RHODORA JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB VoK 96 January 1994 No. 885 RHODORA, Vol. 96, No. 885, 1-29, 1994 pp. A AND VEGETATION FLORISTIC ANALYSIS OF A CREATED WETLAND SOUTHEASTERN IN NEW HAMPSHIRE Donald Padgett and Garrett Crow E. J. ABSTRACT The common artificial creation of wetlands has become a means to mitigate the effects of wetland alteration due to development. The establishment of diverse, functional plant communities and is a crucial challenging part of a wetland creation project. This study analyzes plant community structure of a seven year old, ar- One tificially created wetland. hundred and four vascular plant species are doc- umented from New the study site including two species listed as rare for Hampshire and the first state record for an introduced weedy species. TWINSPAN The computer program used species abundance data to classify the vegetation into seven cover types. The three plant associations of the open water Chara habitat were: cover type, Potamogeton pusillus cover type, and the Potamogeton The natans cover type. four plant associations of the emergent habitat were: Eleocharis smallu cover Typha type, cover Juncus latifolia type, effusus-Phalahs arundinacea cover and Carex type, a cover Each stricta type. cover type and associated habitat was described and delineated. Key Words: wetlands, wetland floristics, mitigation, rare plants INTRODUCTION Encompa ms marshes as freshwater swamps, bogs, sloughs, ponds and lakes, rivers and streams marshes biomes Mitsch and productivity associated food-chain support, wetland ecosys- tems oerform are vital in (Goodwin and Burke Niering, 1974; 1988; Larson, et 1988). al., The rather late recognition of wetland values and functions over rm 1 Rhodora 96 2 [Vol. Most which wetlands are being destroyed. important, the rale at environment vital role wetlands play in the overall quality of the has been recognized and a variety of efforts are being im- finally means plemented provide conservation. to a for their One has been developed conserve wetlands strategy that to is Wetland com- termed wetland mitigation. mitigation reduces or pensates for the negative impacts on aquatic ecosystems imposed by development by creating or restoring wetland areas artificially become (Savage, 1986). Mitigation has an integral part of wetland and Wetland protection conservation mitigation policies. at- tempts to restore to an area those functions of a wetland that are impacted or destroyed. The primary goal of wetland creation or restoration projects aims wetland Wetland typically at restoring lost functions. cre- aim an ation efforts additionally at creating area that at least is the same size as the area lost and the estabhshment of certain wetland vegetation types (Lowry, 1990; Kruczynski, 1990). The establishment of hydrophilic vegetation wet- crucial in is land mitigation projects they are to imitate successfully natural if Wetland some wetland ecosystems. vegetation provides of the important functions and values inherent wetlands (Good in et al., Mitsch and and 1978; Gosselink, 1986) creates suitable habitats The for wildlife (Weller, 1978). establishment of diverse, func- tional wetland plant communities a challenging part of miti- is and Crow, gation projects (Padgett in press). The objective of the present study to provide a detailed is vegetation analysis and total floristic inventory of a seven year old, artificially created, freshwater wetland located in southeastern New Hampshire. History Site During the winter of 1985-86, the Hospital Corporation of America (HCA) created a freshwater wetland under the Section Army compensate 404 guidelines of the U.S. Corps of Engineers to new for the filling of a wetland at the construction site of a regional New The Portsmouth, Hampshire. land area desig- hospital in nated for the creation of the compensatory wetland was an aban- from doned gravel pit located approximately 3 miles the hospital NH (Rockingham the southern end of Portsmouth, Co.) in site The (Michener of the mitigation area 13 ct al, 1986). size (ca. Crow-Wetland and Padgett 1994] 3 was acres) similar to the size of the area lost due to the hospital construction. Wetland construction began and in the of 1985 continued fall through the winter of 1986. Basins were excavated and graded to form a sinuous configuration of open water pools and marsh areas (Michener The wetland and new et al., 1986). soils peat at the im material through and 6-12" the winter eventually spread as a top dressing over the excavated basins. The primary strategy to vegetate the newly created basins was allow to for the natural colonization of wetland Therefore plants. m serve holding urally large quantities of plant seeds and vegetative pro- (Michener primary Groundwater flow westward, discharging into the adjacent is for- swamp summer dammed after construction, beavers the small outflow channel. approximately above the design elevation. Site Description The H.C.A. Portsmouth wetland a created freshwater wetland is Some consisting of primarily herbaceous and aquatic plants. tree shrub areas exist on several upland islands and pool margins. Open water areas form sinuous configurations around these is- maximum The lands. depth of the open water areas approxi- is mately two The meters. created wetland surrounded by upland is forest communities on the northeastern and southern and sides, A community wetland on northwestem a forest the side. dis- woody turbed, open, sandy area with few surrounds plants the wetland on the southeast side. AND MATERIALS METHODS Vegetation Analysis Samphng Vegetation data were collected during July of 1992. was done using a systematic sampling method (Mueller-Dombois Rhodora 4 [Vol. 96 and Ellenberg, 1974). Thirteen transects were established across m A the wetland at 20 meter intervals. single 0.5 x 0.5 quadrat was placed at every five meters of transect for a total of 284 sample and on banks and upland Trees, shrubs herbs occurring the plots. on the islands throughout the wetland were not included in sites the sampling. A was visual estimate of percent cover recorded to estimate The was abundance. primary focus of the vegetation analysis vascular plant species. However, because of their relative abun- dance Chara cf vulgaris and Ricciocarpus natans, two non-vas- cular aquatic plants, were included in the vegetation analysis. Cover was defined as the vertical crown or shoot-area projection (Mueller-Dombois per species in relation to the reference area and EUenberg, 1974). TWINSPAN (Two-Way The data were analyzed using Indi- program. cator Species Analysis), a fortran classification analysis The program of samples and then constructs a classification first based on this classification, constructs a classification of species according to their ecological preferences (Hill, 1979). TWINSPAN groups the quadrat samples according to the flo- similarity of members. Using these groups as a basis, ristic its form There species are clustered to hierarchial dichotomies. are three ordinations involved in determining a dichotomy: a pri- 1) mary made method ordination, by a of reciprocal averaging; 2) determined from a refined ordination, using differential species the primary ordination; and an indicator ordination, using 3) Corresponding highly preferential species (Hill, 1979). to the Braun-Blanquet cover-abundance of 50 and 75 scale 25, 0, 5, percent (Mueller-Dombois and Ellenberg, 1974), abundance data TWINSPAN of "pseudo-species" were classified at cut levels of 4 and 5 respectively. Ultimately, a two-way table gen- 1, 2, 3, is and erated grouping similar quadrat samples across the top similar down species the side (Padgett, 1993). left A map was prepared based on cover types vegetation (Figure 2) of quadrats along the 13 transects to give a visual estimate of the pattern of distribution of the seven cover types. Study Floristic An inventory of the vascular plants of the was undertaken site during the 1992 season. Voucher specimens were collected field Crow— and Wetland Padgett 1994] 5 284 II 143 0|x:ii walci ly[x:s EiTicigcnl types 54 87 134 Cham co\'cr Eleocharis swaUii I}'|x: cover type 68 55 79 \9 roinnwgeioii iuuhjis Pofamogcfr)!! pusilliLS Typha lalifolia cover ly|x; cover type cover type 65 14 Carex JuUCliS slrirln efflLSlLS- Phalaris arimdiuacea co\'cr type cover type TWINSPAN Figure analysis showing the seven plant cover types classified 1. Numerals number at four hierarchical levels. represent the of quadrats included each dichotomy. in Herbarium Hampsh Nomenclature follows Aquatic and Wetland Plants North- of North America (Crow and and Manual eastern Hellquist, in press) and of Vascular Plants of Northeastern United States Adjacent Canada and (Gleason Cronquist, 1991). AND RESULTS DISCUSSION The construction of the H.C.A. Portsmouth wetland has created a unique environment, allowing for a natural colonization and emergence of wetland plant The various micro-habitats species. formed by the gradual slopes of the basins and integration of As raised islands provide a heterogeneous ecosystem. a result, concentric vegetation zones have developed according to the eco- logical affinities of species characteristic of natural wetland sys- The tems. vegetation patterns of the are typical of those site and described for inland freshwater wetlands (Mitsch Gosselink, Hammer, 1986; 1992; Weller, 1978; Curtis, 1959). Generally, for most inland wetlands, sedges {Carex) and rushes (Juncus) occupy Rhodora 6 [Vol. 96 wet-meadow the areas gradually passing into the shallow water areas colonized by Cattails (Typha), Bulrush {Scirpus) and Pick- Weed The open {Ponlederia cordata). deeper, water areas are erel colonized by submerged species {Utricularia) and floating leaved {Potamogeton and Nuphar), species Plant Cover Types TWINSPAN The was wetland vegetation classified using into seven cover types (Figure Data from 284 quadrats were an- 1). alyzed and 67 species were clustered four hierarchical divisions at into the seven cover types which could be visually recognized in the field. The dichotomy grouped the 284 quadrats into two major first groupings, with 141 quadrats representing the open water cover types, and 143 quadrats representing the emergent cover types. At the second level of clustering, 54 quadrats of open water types Chara were distinguished as the cover type, with 87 quadrats remaining to be sub-clustered. At the third level 68 quadrats represented the Potamogeton natans cover type and 19 quadrats Potamogeton defined the pusillus cover type. TWINSPAN grouped the 143 quadrats the emergent cover in type group into four cover types (Figure At the second level 1). the Eleocharis smallii cover type was recognized quadrats). At (9 the third level the Typha cover type was distinguished latifolia (55 quadrats), while the Juneus ejfusus-Phalaris arundinacea (65 and Carex quadrats) stricta (14 quadrats) cover types were not defined the fourth until level. The clustering of the quadrats at the divisional level clearly first correlates with the two primary habitats designed for the site The open water cover occupy where creation. types the areas the Most water too deep be colonized by emergent of to species. is the open water areas were vegetated by floating-leaved, free-float- and submerged However, some remained ing species. areas de- void of vascular plants. For instance, one open pool was heavily colonized by filamentous Although cover were algae. three types recognized as the open water areas, these were not characterized by distinct vegetation zones. Instead these cover types occurred mosaic in a pattern. The meadow emergent cover types represent the marsh or wet regions designed for the wetland. These areas are defined by a Crow— Padgett and Wetland 1994] 7 topographic gradient beginning in the shallow fringe areas of the marsh Consequently, the emergent when compared nal patterns to th emergent cover most types are the mmance emergent mergent cover types than in the open water cover areas OPEN WATER COVER TYPES Potamogeton Cover Type natans The Potamogeton natans cover type the largest plant asso- is ciation of the open water types (Figure defined by the 68 2), TWINSPAN. Of quadrats clustered by the seventeen species that dom- community, characterize the the floating-leaved natans P. is 35% 90% and The inant with cover frequency (Table sub- 1). merged Najas minor and Utricularia gibba are sub-dominants 9% 11% and with covers, respectively, although U. gibba has a The morphology significantly higher percent frequency. delicate of U. gibba does not allow for high percent coverage. The cover type easily discernible by the presence of broad is floating leaves of Potamogeton natans that form an expansive cover on the surface of the water. This floating-leaf layer was so extensive in certain portions of the wetland, particularly in the northwestern few open water were pool, that surface areas present. mean Although the cover of P. natans was very extensive, the percent cover of P. natans throughout the entire cover type was only 35%. In areas that were sparsely populated with natans P. submerged species, such as Najas minor, Utricularia gibba and were P. pusillus, frequent. Chara Cover Type The Chara cover type one of the larger plant associations of is open The community the water types (Figure can be charac- 2). m minates Rhodora 96 8 [Vol. Chum CT sp. CT Potaniogeton nutans CT Polumogeton pusillus CT Eleocharis smullii CT Typha latifoUa JuncKs ejfitsus-Phalaris CT anindinucea CT Carcx strictu The Figure distribution of the seven plant cover types of the H.C.A. Ports- 2. CT = mouth Cover created wetland. type. Crow-Wetland and Padgett 9 1994] Mean Table percent cover and percent frequency for species in the Pota- 1, < mogeton natans cover type (species with cover percent are excluded). 1 % % Species Cover Frequency Potamogeton natans 35 90 Najas minor 11 15 Utricularia gibba 9 82 Potamogeton pusillus 2 15 Eleocharis acicularis 2 6 Lemna minor 19 1 Ludwigia palustris 1 1 commonly oTChara were observed as monotypic colonies or with few other species. However, Utricularia gibba and Potamogeton natans are sub-dominant components of the community. The community remaining species of the (Table with the exception 2), more of of emergent P. pusillus, are characteristic areas. The weedy Chara nature of has allowed to successfully col- it More open onize the water areas of the typically associated site. with hard water Chara has been found dominate deeper to sites, amounts waters of lakes producing large of biomass (Rickett, The Chara community 1921). exists in the deeper pools of the where wetland, the cover of floating leaved species relatively is low. Potamogeton Cover Type pusillus This cover type characterized by the submerged Pondweed is Potamogeton species, pusillus (Table This plant association 3). Table Mean percent cover and percent frequency for species in the Chara 2. < cover type (species with cover percent are excluded). 1 % % Cover Frequency Species Chara 98 vulgaris 81 cf. gibba 56 Utricularia 11 Potamogeton natans 37 9 Eleocharis acicularis 3 7 Ludwigia palustris 9 1 Potamogeton pusillus 7 1 Juncus 4 articulatus 1 Typha 4 latifolia 1 Rhodora 10 [Vol.96 Mean Table percent cover and percent frequency for species in the Pota- 3. < mogeton pusillus cover type (species with cover percent are excluded). 1 % % Cover Frequency Species Potamogeton 36 95 pusillus Pontederia cordata 12 21 Elcocharis aciculans 26 5 Potamogeton nutans 4 47 42 Utricularia gibba 3 Lemna minor 37 1 Najas minor 11 1 Potamogeton amplifolius 5 1 covers the smallest portion of the open water areas and was rep- resented by only 19 quadrats. The Potamogeton pusillus cover type occurs in scattered areas within the open water portions of most the wetland, typically close to the emergent vegetation cover Two types. of the largest areas where this cover type predom- is shown inate are in Figure Although, this cover type primarily 2. is composed submerged and of emergent free-floating species, the Pontederia cordata sub-dominant component. a is EMERGENT COVER TYPES Juncus ejfusus-Phalaris arundinacea Cover Type The Juncus arundinacea cover most effusus-Phalaris type the is Of floristically diverse cover type of the entire wetland. the 55 species occurring in the cover type, Juncus ejfusus and Phalaris arundinacea are co-dominant, with a collective cover value of 41% (Table This cover type generally extends from the upland 4). shores or islands to the edge communities of the Typha latifolia cover type (Figure However, some pool margins are directly 2). surrounded by cover where Typha type does not dominate. this The areas of the Juncus arundinacea cover effusus-Phalaris type are rarely inundated for any extended period of time. Most por- tions remain wet damp, allowing wide range of wetland to for a species to exist. This mostly low-growing, herbaceous plant com- munity some woody components has such Alnus incana as ssp. rugosa, Acer rubrum and Cornus occupying various spp. portions. among Unique cover both annuals and types, perennials are well The represented within Cyperaceae this association. best repre- is