 KB DOCX).
Data S1. Data in support of study results in Microsoft Excel file format. There are three worksheets in this file. The first worksheet titled “Metadata” contains readme file information for worksheets two and three. The second worksheet is titled “Raw Data” and contains quadrat-level data for all sites, including the name of each species detected and its estimated proportional cover. The third worksheet is titled “Restored–Reference Pairs” and contains the site codes for the 18 restored and reference wetlands used for pair-wise comparison of vegetation assemblage metrics. We collected data at restored (n = 47) and reference (n = 18) wetlands in the St. Lawrence River valley of New York during 2009–2011 and 2014.
Found at DOI: https://doi.org/10.3996/092018-JFWM-080.S2 (254 KB XLSX).
Reference S1. Albert DA. 2008. Chapter 3—vegetation community indicators. Pages 32–54 in Burton TM, Brazner JC, Ciborowski JJH, Grabas GP, Hummer J, Schneider J, Uzarski DG, editors. Great Lakes Coastal Wetlands Monitoring Plan. Ann Arbor, Michigan: Great Lakes Coastal Wetlands Consortium.
Found at DOI: https://doi.org/10.3996/092018-JFWM-080.S3 (15.6 MB PDF); also available at https://www.glc.org/wp-content/uploads/2016/10/Great-Lakes-Coastal-Wetlands-Monitoring-Plan-FINAL-March-2008.pdf
Reference S2. Cowardin LM, Carter V, Golet FC, LaRoe ET. 1979. Classification of wetlands and deepwater habitats of the United States. Report No. FWS/OBS/-79/31. Washington, D.C.: U.S. Fish and Wildlife Service.
Found at DOI: https://doi.org/10.3996/092018-JFWM-080.S4 (17.45 MB PDF); also available at https://www.fws.gov/wetlands/Documents/Classification-of-Wetlands-and-Deepwater-Habitats-of-the-United-States.pdf
Reference S3. Ducks Unlimited. 2018. Ducks Unlimited 2018 Annual Report. Menphis, Tennessee: Ducks Unlimited.
Found at DOI: https://doi.org/10.3996/092018-JFWM-080.S5 (3.27 MB PDF); also available at https://duckscdn.blob.core.windows.net/imagescontainer/landing-pages/aboutDU/AR2018.pdf
Reference S4. Filsinger M, Milmoe J. 2012. Restore and enhance: 25th anniversary of the Service's Partners of Fish and Wildlife Program. Washington, D.C.: U.S. Fish and Wildlife Service.
Found at DOI: https://doi.org/10.3996/092018-JFWM-080.S6 (208 KB PDF); also available at https://www.fws.gov/partners/docs/PFW_25th_Celebration_article.pdf
Reference S5. Fredrickson LH, Taylor TS. 1982. Management of seasonally flooded impoundments for wildlife. Washington, D.C.: U.S. Fish and Wildlife Service. Resource Publication 148.
Found at DOI: https://doi.org/10.3996/092018-JFWM-080.S7 (2.54 MB PDF); also available at http://www.dtic.mil/docs/citations/ADA323232
Reference S6. Nelms KD, editor. 2007. Wetland management for waterfowl handbook. Washington, D.C.: U.S. Fish and Wildlife Service.
Found at DOI: https://doi.org/10.3996/092018-JFWM-080.S8 (3.08 MB PDF); also available at https://www.nrcs.usda.gov/Internet/FSE_DOCUMENTS/nrcs142p2_016986.pdf
Reference S7. [NYSDEC] New York State Department of Environmental Conservation. 2015. New York State Wildlife Action Plan.
Found at DOI: https://doi.org/10.3996/092018-JFWM-080.S9 (5.53 MB PDF); also available at https://www.dec.ny.gov/animals/7179.html
Reference S8. [NAWMP] North American Waterfowl Management Plan. 2012. People conserving waterfowl and wetlands.
Found at DOI: https://doi.org/10.3996/092018-JFWM-080.S10 (2.67 MB PDF); also available at https://www.fws.gov/migratorybirds/pdf/management/NAWMP/2012NAWMP.pdf
Reference S9. [NAWMP ACJV] North American Waterfowl Management Plan, Atlantic Coast Joint Venture. 2005. Atlantic Coast Joint Venture Waterfowl Implementation Plan, Revision.
Found at DOI: https://doi.org/10.3996/092018-JFWM-080.S11 (7.51 MB PDF); also available at http://acjv.org/planning/waterfowl-implementation-plan/
Reference S10. Petrie MJ, Brasher MG, Soulliere GJ, Tirpak JM, Pool DB, Reker RR. 2011. Guidelines for establishing joint venture waterfowl population objectives. North American Waterfowl Management Plan.
Found at DOI: https://doi.org/10.3996/092018-JFWM-080.S12 (169 KB PDF); also available at https://www.fws.gov/migratorybirds/pdf/management/NAWMP/GuidelinesforEstablishing%20JVPopulationObjectives%20FinalReport_draft_alt_format.pdf
Reference S11. Reschke C. 1990. Ecological communities of New York State. Albany, New York: New York State Department of Environmental Conservation.
Found at DOI: https://doi.org/10.3996/092018-JFWM-080.S13 (4.71 MB PDF); also available at https://www.dec.ny.gov/docs/wildlife_pdf/ecocomm1990.pdf
Reference S12. Rewa CA. 2005. Wildlife benefits of the Wetlands Reserve Program. Pages 133–146 in Haufler JB, editor. Fish and wildlife benefits of Farm Bill programs: 2000–2005 update. Bethesda, Maryland: The Wildlife Society.
Found at DOI: https://doi.org/10.3996/092018-JFWM-080.S14 (2.99 MB PDF); also available at https://digitalcommons.unl.edu/cgi/viewcontent.cgi?article=1105&context=usdafsfacpub
Reference S13. Rewa CA. 2007. Fish and Wildlife benefits associated with wetland establishment practices. Pages 71–82 in Haufler JB, editor. Fish and wildlife response to Farm Bill conservation practices. Bethesda, Maryland: The Wildlife Society.
Found at DOI: https://doi.org/10.3996/092018-JFWM-080.S15 (2.33 MB PDF); also available at https://www.fws.gov/partners/docs/farmbill/fwfball.pdf
Reference S14. Strader TW, Simpson PH. 2005. Moist-soil management guidelines for U.S. Fish and Wildlife Service Southeast Region. Jackson, Mississippi: Division of Migratory Birds, U.S. Fish and Wildlife Service.
Found at DOI: https://doi.org/10.3996/092018-JFWM-080.S16 (1.54 MB PDF); also available at https://www.fws.gov/columbiawildlife/MoistSoilReport.pdf
Reference S15. [USFWS] U.S. Fish and Wildlife Service. 2006. St. Lawrence Wetlands and Grassland Management District. Washington, D.C.: U.S. Fish and Wildlife Service.
Found at DOI: https://doi.org/10.3996/092018-JFWM-080.S17 (337 KB PDF); also available at https://www.fws.gov/northeast/facts/StLawrence06.pdf
Acknowledgments
A New York State Department of Environmental Conservation State Wildlife Grant “Value of Wetland Restoration Incentive Programs on Privately Owned Land for Species of Greatest Conservation Need,” and an Erb Family Foundation–University of Michigan Water Center Grant “Environmental and Socioeconomic Factors Associated with Public–Private Partnership Wetland Restoration Projects Benefiting Wildlife in the Great Lakes Watershed” provided funding in support of this project. We thank all of the landowners who participated in this study. Without your cooperation and dedication to the stewardship of our natural resources, this work would not have been possible. We also thank our partners including: Angelena Ross (New York State Department of Environmental Conservation); Jim Pullano and Kim Farrell (USDA Natural Resources Conservation Service); Gian Dodici, Carl Schwartz, and Tom Jasikoff (USFWS); Glenn Johnson (State University of New York Potsdam); Sarah Fleming (Ducks Unlimited), and all of the folks that have helped along the way: Matthew Valente, Kinga Stryszowska, Cody Merrill, Jayson Hajek, Maria Hargis, Robyn Andrusyszyn, Eric Marcy, Kallen Frey, Brittany Guarna, Laura Barlow, Jeremy Ozolins, Nychele Carley, John Sherry, and Felix Grimberg. We also thank two anonymous reviewers and the Associate Editor who provided comments that greatly improved this manuscript.
Any use of trade, product, website, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government.
References
Author notes
Citation: Benson CE, Carberry B, Langen TA. 2019. Public–private partnership wetland restorations provide quality forage for waterfowl in northern New York. Journal of Fish and Wildlife Management 10(2):323–335; e1944-687X. https://doi.org/10.3996/092018-JFWM-080
The findings and conclusions in this article are those of the author(s) and do not necessarily represent the views of the U.S. Fish and Wildlife Service.