The vascular plant community was surveyed at Haskamp Woods , Allen County, Indiana, which we compared to six neighboring forests. We encountered 54 unique species across the understory, midstory, and overstory strata. Species richness and di versity at Haskamp Woods were greatest in the understory. Coefficient of conservatism at Haskamp Woods was only greater than one of the neighboring forests, while floristic quality index was the median val ue. Additionally, using nonmetric multidimensional scaling to visualize dissimilarity between the seven forests, Haskamp Woods was more dissimilar to the two properties with the longest protect ion history and the two privately owned properties. Overstory and midstory at Haskamp Woods were dominated by Acer saccharum. The overstory composition was relatively similar to five of the neighborin g forests, with the sixth differing because it is an uncommon forest type in the region. Overall, Haskamp Woods ha similarities and differences in species composition compared to the six neighboring forests. Va riability in protection and disturbance histories, as well as edaphic conditions, result in variable fores t communities. In the Midwest United States, agriculture, including culti vated crops, pastureland, or other open-field agricultural practices, dominates the land cover t ypes (Fry et al. 2011). Across the entire region, agriculture accounts for 60% of land use types. H owever, focusing on the southern half of the Midwest, cultivated and pastoral agriculture accounts fo r 80-90% of land use, with fragmentation of forests further driven by urban and suburban development (Radeloff et al. 2005). The majority of forests in the region are relatively small woodlots, most of which are privately owned (Fuelling 2014). Forest fragments serve as important, suitable habitat f or a wide range of organisms due to the often-unsuitable nature of surrounding matrix (Bouma et al . 2013). Influence by the surrounding matrix is variable depending on fragment isolation, quality of the matrix, and other underlying factors (Davies et al. 2000; Vandermeer & Carvajal 2001). This surro unding matrix is important as seed sources for plant community development through colonization and extinction processes (Fröborg & Eriksson 1997). Additionally, interactions between soil type s and disturbance histories can lead to the development of uncommon forest types, which may differ fr om surrounding forests (Adkins et al. 2016). Haskamp Woods is a 31.8 ha property that was privately owned u til 2002, when it was purchased by the New Haven Parks and Recreation Department (pers. comm., A. Gurney). Over twothirds of the property is currently cultivated row crop agriculture with 10.1 ha forested (pers. observ.). The forest canopy was patchy with large gaps visible in aeri al images from 1938 and 1964, with canopy closure happing at some point after 1972 (IHAPI 2017a-c). Ha skamp Woods is surrounded by privately owned land, which is a mixture of agriculture, f orests, and suburban development. Close Bisht et al.: Survey of Haskamp Woods 2 proximity to agriculture and suburban development is typical f or forest fragments within Indiana and the surrounding region (e.g. Brothers & Spingarn 1992; Fuelling 2014; A rvola et al. 2014; Adkins et al. 2016). Understanding the baseline community composition and structure is important for development of sound management strategies. The objectives of this study were to conduct an ecological plant survey of understory, midstory, and overstory species in Haskamp Woods, Allen County, Indiana, and compare the plant community with neighbori ng forest properties. Results from this study regarding plant community composition, environmental variables, and forest structure will assist New Haven Parks and Recreation in making manageme nt decisions about Haskamp Woods. MATERIALS AND METHODS Study Design Five transects were established at Haskamp perpendicular to the eastern edge of the property, with five plot centers along each transect 30 m apart (Fig ure 1). The eastern most plots were 33 m from the eastern edge of the property. At each plot center , two 1 m understory quadrats were located Figure 1. Study design in the forested portion of Haskamp Woods (locati on marked with star), Indiana, USA. Bisht et al.: Survey of Haskamp Woods 3 randomly within a 5 x 5 m area. A 10 x 10 m midstory plot and 15 x 15 m overstory plot were established at each plot center. Mapping and spatial analys es were conducted in QGIS (version 2.18.3). Field data collection occurred 17-30 September 2016. Plant Surveys Within each 1 m understory quadrat, all plants ≤ 2 m tall were counted and identified to species. Understory data was pooled at each plot center. Voucher specimens of understory plants were deposited in the Indiana University-Purdue University herbarium. Within each 100 m 2 midstory plot, all plants > 2 tall and ≤ 8 cm in diameter at breast height (dbh, 1.3 m above the soil surface) were counted and identified to species. Within each 225 m overstory plot, all plants > 8 cm in dbh were counted, identified to species, and dbh was reco rd d. The largest diameter overstory individual in each plot was selected and two cores were collected at breast height, perpendicular to each other, with a 4.3 mm diameter increment borer (Haglöf Sweden AB , Långsele, Sweden). Cores were air dried, mounted on wood rails for support, progressively sa nded with 220 to 500 grit sandpaper, and rings were counted. Environmental Data We measured percent volumetric soil moisture content, per cent canopy cover, soil pH, litter depth, soil compaction, and percent available light at the corner of each 5 x 5 m, 10 x 10 m, and 15 x 15 m plot. Percent volumetric soil moisture content (VMC) was measured with a 12 cm long probe attached to a FieldScout TDR moisture meter (Spectrum Tec hnologies Inc., Aurora, IL, USA). Percent canopy cover was measured with a concave spherical densiometer (Forestry Suppliers, Jackson, MS, USA). Soil pH was measured with a Fields cout SoilStik meter (Spectrum Technologies Inc., Aurora, IL, USA). Litter depth was measured with a meter stick to the nearest 0.1 cm. Soil compaction was measured with a Lang penetrometer (Forestry S uppliers, Jackson, MS, USA) as insertion force (kgf). Percent available light was calcul ted from a six-sensor bar at each plot corner and an unattended single sensor (Spectrum Technologies Inc., Aurora, IL, USA) set outside of the forest in full sun. Light data was collected as μmol/m/sec of photons and converted to percent available light. Soil types and drainage information we re defined by the USDA NRCS Web Soil Survey (http://websoilsurvey.sc.egov.usda.gov/) as the dominant values at the site. Data Analysis Species richness (S = number of species), Shannon’s diversi ty index (H'=-Σpi ln pi, where pi is the proportion of ith species in plot), and Pielou’s evenness index (J'=H'/l n S) were calculated for understory, midstory, and overstory plants at each plot. Fl oristic Quality Index (FQI) was calculated for the property for understory species (FQI = C mean x square root of the number of species, where Cmean is the mean coefficient of conservatism value for the e ntire property). Coefficient of conservatism values were attained from Rothrock (2004). Sørensen similarity index was calculated between midstory and overstory species (2 * number of sha red species / sum of mistory and overstory richness). Overstory species relative importance values (RIV) were calculated as sum of relative frequency, relative dominance, and relative density. Rela tiv frequency was calculated as frequency of species i / sum of all frequencies, where frequency of species i is the number of plots species i occurred / number of plots surveyed. Relative dominance was calculated as basal area of species i / sum of all basal areas, where basal area was the crosssectional area of each species per ha calculated from dbh data. Relative density was calculated as densi ty of species i / sum of all densities. We used RIV to determine the forest type of Haskamp Woods. Haskamp Woods understory, midstory, and overstory composi tions were compared to those of six neighboring forests in Allen County, Indiana (Figur e 2). Detailed property descriptions can be found in their respective citations but are presented here briefly. Fogwell Forest Nature Preserve is a 12.3 ha state designated nature preserve owned and managed by AC RES Land Trust, which has Bisht et al.: Survey of Haskamp Woods 4 been protected since the 1930s (Arvola et al. 2014). Indiana University-Purdue University Fort Wayne (IPFW) forest is a 13.8 ha forest adjacent to the university, which has been owned by IPFW since 2004 and does not have a formal protection status (Arvola e t a . 2014). Mengerson Nature Preserve is a 14.4 ha state designated nature preserve owned and managed by ACRES Land Trust, which has been protected since 1973 (Arvola et al. 2016). Moser Park is a 5.6 ha forest managed by New Haven Parks and Recreation since 1962, which had previous ly been used for railroad storage and landings (Adkins et al. 2016). The two privately owned prop erties (Private A and Private B) were 10.4 ha and 6.7 ha, respectively (Fuelling 2014). Plot means for each environmental data were calculated. Pearson correlation was used to identify relationships between environmental variables. We us d non-metric multidimensional Figure 2. Geographic relationship between Haskamp Woods and six neighbori ng f ests in Allen County, Indiana. Grey polygon represents City of Fort Wayne boundary. Bisht et al.: Survey of Haskamp Woods 5 scaling (NMDS) ordination to visualize relationships betwe en understory communities in Haskamp and six neighboring forests, and environmental data (percent c anopy cover and overstory richness), displayed as joint vectors. Understory species presence/abse nce values were used for the six properties from Adkins et al. (2016), Arvola e