Printer Friendly

Vegetative ecology of natural and constructed wetlands along the Leon River in Comanche County, Texas.

Abstract.--The vegetation found in two riverine wetlands, located along the Leon River in the West Cross Timbers, Comanche County, Texas, was compared for one year. A wetland constructed during 1999 and an adjacent natural reference wetland, established by beaver, were compared using relative coverage, density, frequency, and importance values. Wetland indicator status for plants and diversity were also assessed. Over four sampling periods, the constructed wetland underwent succession to become similar to the adjacent natural wetland. Common cat-tail (Typha latifolia), an obligate wetland species, dominated the inner core of the constructed wetland with marsh-elder (Iva annua) occupying the drier periphery. Ninety-five percent of the species sampled in the reference wetland were hydrophytes and plant succession followed a centrifugal model. The constructed wetland was dominated by marsh-elder and knot grass (Paspalum distichum) with only 67% of species being hydrophytes. Diversity was greater in the constructed wetland because it was in the arrival and establishment phase of wetland succession.

**********

Wetlands are defined by the presence of hydrophytic vegetation, periodic inundation, and hydric soils (United States Army Corps of Engineers (USACOE) 1987; Tiner 1999; van der Valk 2006; Mitsch & Gosselink 2007; Keddy 2010). Texas' wetlands are among its most valuable natural resources. These lands provide many economic and ecological benefits, including flood control, improved water quality, harvestable products, and habitat for abundant fish, shellfish, and wildlife resources (Texas Parks and Wildlife Department (TPWD) 1997; van der Valk 2006; Mitsch & Gosselink 2007; King et al. 2009). Texas contains a diversity of wetland habitat types, both natural and anthropogenic. Although wetlands in Texas comprise less than 5% of the state's total land area (TPWD 1997), Texas is one of 19 states that has exhibited the most significant losses of wetland ecosystems (Tiner 1984) with a 56% loss during the preceding 200 years (TPWD 1997). Texas also has 40 to 60% of its freshwater plant communities at moderate to high risk of being eliminated (Heinz Center 2008). Few wetlands have been studied in the West Cross Timbers region of Texas with one investigation involving only constructed wetlands (Williams & Hudak 2005) and another examining a constructed wetland with comparisons to a beaver-created marsh (Brister 2005), so little is known about wetland plant communities and succession.

The West Cross Timbers area of Texas (Figure 1) lies south of the Red River, west of the Fort Worth Prairie, north of the Lampasas Cut Plain, and east of the Rolling Plains. It encompasses an area of 1,665,686 ha and is considered a mosaic of prairie, woodland, and savannah vegetation characterized by post oak (Quercus stellata) and blackjack oak (Quercus marilandica) as the dominant woody species (Dyksterhuis 1948; Diggs et al. 1999; Hoagland et al. 1999). The Red, Trinity, Brazos, and Colorado River basins all drain a portion of the West Cross Timbers area of Texas, with numerous tributaries scattered throughout the area. Each of these rivers and their associated tributaries contribute a significant amount of acreage to the Texas riparian wetland component. One such tributary is the Leon River.

The headwaters of the Leon begin in Eastland County, near the town of Eastland, and flow southeastward through Comanche, Hamilton, and Coryell Counties to be joined by the Lampasas River in Bell County and the San Gabriel River in Milam County to form the Little River, a major tributary of the Brazos River. Two reservoirs, Lake Leon and Proctor Lake, are located on the upper reaches of the Leon. The Leon River contributes water to a number of riparian wetland complexes as it flows into and out of Proctor Lake.

One complex, approximately 32 ha in size, is located below the dam at Proctor Lake. It consists of two wetland areas located between the Leon River channel and the outflow channel that was built by the USACOE when the lake was constructed. One of the wetland areas, a freshwater marsh, has been created by beaver activity, resulting in dams which trap water from the Leon River channel and from seepage through the earthen portion of the lake dam. The age of the marsh is unknown, but according to lake personnel, it is thought to be as old as the lake, which became operational in 1963. For purposes of this study, the marsh will be referred to as the reference wetland.

The second of the wetland areas, a constructed wetland built by USACOE, TPWD, and Ducks Unlimited to increase waterfowl habitat is located to the southeast (downriver) of the reference wetland. It was completed in 1999 and did not receive any significant hydrologic input until the Winter/Spring of 2001. The constructed wetland was not seeded, nor was any vegetation introduced into the wetland. The constructed wetland lies adjacent to the reference wetland and the two are separated by a berm in order to allow independent control of water levels. Both wetland areas would be considered inland, riverine, freshwater marshes (Mitsch & Gosselink 2007).

Riverine marshes often border riparian forests or occupy pockets within them, sometimes in abandoned oxbows and at the headwaters of rivers. In addition, beavers build wetlands by damming streams, creating small riverine marshes that attract wildlife, especially waterfowl (van der Valk 2006; Mitch & Gosselink 2007). Beavers often play significant roles in wetland development and maintenance. If their dams hold, this creates a stable hydroperiod. If their dams frequently wash out, they cause water level fluctuations, thus increasing plant diversity. Increased plant diversity usually causes increased waterfowl and mammal diversity in and around the wetland. Vegetation in freshwater marshes created by beavers is often characterized by monocots such as Typha, moisture-adapted grasses, and sedges (Mitsch & Gosselink 2007). The additional nutrients added to the water as a result of beaver and other mammals, waterfowl, and organic material from decaying vegetation, increase the biodiversity of the wetland by feeding plankton, invertebrates, crustaceans, amphibians, fish, and reptiles. The beaver help to support the entire ecosystem of organisms, thus becoming a keystone species (Primack 2010). Due to their dam construction regulating water level and stream flow, beavers can convert an otherwise unfavorable area into one that is habitable by wetland biota (Schmidly 2004).

Loss of wetland habitat is a serious environmental problem (Mitsch & Gosselink 2007; Heinz Center 2008). As wetlands continue to decrease, not only in Texas but throughout the world, it is becoming increasingly important that wetland ecosystems be better understood in order to facilitate quality mitigation and construction practices. Since few wetlands in the West Cross Timbers region of Texas have been examined, the ecological data for this investigation will serve as a baseline for future studies of wetlands in the region. The objective of this investigation is to assess changes in vegetational ecology of the reference and constructed wetlands at Proctor Lake Reservoir.

MATERIALS AND METHODS

The ecological comparisons were conducted below the dam at Proctor Lake in the two previously mentioned wetland areas located in Comanche County, Texas (Figure 1). Both wetlands were sampled to examine vegetation over four time periods. Sampling dates for vegetation were chosen based on growing season data for the area. The mean length of the frost-free period for Comanche County, Texas is about 230 days. The frost-free period begins in mid-March and lasts until mid-November. Sampling was done from August-September 2001, March-April 2002, June-July 2002, and September-October 2002.

Soil samples from each quadrat where vegetation was sampled were taken with a 38 cm soil sampling probe and compared to a Munsell soil color chart for hydric soil color indicators. Based on this test, all soils sampled in the reference wetland were hydric and two-thirds of the samples from the constructed wetland were classified as hydric.

Based on visual observations of inundation, soil saturation, and watermarks on surrounding vegetation, hydrology between the two sites was similar except in the Fall of 2001, at the start of the investigation, when about one-third of the quadrats in the constructed wetland had no visual signs of inundation, soil saturation, or watermarks on surrounding vegetation. Thus, when vegetation was sampled for the first time in the investigation, hydrology in the constructed wetland was more variable than the hydrology of the reference wetland, but during the three following sample periods, hydrology was constant in quadrats for both wetlands.

The ecological comparison for vegetation was conducted by first establishing a baseline that ran parallel to the Leon River and included both wetland areas. The baseline was approximately 890 m and was intersected by 10 transects that traverse the wetland areas. The baseline was divided into equally spaced segments (107 m apart) in the natural wetland and in the constructed wetland (71 m apart). During each sampling period, transect starting point for each segment was determined using a random numbers table, and the transect was situated perpendicular to the baseline of the wetland. Each transect had two randomly-stratified observation points determined by using a random numbers table where a 0.25 [m.sup.2] quadrat was used to determine the number and percent coverage of plants. Taxonomy for plants followed Diggs et al. (1999) and scientific names with authorities are included in Tables 1-4. Once identified, vegetation was characterized at each observation point along each transect utilizing the following methodology as modified from USACOE (1987) and Brower et al. (1990).

Herbs were sampled in both wetlands. Herbs according to USACOE (1987) include all non-woody plants and woody plants < 1 m in height. In order to measure herbaceous vegetation, a 0.25 [m.sup.2] quadrat was randomly placed on the transect within the 1 [m.sup.2] observation point. Each species found within the quadrat was then identified, counted, and its percent cover estimated. An importance value (IV) was assigned to each herbaceous or woody seedling species found at each observation point, based on the sum of the relative coverage, relative density, and relative frequency of each species, while the percent relative importance value (%RIV) was determined by comparing the IV of each particular species to the total IV of all of the species (Fredrickson 1979; Brower et al. 1990, Smith 1996). A diversity index for all herbaceous and woody seedling species was applied using the Shannon-Weaver index, based on information obtained from the above data (Miller 1985; Brower et al. 1990). In addition, richness and evenness were calculated (Brower et al. 1990).

Those plants that have adapted to growth in water or in soil, which is at least periodically saturated or inundated, are often referred to as hydrophytic plants, or simply hydrophytes (USACOE 1987). Plants have been categorized, based on their life history; into five separate groups called wetland indicator categories (USACOE 1987; United States Fish and Wildlife Service (USFWS) 1997; Tiner 1999). These include obligate wetland species (OBL) that occur in wetlands greater than 99% of the time. Facultative wetland plants (FACW-, FACW, and FACW+) occur in wetlands 67-99% of the time. Facultative plants (FAC-, FAC, and FAC+) occur in wetlands 34-66% of the time and facultative upland plants (FACU+, FACU, and FACU-) occur in wetlands 1-33% of the time. Upland plants (UPL) occur in wetlands less than 1% of the time. This classification system allows one to categorize vegetation in sampling units, and if greater than 50% of the dominant vegetation is OBL, FACW+, FACW, FACW-, FAC+, or FAC it is considered hydrophytic vegetation and the criterion for wetland vegetation in that sampling unit has been established (USACOE 1987). Each component of the vegetation data was classified by its wetland indicator status (USFWS 1997) and the 50/20 rule was then applied in order to determine dominance of species. The species were ranked in descending order of dominance and those species that immediately exceeded 50%, plus any additional species comprising 20% or more of the total dominance, were considered dominant. If more than 50% of the dominant plant species sampled within each observation point were obligate wetland (OBL), facultative wetland (FACW+, FACW, FACW-), or facultative (FAC+, FAC), the area covered by the observation point was considered positive for hydrophytic vegetation (USACOE 1987). In addition, a Mann-Whitney test was used to determine if there were significant differences between the constructed and natural wetland for herbaceous and woody seedling species (Brower et al. 1990).

RESULTS

Relative importance values (RIV%), a measure of dominance for herbaceous vegetation in the reference and constructed wetlands, were determined for each sampling period. During the Fall 2001 sampling period, common cat-tail (Typha latifolia; 49.8%) and Bermuda grass (Cynodon dactylon; 12.7%) were the most important plants identified in the reference wetland, while frogfruit (Lippia nodiflora; 21.3%), Johnson grass (Sorghum halepense; 19.2%), and barnyard grass (Echinochloa muricata; 15.5%) were most important in the constructed wetland (Table 1). During the Spring 2002 sampling period, common cat-tail (77.3%) was found to be the most important species in the reference wetland, while in the constructed wetland, marsh-elder (Iva annua; 34.0%) and fiddle dock (Rumex pulcher, 19.6%) were found to be dominant (Table 2). During the Summer 2002 sampling period the reference wetland was dominated by marsh-elder (42.7%) and Bermuda grass (19.6%), whereas the constructed wetland had marsh-elder (31.8%), and (for the first time) common cat-tail (24.4%) and knot grass (Paspalum distichum; 23.7%) as dominants (Table 3). Knot grass (32.1%) and marsh-elder (18.7%) were found to be most important in the reference wetland during the Fall 2002 sampling period, and knot grass (20.0%), water-hemp (Amaranthus rudis; 12.7%), cocklebur (Xanthium strumarium; 12.4%), and marsh-elder (10.4%) all shared co-importance in the constructed wetland (Table 4). Total %RIV for the reference wetland, when the data was analyzed as a combined total, indicated common cat-tail (25.7%) being most important, followed by marsh-elder (25.6%) (Table 5). The constructed wetland was dominated by marsh-elder (17.6%), but had relatively high importance values for knot grass (17.5%), common cat-tail (10.6%), and cocklebur (8.9%) (Table 5).

In the Fall of 2001, the percentage of herbaceous plant species sampled in the reference wetland that were classified as hydrophytes was 70%, whereas 56% of the plant species sampled in the constructed wetland were classified as hydrophytes (Tables 1 and 5). One hundred percent of the plant species were classified as hydrophytes in the reference wetland Spring 2002, whereas 88% were classified as hydrophytes in the constructed wetland (Tables 2 and 5). During the next two sampling periods, the percentage of plant species classified as hydrophytes were about the same, 80 and 82% in the reference wetland vs. 83 and 78%, respectively, in the constructed wetland (Tables 3, 4, and 5). In the analysis of the data as a combined total, the reference wetland had a greater percentage of plant species classified as hydrophytes than did the constructed wetland, 83% and 72% respectively (Table 5).

Thirty-two species were identified from 80 quadrats in the two wetland areas. Of those 32 species, 13 (40.6%) were found to be present in both wetland areas, eight (25%) were found exclusively in the reference wetland and 11 (34.4%) were found exclusively in the constructed wetland (Table 5). Twenty-eight of the 32 species were natives, with four species being introduced. All four introduced species were found in the reference wetland (Bermuda grass, knot grass, Johnson grass, and fiddle dock and three of the four (excluding knot grass) were identified in the constructed wetland. The species identified in the reference wetland consisted of eight annuals and 13 perennials (Table 5), whereas the species identified in the constructed wetland consisted of 10 annuals and 14 perennials (Table 5).

Herbaceous plant species richness for Fall 2001 (Table 6) was slightly greater in the reference wetland (10) than in the constructed wetland (nine). However, during the Spring 2002 sampling period, the constructed wetland surpassed the reference wetland in plant species richness (eight and five, respectively) (Table 6). During the Summer 2002 sampling period, the reference wetland had higher richness with 11 species compared to eight in the constructed wetland (Table 6). During Fall 2002 (Table 6), the constructed wetland surpassed the reference wetland in species richness with 18 species, while the reference wetland had 11. When the data were analyzed as a combined total, the constructed wetland was found to have a slightly greater richness of 27 species, compared to 25 for the reference wetland (Table 6).

Evenness values (Table 6) were greater in the constructed wetland for every sampling period. Reference wetland evenness values ranged from a low of 0.38 in Spring 2002 to a high of 0.67 in the Fall 2001, for a combined total value of 0.48 (Table 6). Constructed wetland values ranged from a low of 0.53 in Summer 2002 to a high of 0.78 in Fall 2001, for a combined total value of 0.64 (Table 6).

Shannon's diversity index was calculated for each sampling period. Like evenness, diversity (Table 6) was found to be greater in the constructed wetland than in the reference wetland during every sampling period. Diversity values in the reference wetland ranged from a low of 0.62 in Spring 2002, to a high of 1.53 during Fall 2001. When the data were analyzed as a combined total, the Shannon's diversity value was 1.56. Diversity values in the constructed wetland ranged from a low of 1.10 during the Summer 2002 sampling period to a high of 1.86 in Fall 2002. When the data were analyzed as a combined total, a Shannon's diversity value of 2.11 was obtained (Table 6).

In the Fall 2001 sampling period, 100% of the quadrats sampled in the reference wetland met hydrophytic plant requirements (greater than 50% of the dominant species were FAC or wetter), while only 60% of those in the constructed wetland met the same requirements. During the Spring 2002 sampling period, 80% of the reference wetlands' quadrats met hydrophytic vegetation requirements compared to 40% of the constructed wetlands' quadrats. During the Summer 2002 sampling period, the percentage of quadrats meeting the hydrophytic vegetation requirements in both the reference and constructed wetlands had increased to 100 and 80% respectively. During the Fall 2002 sampling period, the percentage of quadrats meeting hydrophytic vegetation requirements in the reference wetland remained at 100%, while the constructed wetlands percentage dropped to 70. These results, when analyzed as a whole, show that the reference wetland met its hydrophytic vegetation requirements 95% of the time, whereas the constructed wetland met its hydrophytic vegetation requirements 63% of the time.

In addition to the above analyses, a non-parametric Mann-Whitney statistical test was conducted on richness, evenness, diversity, plant density, and percent plant coverage, but none were found to be significantly different in comparisons between the reference and constructed wetlands.

DISCUSSION

Using the standards of hydrophytic vegetation, soils, and hydrology (USACOE 1987), the reference area would be considered a wetland throughout this investigation. The constructed wetland had vegetation that was increasingly hydrophytic throughout the investigation, had about two-thirds of its soil samples classified as hydric, and had a constant hydrology, similar to the reference wetland, following the first sampling period.

Based on Relative Importance Values (RIV%) for the Fall 2001 sampling period, the most important species in the reference wetland were common cat-tail and Bermuda grass (Table 1). Common cat-tail can be an invasive species that has the ability to create large monocultures and out-compete other species. This is due to its rapid growth rate and the fact that it asexually reproduces using rhizomes, which spread quickly. Cat-tail invasion can be problematic because dense monocultures reduce the ratio of open water to vegetation, reduce the diversity of vegetation, and therefore reduce the habitat value of the site (Brown and Bedford 1997; Mitsch & Gosselink 2007). In addition, common cat-tail is an obligate wetland species (found in wetlands >99% of the time) and most of the reference wetland was found to be saturated for the majority of the year during the entire study. Therefore, common cat-tail would be well adapted for this environment. Each individual has the ability to release thousands of seeds, which can drift considerable distances in either the air or the water, giving it the ability to easily disperse.

Bermuda grass was also found to be one of the dominants in the reference wetland. However, it is a FACU+ species, which gives it about a 10% likelihood of being found in a wetland. The probability is low, but wetlands are dynamic ecosystems, which are constantly changing and Bermuda grass is an introduced species that can survive in a variety of conditions, often as a weed, in many parts of Texas (Diggs et al. 1999). It occurred along the wetland edge during the drier months of the growing season because the soil surface along the edge of the wetland had a tendency to dry, allowing species not generally adapted to wetlands a chance to compete with some of the wetland species, which may have died or become dormant. Bermuda grass has also been extensively introduced for livestock hay and grazing in fields adjacent to the Leon River.

In the constructed wetland, during Fall 2001, frogfruit, Johnson grass, and barnyard grass were found to be the most important species present (Table 1). Frogfruit has a wetland indicator status of FAC+ (about 60% chance of occurring in a wetland), Johnson grass is FACU (20% chance of occurring in a wetland), and barnyard grass is FACW (80% chance of occurring in a wetland), averaging a little more than 50% chance of occurring in the wetland. These species are often considered to be weedy, inhabiting disturbed areas (Diggs et al. 1999).

During the Spring 2002 sampling season, the reference wetland was dominated by common cat-tail, while the constructed wetland was co-dominated by marsh-elder and fiddle dock (Table 2). They were classified as FAC+ and FACW- (70% chance of occurring in a wetland) respectively. Marsh-elder is native, occurring in low, moist habitats, whereas fiddle dock is introduced and usually occurs in wet, disturbed areas (Diggs et al. 1999). The constructed wetland was being colonized by more wetland species, which would be expected following a fall and winter of flooding by reservoir personnel, in order to create waterfowl habitat. If the wetland hydrology stabilizes over time, the species composition should become more hydrophytic and change into a secondary seral stage or even a climax community.

During the Summer 2002 sampling period, the reference wetland was dominated by marsh-elder, and Bermuda grass (Table 3). The presence of Bermuda grass as a dominant is not necessarily the result of a shift in the reference wetland's vegetation, but a result of random sampling that occurred along the drier margins during that sampling period. Marsh-elder also dominated the outer core of the reference wetland, where the soil was drier, while common cat-tail dominated the inner core of the wetland, which was saturated most of the time.

During this same time period (Summer 2002), the constructed wetland was also dominated by marsh-elder, as well as common cat-tail and knot grass (Table 3), a FACW+ species (chance of occurring in a wetland 90%). Following the Fall/Winter inundation instigated by reservoir personnel, the constructed wetland was colonized with more hydrophytic species.

During the Fall 2002 sampling period, the reference wetland was dominated by knot grass and marsh-elder (Table 4). During this time of the year there was generally little precipitation, allowing the edge of the reference wetland to recede toward the core. This drying effect around the edges of the wetland allowed more facultative species a chance to re-establish themselves.

The constructed wetland was dominated by knot grass, waterhemp (FAC; with a 50% chance of being found in a wetland), cocklebur (FAC-; with a 40% chance of occurring in a wetland), and marsh-elder (Table 4). Water-hemp and cocklebur were two new co-dominants, both colonizing the edges of the wetland after drawdown of water levels.

Overall, the reference wetland was dominated by common cattail and marsh-elder (Table 5). The common cat-tail was found throughout the inner core of the wetland, while marsh-elder was found outside the core and toward the edges. These results follow the centrifugal organization model (Mitsch & Gosselink 2007; Keddy 2010), which states that the core habitat in wetlands has low disturbance and high fertility and is dominated by species that form dense canopies such as Typha. Peripheral habitats represent different kinds and combinations of stresses (infertility, disturbance) and support distinctive plant associations (Mitsch & Gosselink 2007). In this case, the stressor was the periodic draw down of the wetlands, allowing the core habitat to pulse inward and outward depending on the hydroperiod.

The dominant plant in the constructed wetland, overall, was marsh-elder (Table 5). In a study of excavated depressions near an earthen dam in north-central Texas, the dominant herb was also marsh-elder (Williams & Hudak 2005). However, in the Proctor Lake constructed wetland, relatively high importance values were found for knot grass, common cat-tail, and cocklebur. Knot grass and common cat-tail were not found in excavated depressions and cocklebur was rarely encountered in the excavated depressions (Williams & Hudak 2005). Each of these species are classified as FAC- or wetter. Overall, the constructed wetland was dominated by greater than 50% hydrophytes and like the reference wetland seemed to be developing centrifugal organization (Deberry & Perry 2005).

Species richness, overall, was found to be slightly greater in the constructed wetland (Table 6). This may be attributed to the fact that the constructed wetland was recently uninhabited due to the disturbance from construction. Early colonizing species took advantage of the uninhabited space and colonized the area causing a greater species richness, like that reported for constructed wetlands in Grimes County, Texas (Noon 1996). Successful early colonists have adaptations that enable them to establish quickly on open sites. They are generally small and low growing, have short life cycles, and reproduce annually by seeds or send out new shoots from buds near the ground (van der Valk 1981; Mitsch & Gosselink 2007). They produce large numbers of easily dispersed small seeds, respond quickly to disturbance (especially exposure of mineral soil), and attain dominance quickly by suppressing growth of any later serai stage plants that might exist as seedlings beneath them (Smith 1996; Mitsch & Gosselink 2007). These early colonists are tolerant of fluctuating environmental conditions; particularly a wide range of daily temperatures on the soils surface, alternate wetting and drying, and intense light (Smith 1996; Mitsch & Gosselink 2007). These species represent what has been termed the arrival and establishment phase of wetland succession (Noon 1996) and the higher diversity found in the constructed wetland likely is a result of early successional phase. Richness values in restored and natural prairie wetlands that were adjacent to each other in northern Iowa, three years following restoration, had natural wetlands with 46 total species compared to only 27 total species occurring in restored wetlands (Galatowitsch & van der Valk 1996). Noon (1996), using data from eight constructed wetlands in Texas, found that the older the wetlands became, the fewer species occurred. Therefore, it is likely that, once the constructed wetland moves out of the arrival and establishment phase, species numbers will decrease.

Species present during the first few years of vegetation establishment that were not planted or seeded are assumed to be volunteers from offsite sources, and therefore represent the primary seral stage of vegetation succession (Reinhartz & Wame 1993; DeBerry & Perry 2004). In addition, vegetation in adjacent wetlands may be a potential seed source; particularly from habitats dominated by herbaceous species that are more likely to colonize a young substrate (DeBerry & Perry 2004). Shared species composition between the wetlands was greater than 40% because the reference wetland likely served as a seed source for colonization of the constructed wetland.

The constructed wetland had greater diversity then the reference wetland during each sampling period (Table 6). Data from eight constructed wetlands in Texas (Noon 1996) indicated that diversity was highest at younger sites. Marshes, when compared to savannah type habitat, exhibited low species diversity in a study of wetland habitats along a portion of the North Fork of the Forked Deer River in West Tennessee. The savannah habitat was seasonally flooded, bringing in new seed sources and nutrients, while the marsh habitat retained water year around (Miller 1985). Ice scouring, infertile sandy soils, flooding by beavers, and open shorelines are among the stresses that shift communities to be less vegetatively productive, albeit possibly more diverse, assemblages (Mitsch & Gosselink 2007). Fluctuating hydroperiod in wetlands can promote dominance by annuals and nonclonal perennials in zones or patches within the wetland (Collins & Wein 1995; Mitsch & Gosselink 2007), and this patchiness and annual duration may increase diversity. Because, with the exception of the first sampling period, hydrology of the two wetlands was similar, differences in diversity are likely due to differences in successional stages between the two wetlands. The reference wetland is older and exhibited hydrophytic vegetation 95% of the time whereas the younger constructed wetland exhibited hydrophytic vegetation only 63% of the time. Assuming hydrology remains constant over time, the constructed wetland should develop more hydrophytic vegetation.

The adjacent riverine forest does not invade the marsh because the beaver dam and constructed wetland hydrology maintains a habitat too wet for the establishment of most tree species. Mitchell & Niering (1993) found that 30 year old beaver wetlands decimated the anchored forest wetland and replaced it with an abundance of graminoids in a northwestern Connecticut bog. The beaver-created marsh functions in a similar fashion and maintains separate marsh and upland forest communities. Because of more consistent hydrology and the older age of the reference wetland, common cattail, an obligate wetland species, dominates the inner core area with marsh-elder, a FAC+ species, occupying the drier periphery. Ninety-five percent of the species sampled in the reference wetland were hydrophytes. These results closely resemble a successional model of centrifugal organization that describes the distribution of species and vegetation types along gradients caused by a combination of environmental constraints (Wisheu & Keddy 1992; Mitsch & Gosselink 2007; Keddy 2010). In eastern North America, the core habitats are often dominated by species that form dense canopies such as Typha. Habitats peripheral to the core depend on different kinds and combinations of stress and disturbance. When beavers create disturbance, the peripheral habitats include species of sedge near the Typha core and wet-adapted forb species near the edges (Wisheu & Keddy 1992; Mitsch & Gosselink 2007; Keddy 2010). A similar plant community was observed in the reference wetland at Proctor Lake.

In contrast, the constructed wetland was dominated by marsh-elder and knot grass, both FAC+ species, with a greater diversity of vegetation of which 67% were hydrophytes. Richness, evenness, and Shannon's diversity were greater in the constructed wetland. This was likely due to its earlier successional stage with plants being a part of the arrival and establishment phase (Noon 1996).

Because common cat-tail was more important in later sampling in the constructed wetland it may develop according to the centrifugal organization with a Typha core (Deberry & Perry 2004). Mitsch et al. (2005) describes a naturally colonizing, riverine wetland in Ohio as being dominated by Typha. Noon (1996) proposed two phases in early constructed wetland primary succession called the arrival and establishment phase, which is characterized by random arrival and establishment of species, and the autogenic dominance phase, which is characterized by biomass production and competition between species. Because of its greater diversity, the constructed wetland is probably beginning to finish the first successional stage and is moving into the second phase whereas the reference wetland is in the autogenic dominance phase. In van der Valk's (1981) model of freshwater wetland vegetation dynamics, two basic types of wetland species are recognized: (1) species with long-lived propagules that are in the wetland seedbank and can grow when suitable conditions occur, and (2) species with short-lived propagules that can only grow in the wetland if they reach it during a period when conditions are suitable for germination. This causes the wetland to function as a sieve, allowing only establishment of certain species at any given time and dependent on whether the wetland is flooded or in a drawdown (van der Valk 1981; Mitsch & Gosselink 2007). At Delta Marsh in Canada and marshes of Eagle Lake in Iowa, flooding the marshes after drawdowns resulted in plant communities that contained or were dominated by a species of Typha (van der Valk 1981). Of herbaceous species identified in the Leon River wetlands, there were equal numbers of perennial species, but numbers of annuals differed (Table 5). The constructed wetland had more annual species than the reference (10 vs. 7) likely due to its early successional stages (Noon 1996) and more variable early hydrology, while the older riverine marsh, created by the beavers, was centrifugally organized (Wisheu & Keddy 1992; Deberry & Perry 2004; Mitsch & Gosselink 2007; Keddy 2010) with a perennial common cat-tail core.

LITERATURE CITED

Brister, J. S. 2005. Soils, hydrology, vegetative ecology, and floristics of natural and constructed wetlands along the Leon Riverin the West Cross Timbers, Comanche County, Texas. Unpublished Masters Thesis. Tarleton State University, Stephenville, Texas, 115 pp.

Brower J. E., J. H. Zar, & C. von Ende. 1990. Field and laboratory methods for general ecology, Third Edition. William C. Brown Publishers, Dubque, Iowa, 226 pp.

Brown, S. C. & B. L. Bedford. 1997. Restoration of wetland vegetation with transplanted wetland soil: An experimental study. Wetlands, 17(3):424-437.

Collins, B. & G. Wein. 1995. Seed bank and vegetation of a constructed reservoir. Wetlands, 15(4):374-385.

DeBerry, D. A. & J. E. Perry. 2004. Primary succession in a created freshwater wetland. Castanea, 69(3): 185-193.

Diggs, G. M., Jr., B. L. Lipscomb & R. J. O'Kennon. 1999. Shinners and Mahler's Illustrated Flora of North Central Texas. Botanical Research Institute of Texas, Fort Worth, Texas, 1626 pp.

Dyksterhuis, E. J. 1948. The vegetation of the Western Cross Timbers. Ecol. Monogr., 18(3):325-376.

Fredrickson, L. H. 1979. Floral and faunal changes in lowland hardwood forests in Missouri resulting from channelization, drainage, and impoundment. United States Fish and Wildlife Service and University of Missouri Cooperative Wildlife Research Unit, Columbia, Missouri, 120 pp.

Galatowitsch, S. M. & A. G. van der Valk. 1996. The vegetation of restored and natural prairie wetlands. Ecol. Appli., 6(1): 102-112.

Heinz Center 2008. The state of the nation's ecosystems 2008: measuring the land, water, and living resources of the United States. Island Press, Washington D. C., 352 pp.

Hoagland, B. W., I. H. Butler, F. L. Johnson, & S. Glenn. 1999. The Cross Timbers. Pp. 231-245, in Savannas, Barrens, and Rock Outcrop Plant Communities of North America. (Anderson, R. C., J. S. Fralish, & J. M. Baskin, eds.), Cambridge University Press, New York, New York, 470 pp..

Keddy, P. A. 2010. Wetland Ecology: Principles and Conservation, Second edition. Cambridge University Press, New York, New York, 497 pp.

King, S. L., R. R. Sharitz, J. W. Groninger, & L. L. Battaglia. 2009. The ecology, restoration, and management of southeastern floodplain ecosystems: A synthesis. Wetlands, 29(2):624-634.

Miller, N. A. 1985. A vegetation-habitat study along a portion of the North-Fork-Forked Deer River in West Tennessee. J. Tennessee Acad, of Sci., 60(2):51-56.

Mitchell, C. C. & W. A. Niering. 1993. Vegetation change in a topogenic bog following beaver flooding. Bull. Torrey Bot. Club, 120(2): 136-147.

Mitsch, W. J. & J. B. Gosselink. 2007. Wetlands, Fourth Edition. John Wiley and Sons, Hoboken, New Jersey, 582 pp.

Mitsch, W. J., L. Zhang, C. J. Anderson, A. E. Altor, & M. E. Hernandez. 2005. Creating riverine wetlands: Ecological succession, nutrient retention, and pulsing effects. Ecol. Eng., 25(5):510-527.

Noon, K. F. 1996. A model of created wetland primary succession. Landscape Urban Planning, 34(2):97-123.

Primack, R. B. 2010. Essentials of Conservation Biology, Second Edition. Sinauer Associates, Sunderland, Massachusetts, 601 pp.

Reinhartz, J. A. & E. L. Wame. 1993. Development of vegetation in small created wetlands in southeastern Wisconsin. Wetlands, 13(3): 153-164.

Schmidly, D. J. 2004. The mammals of Texas, Sixth Edition. University of Texas Press, Austin, Texas, 501 pp.

Smith, R. L. 1996. Ecology and Field Biology, Fifth Edition. Harper Collins College Publishers, New York, New York, 740 pp.

Texas Parks & Wildlife Department (TPWD). 1997. Texas Wetland Conservation Plan. Texas Parks & Wildlife Department, Austin, Texas, 64pp.

Tiner, R. 1984. Wetlands of the United States: Current status and recent trends. United States Fish and Wildlife Service, Habitat Resources, Newton Comer, Massachusetts, 57 pp.

Tiner, R. 1999. Wetland indicators: A guide to wetland identification, delineation, classification, and mapping. Lewis Publishers, New York, New York, 392 pp.

United States Army Corps of Engineers (USACOE). 1987. Corps of Engineers Wetlands Delineation Manual. U.S. Army Corps of Engineers Waterways Experiment Station, Vicksburg, Mississippi, 92 pp.

United States Fish & Wildlife Service. 1997. National list of vascular plant species that occur in wetlands: 1996 National Summary. United States Department of the Interior, Washington, D.C., 209 pp.

van der Valk, A. G. 1981. Succession in wetlands: A Gleasonian approach. Ecology, 62(3):688-696.

van der Valk 2006. The biology of freshwater wetlands. Oxford University Press, New York, New York, 173 pp.

Williams L. L. & P. F. Hudak. 2005. Hydrology and plant survival in excavated depressions near an earthen dam in north-central Texas. Environ. Geol., 48(6):795804.

Wisheu, I. C. & P. A. Keddy. 1992. Competition and centrifugal organization of plant communities: theory and tests. J. Veg. Sci., 3(2): 147-156. ADN at: nelson@tarleton.edu

Jeffrey S. Brister (1) and Allan D. Nelson (2)

(1) Natural Resource Conservation Service, United States Department of Agriculture Service Center, 5040 Loop 340, Waco, Texas 76706

(2) Department of Biological Sciences, Box T-0I00, Tarleton State University, Stephenville, Texas 76402

Table 1. Species with common names in parentheses as well as
relative coverages (RC), relative densities (RD), relative
frequencies (RF), and percent relative importance values (%RIV)
for plants in the reference and constructed wetlands for the Fall
of 2001. Asterisk (*) indicates dominant species based on the 50/
20 rule for %RIV (see methods).

Species/Common name                          RC     RD     RF    %RIV

Reference Wetland
Cardiospermum halicacabum L. (Common        12.7    2.8   13.6    9.7
balloonvine)
Cynodon dactvlon (L.) Pers. (Bermuda         3.9   25.0    9.1   12.7
grass) *
Eupatorium serotinum Michx. (Fall           10.6    2.8    9.1    7.5
boneset)
Heliotropium curassavicum L. (Salt           0.4    0.9    4.5    2.0
heliotrope)
Lippia nodiflora (L.) Michx. (Frogfruit)     0.6    0.9    4.5    2.0
Paspalum dilatatum Poir. (Dallis grass)      2.7    3.7    9.1    5.2
Salix nigra Marshall (Black willow)          1.1    3.7    4.5    3.1
Teucrium canadense L. (Wood-sage)            3.3    2.8    4.5    3.6
Typha latifolia L. (Common cat-tail) *      64.1   49.1   36.4   49.8
One unknown plant in quadrats                0.6    8.3    4.5    4.5

Constructed Wetland
Amaranthus rudis J. D. Sauer (Water-         4.8   13.2   16.0   11.3
hemp)
Chamaesyce serpens (Kunth) Small (Mat        2.6    4.4    8.0    5.0
euphorbia)
Echinochloa muricata (P. Beauv.)            24.5    6.0   16.0   15.5
Femald (barnyard grass) *
Iva annua L. (Marsh-elder)                   7.1    9.9    8.0    8.3
Lippia nodiflora (L.) Michx. (Frogfruit) *  16.5   29.3   16.0   21.3
Schedonnardus paniculatus (Nutt.) Trel.      5.3    0.5    4.0    3.3
(Tumble grass)
Sorghum halepense (L.) Pers. (Johnson       16.8   32.9    8.0   19.2
grass) *
Typha latifolia L. (Common cat-tail)         1.9    1.1    8.0    3.7
Xanthium strumarium L. var. canadense       18.5    2.7   16.0   12.4
(Mill.) Torr. & A. Gray (Cocklebur)

Table 2. Species with common names in parentheses as well as
relative coverages (RC), relative densities (RD), relative
frequencies (RF), and percent relative importance values (%RIV)
for plants in the reference and constructed wetlands for the
Spring of 2002. Asterisk (*) indicates dominant species based on
the 50/20 rule for %R1V (see methods).

Species/Common name                          RC     RD     RF    %RIV

Reference Wetland
Panicum obtusum Kunth (Vine-mesquite)        0.5    1.2    9.1    3.6
Pluchea odorata (L.) Cass (Purple pluchea)   0.5    1.2    9.1    3.6
Tvpha latifolia L. (Common cat-tail) *      96.4   61.0   54.5   77.3
Two unknown plants in quadrats               2.8   16.7   27.3   12.0

Constructed Wetland
Carex tetrastachva Scheele (Four-angle       4.4    8.7   10.0    7.7
caric sedge)
Cvnodon dactylon (L.) Pers. (Bermuda         2.8    0.4   10.0    4.4
grass)
Cyperus acuminatus Torr. & Hook, ex          1.1    0.8   10.0    4.0
Torr. (Taper-leaf flat sedge)
Elymus virginicus L. (Virginia wild rye)     1.1    5.1   10.0    5.4
Iva annua L. (Marsh-elder) *                32.2   49.8   20.0   34.0
Paspalum distichum L. (Knot grass)          30.6    7.1   20.0   19.2
Rumex pulcher L. (Fiddle dock) *            22.2   26.5   10.0   19.6
Typha latifolia L. (Common cat-tail)         5.6    1.6   10.0    5.7

Table 3. Species with common names in parentheses as well as
relative coverages (RC), relative densities (RD), relative
frequencies (RF), and percent relative importance values (%RIV)
for plants in the reference and constructed wetlands for the
Summer of 2002. Asterisk (*) indicates dominant species based
on the 50/20 rule for %RIV (see methods).

Species/Common name                        RC     RD     RF    %RIV

Reference Wetland
Ammonia coccinea Rottb. (Toothcup)         0.1    0.2    3.3    1.2
Cardiospermum halicacabum L. (Common       1.0    0.4   10.0    3.8
balloonvine)
Cynodon dactylon (L.) Pers. (Bermuda       8.2   40.7   10.0   19.6
grass) *
Cyperus sp. (Flat sedge)                   0.7    0.5    3.3    1.5
Iva annua L. (Marsh-elder) *              55.4   52.7   20.0   42.7
Rumex pulcher L. (Fiddle dock)             0.3    0.2    6.7    2.4
Salix nigra Marshall (Black willow)        1.4    0.0    3.3    1.6
Teucrium canadense L. (Wood-sage)          0.1    0.0    3.3    1.2
Typha latifolia L. (Common cat-tail) *    29.3    2.7   20.0   17.3
Xanthium strumarium L. var. canadense      3.2    1.2   16.7    7.0
(Mill.) Torr. & A. Gray (Cocklebur)
One unknown plant in quadrats              0.3    1.2    3.3    1.6

Constructed Wetland
Ammonia coccinea Rottb. (Toothcup)         5.1    3.9   17.4    8.8
Iva annua L. (Marsh-elder) *              33.7   53.0    8.7   31.8
Lippia nodiflora (L.) Michx. (Frogfruit)   0.7    1.0    8.7    3.5
Paspalum distichum L. (Knot grass) *      18.3   35.3   17.4   23.7
Typha latifolia L. (Common cat-tail) *    37.3    5.5   30.4   24.4
Xanthium strumarium L. var. canadense      0.2    0.1    4.3    1.6
(Mill.) Torr. & A. Gray (Cocklebur)
Two unknown plants in quadrats             4.5    1.2   13.0    6.3

Table 4. Species with common names in parentheses as well as
relative coverages (RC), relative densities (RD), relative
frequencies (RF), and percent relative importance values (%RIV)
for plants in the reference and constructed wetlands for the Fall
of 2002. Asterisk (*) indicates dominant species based on the 50/
20 rule for %RIV (see methods).

Species/Common name                      RC     RD     RF    %RIV

Reference Wetland
Aster subulatus Michx. var. ligulatus    9.2    1.7    4.3    5.1
Shinners (Wireweed)
Cardiospermum halicacabum L. (Common    17.4    2.6   34.8   18.3
balloonvine)
Ecliptaprostrata (L.) L. (Pieplant)      1.9    0.7    4.3    2.3
Eupatorium serotinum Michx. (Fall        3.7    1.2    4.3    3.1
boneset)
Iva annua L. (Marsh-elder) *            25.2   17.7   13.0   18.7
Panicum dichotomiflorum Michx. (Fall     0.9    0.5    4.3    1.9
panicum)
Paspalum distichum L. (Knot grass) *    17.8   69.7    8.7   32.1
Sorghum halepense (L.) Pers. (Johnson    1.9    0.6    4.3    2.3
grass)
Typha latifolia L. (Common cat-tail) *  21.5    3.7   13.0   12.8
Ulmus crassifolia Nutt. (Cedar elm)      0.4    1.0    4.3    1.9
One unknown plant in quadrats            0.2    0.4    4.3    1.6

Constructed Wetland
Amaranthus rudis J. D. Sauer (Water-     9.4   22.4    6.4   12.7
hemp) *
Ammonia coccinea Rottb. (Toothcup)       1.1    0.4    2.1    1.2
Aster subulatus Michx. var. ligulatus    6.4    6.4   10.6    7.8
Shinners (Wireweed)
Cardiospermum halicacabum L. (Common     0.1    0.1    2.1    0.8
balloonvine)
Chamaesyce serpens (Kunth) Small (Mat    5.1    4.4    4.3    4.6
euphorbia)
Digitaria cognata (Schult.) Pilg.        0.7    0.2    2.1    1.0
subsp. cognata (Fall witch grass)
Echinochloa muricata (P. Beauv.)         0.6    0.5    2.1    1.1
Fernald (barnyard grass)
Eragrostis reptans (Michx.) Nees.        0.1    0.1    2.1    0.8
(Creeping love grass)
Heliotropium curassavicum L. (Salt       0.1    0.1    2.1    0.8
heliotrope)
Iva annua L. (Marsh-elder) *            15.6    4.8   10.6   10.4
Leptochloa mucronata (Michx.) Kunth      1.5    2.3    4.3    2.7
(Slender grass)
Lippia nodiflora (L.) Michx.             3.7    4.2   14.9    7.6
(Frogfruit)
Panicum obtusum Kunth (Vine-mesquite)    1.1    0.7    2.1    1.3
Paspalum distichum L. (Knot grass) *    11.6   41.9    6.4   20.0
Tridens texanus (Wats.) Nash. (Texas     0.6    0.9    2.1    1.2
tridens)
Typha lalifolia L. (Common cat-tail) *  14.3    1.4   10.6    8.8
Xanthium strumarium L. var. canadense   24.5    2.0   10.6   12.4
(Mill.) Torr. & A. Gray (Cocklebur)
Two unknown plants in quadrats           3.5    7.3    4.3    5.0

Table 5. Longevity (L) as annual (A) or perennial (P), wetland
indicator status (WIS) (see methods) with hydrophytic (OBL,
FACW+, FACW, FACW-, FAC+), neutral (FAC), non/hydrophytic (FAC+,
FACU+, FACU, FACU+, UPL), and not in database (NID) designations
for plants identified to species, as well as relative coverages
(RC), relative densities (RD), relative frequencies (RF), and
percent relative importance values (%RIV) for plants in the
reference and constructed wetlands for the entire study. Asterisk
(*) indicates dominant species based on the 50+20 rule for %RIV
(see methods).

Species                         L    WIS     RC     RD     RF   %RIV

Reference Wetland
Ammonia coccinea                A    OBL    0.1    0.1    1.0    0.4
Aster subulatus var. ligulatus  A    OBL    2.5    0.5    1.0    1.3
Cardiospermum halicacabum       A    FAC    8.2    1.1   14.1    7.8
Cynodon dactylon                P   FACU+   4.0   28.2    5.1   12.4
Cyperus sp.                     --   --     0.3    0.4    1.0    0.5
Eclipta prostrata               A   FACW    0.5    0.2    1.0    0.6
Eupatorium serotinum            P   FAC-    3.5    0.4    3.0    2.3
Heliotropium curassavicum       P   FACW    0.1    0.0    1.0    0.4
Iva annua *                     A   FAC+   27.7   40.0    9.1   25.6
Lippia nodiflora                P   FAC+    0.6    0.1    2.0    0.9
Panicum dichotomiflorum         A   FACW    0.3    0.1    1.0    0.5
Panicum obtusum                 P   FAC+    0.1    0.0    1.0    0.4
Paspalum dilatatum              P    FAC    0.1    0.0    1.0    0.4
Paspalum distichium             P   FACW+   4.9   18.6    2.0    8.5
Pluchea odorata                 A    OBL    0.1    0.0    1.0    0.4
Rumex pulcher                   P   FACW-   0.1    0.1    2.0    0.8
Salix nigra                     P   FACW+   0.8    0.2    2.0    1.0
Sorghum halepense               P   FACU    0.5    0.2   10.1    3.6
Teucrium canadense              P   FACW-   0.8    0.1    2.0    1.0
Typha latifolia *               P    OBL   43.0    6.8   27.3   25.7
Ulmus crassifolia               P    FAC    0.1    0.3    1.0    0.5
Xanthium strumarium var.        A   FAC-    1.2    0.8    5.1    2.4
canadense
Three unknown plants in         --   --     0.7    1.7    6.0    2.8
quadrats

Constructed Wetland
Amaranthus rudis                A    FAC    5.4   12.6    6.7    8.2
Ammonia coccinea                A    OBL    1.5    1.4    4.8    2.6
Aster subulatus var. ligulatus  A    OBL    2.8    3.2    4.8    3.6
Cardiospermum halicacabum       A    FAC    0.0    0.1    1.0    0.4
Carex tetrastachya              P   FACW    0.4    0.7    1.0    0.7
Chamascyce serpens              A    UPL    2.9    2.7    3.8    3.1
Cynodon dactylon                P   FACU+   0.2    0.0    1.0    0.4
Cyperus acuminatus              P    OBL    0.1    0.1    1.0    0.4
Digitaria cognata subsp.        P    NID    0.3    0.1    1.0    0.5
cognata Echinochloa muricata    A   FACW    7.0    0.9    4.8    4.2
Elymus virginicus               P    FAC    0.1    0.4    1.0    0.5
Eragrostis reptans              A    OBL    0.0    0.0    1.0    0.3
Heliotropium curassavicum       P   FACW    0.0    0.0    1.0    0.3
Iva annua *                     A   FAC+   18.4   24.1   10.5   17.6
Leptochloa mucronata            A   FACW    0.6    1.1    1.9    1.2
Lippia nodiflora                P   FAC+    6.9    5.6   12.4    8.3
Panicum obtusum                 P   FAC+    0.5    0.4    1.0    0.6
Paspalum distichium             P   FACW+  11.4   32.6    8.6   17.5
Rumex pulcher                   P   FACW-   2.0    2.0    1.0    1.6
Schedonnardus paniculatus       P    NID    1.5    0.1    1.0    0.8
Sorghum halepense               P   FACU    4.6    3.6    1.9    3.4
Tridens texanus                 P    NID    0.2    0.4    1.0    0.5
Typha latifolia *               P    OBL   14.8    2.7   14.3   10.6
Xanthium strumarium var.        A   FAC-   15.8    1.3    9.5    8.9
canadense *
Three unknown plants in         --   --     2.4    4.0    3.8    3.8
quadrats

Table 6. Richness, evenness, and Shannon's diversity for
reference and constructed wetlands during each sampling
period including combined total.

                                         Shannon's
Sampling period     Richness   Lveness   diversity

Reference Wetland
Fall 2001              10       0.67       1.53
Spring 2002            5        0.38       0.62
Summer 2002            11       0.41       0.99
Fall 2002              11       0.44       1.06
Combined total         25       0.48       1.56

Reference Wetland
Fall 2001              9        0.78       1.70
Spring 2002            8        0.66       1.38
Summer 2002            8        0.53       1.10
Fall 2002              18       0.64       1.86
Combined total         27       0.64       2.11
COPYRIGHT 2012 Texas Academy of Science
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Brister, Jeffrey S.; Nelson, Allan D.
Publication:The Texas Journal of Science
Date:Feb 1, 2012
Words:8322
Previous Article:Semi-strong continuity.
Next Article:DDT dechlorination electrocatalyzed by a synthetic iron porphyrin in pyridine and dimethyl sulfoxide.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters