Northeastern Naturalist Vol. 25, No. 3 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 399 2018 NORTHEASTERN NATURALIST 25(3):399–417 Regional Patterns of Floristic Diversity and Composition in Forests Invaded by Garlic Mustard (Alliaria petiolata) Dustin F. Haines1,*, Jason A. Aylward2, Serita D. Frey3, and Kristina A. Stinson4 Abstract - The impacts of invasive species on native plant communities are often studied on small spatial scales but may vary across regionally heterogeneous landscapes. Comparisons of vegetation across several similar sites with and without an invasive species present can be logistically challenging but highly informative to both scientists and land managers. We examined regional geographic variation in the diversity and composition of 8 replicate northeastern forest-understory plant communities invaded by the non-native species Alliaria petiolata (Garlic Mustard). Despite variation in underlying soil conditions and horizon development, several native species and their associated functional groups were either negatively or positively associated with Garlic Mustard invasion at the regional scale, and soil moisture and pH were higher in invaded plots across all sites. Most tree species were less common at invaded sites, but high tree-seedling abundances at some sites led to regionally higher seedling abundance in the presence of Garlic Mustard. Our study highlights the importance of species-specific responses, as well as site-specific soil conditions, for better understanding potential impacts of invasion. Introduction It is well established that invasion by non-native plants can impact diversity and composition of native plant communities (Callaway and Ridenour 2004, Callaway et al. 2004, Elton 1958, Klironomos 2002, Simberloff and Von Holle 1999), and that a wide range of environmental disturbances can facilitate invasions of non-native plants in general (Dukes and Mooney 1999). In northeastern North American deciduous forests, research has documented correlations between plant invasions and landscape features such as fragmentation and past land-use (Motzkin et al. 1996, 1999). However, effects of individual invasive plants on the local flora are generally documented for just 1 or a few adjacent forest stands, and management of individual species is generally done on an ad-hoc, local basis (e.g., Kueffer et al. 2013). A fundamental knowledge gap remains in understanding interactions between forest flora and invasion by individual species of concern (Murphy and Romanuk 2014). Understanding whether and how specific invasive plants co-vary with floristic and environmental patterns at a regional scale can improve our understanding of the invasion process and bolster effective management (Kueffer et al. 2013). 1Department of Biology, University of Minnesota Duluth, Duluth, MN 55812. 2Harvard Forest, Petersham, MA 01366. 3Department of Natural Resources and the Environment, University of New Hampshire, Durham, NH 03824. 4Department of Environmental Conservation, University of Massachusetts Amherst, Amherst, MA 01003. *Corresponding author - dhaines@d.umn.edu. Manuscript Editor: Robert Bertin Northeastern Naturalist 400 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 Vol. 25, No. 3 The invasive plant Alliaria petiolata (M.Bieb.) Cavara & Grande (Garlic Mustard; Brassicaceae) is widely known to invade intact forest-understory plant communities. A number of mechanisms including escape from natural enemies, phenotypic plasticity, and early phenology have been postulated for its success in North American forest habitats (Rodgers et al. 2008a). Several studies have found that it suppresses native plant growth in temperate deciduous forests of North America via phytochemical disruption of mutualisms with mycorrhizal fungi (Burke 2008, Cantor et al. 2011, Castellano and Gorchov 2012, Hale et al. 2016, Koch et al. 2011, Stinson et al. 2006, Wolfe et al. 2008). Garlic Mustard may also alter the native flora through direct competition and/or asynchronous capture of light, water, and nutrients (Myers and Anderson 2003, Whigham 2004), and through effects on nutrient cycling (Rodgers et al. 2008b), suppression of germination (Prati and Bossdorf 2004), and interactions with insect and mammalian herbivores (Dávalos et al. 2015a, 2015b; Kalisz et al. 2014). However, many local studies showing a direct displacement or inhibition of native plants in the field (Nuzzo 1999, Rodgers et al. 2008a, Stinson et al. 2007, Waller et al. 2016) are confounded by others showing little difference in the community composition of invaded and non-invaded sites (Davis et al. 2014, Nuzzo et al. 2009, Rodgers et al. 2008a, Rooney and Rogers 2011), a decrease in impact over time (Lankau et al. 2009), and co-occurrence with other invasions, soil and canopy disturbance, and herbivory—all of which also impact forest flora (Dávalos et al. 2015a 2015b; Eschtruth and Battles 2009; Kalisz et al. 2014; Knight et al. 2009; Nuzzo, et al. 2009). Studies that consider plant community assemblages of invaded forests in the context of geographic variation, particularly in soil variation at a regional scale, are missing from a vast literature on local ecology of Garlic Mustard. Land managers currently working to control Garlic Mustard at a local-stand scale would also benefit from a better understanding of plant community associations with Garlic Mustard invasion and their geographical variation. We conducted comparative community-level analyses in 8 northern hardwood forests to test whether forest patches invaded by Garlic Mustard are floristically distinct from nearby non-invaded patches in diversity, density, and species composition. Based on prior studies, we predicted that Garlic Mustard presence would be associated with reduced plant diversity and that plant community composition would differ between invaded and non-invaded areas. We also predicted that Garlic Mustard invasion would be associated with an increase in other non-native plants and plants associated with disturbance, and with a decline in species dependent on mycorrhizae. We were interested in whether the floristic composition and diversity of invaded/non-invaded forest habitats is consistent across similar forests within a broader regional landscape, and, if not, which environmental and abiotic variables best predict differences in plant community composition. Field-site Description We surveyed and compared the forest understory vegetation in 8 distinct northern hardwood ecosystems of the northeastern US, where Garlic Mustard is a known Northeastern Naturalist Vol. 25, No. 3 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 401 management concern. Study forests were dominated by Acer saccharum (Sugar Maple), Fraxinus americana (White Ash), Quercus rubra L. (Red Oak), and Pinus strobus L. (White Pine), the presence of which suggest secondary regrowth following agricultural abandonment (Hall et al. 2002, Thompson et al. 2013). The study sites were located in an area from eastern and central Massachusetts to the Berkshire and Taconic Mountains, spanning 251 km and elevations of 40 m to 404 m (Table 1). In addition to choosing a similar canopy composition for each site, we controlled for site history by (a) verifying with landowners/stakeholders that Garlic Mustard has been present for at least 2 decades, (b) excluding sites with clear evidence of past agricultural use (i.e., confirming that plots were situated on soils with a shallow and disorganized Ap horizon indicative of use as unimproved pasture and/or woodlot; Motzkin et al. 1996), and (c) mapping spatial coordinates of the sites to available historical forest-cover maps (Motzkin and Foster 2009) and aerial photography (Army Map Service aerial photography courtesy of the US Geological Survey) to verify approximate forest age (Motzkin and Foster 2009). Table 1. Physical descriptions of the 8 study sites, listed from southwest to northeast. Lat. = latitutde, long. = longitude, moist. = moisture, and elev. = elevation. Mean soil Dominant Land Lat. (°N), Soil order, moist. Elev. Slope, canopy Site stewardship long. (°W) texture (%) (m) aspect species West Point Department 41.3793, Inceptisol, 35.6 343.2 20.5% Sugar Maple (WP) of Defense 74.0192 Clay loam 116.5° –Red Oak Black Rock Black Rock 41.4207, Inceptisol, 33.3 212.7 24.8%, Sugar Maple (BR) Forest 74.0104 Sand clay 321.5° –Red Oak Consortium loam –Green Ash Pittsfield State Massachusetts 42.4868, Spodosol, 35.2 360.5 4.6%, Sugar Maple Forest (PF) State 73.2998 Silt clay 102.2° –Black Cherry loam –Beech Questing Trustees of 42.1211, Spodosol, 34.3 404.4 18.7%, Sugar Maple Forest (QF) Reservations 73.2542 Sand clay 297.0° –White Ash loam McLennan Trustees of 42.2215, Spodosol, 22.8 340.7 28.5%, Sugar Maple Forest (MC) Reservations 73.1732 Clay loam 218.9° River Road Private 42.5365, Inceptisol, 19.1 40 12.2%, Sugar Maple (RR) 72.5691 Clay loam 109.4° –White Ash –Silver Maple Harvard Forest Harvard 42.5294, Inceptisol, 26.9 315.7 17.2%, Sugar Maple (HF) University 72.1904 Clay loam 253.2° –White Ash –Black Cherry Drumlin Farm Mass 42.4094, Entisol, 24.2 74.3 13.3%, White Pine (DF) Audubon 71.3272 Clay loam 80.1° –Sugar Maple Northeastern Naturalist 402 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 Vol. 25, No. 3 Methods Understory-community sampling At each site, we established six 2 m × 2 m plots within patches of forest where Garlic Mustard has had a known presence for at least 2 decades. Garlic Mustard mean densities were 65 ± 12.6 SE plants per m2 across all sites, including first- and second-year plants. We located the plots at random points along a transect spanning the length of the invaded patch with a minimum interval of 5 m. We chose invaded sites with a minimum density of 20 adult Garlic Mustard plants per m2, a cut-off value used in prior work (Stinson et al. 2007). Given that the phytochemical effect of Garlic Mustard is known to decline at distances of ≥20 m from a patch (Wolfe et al. 2008), we selected non-invaded plots within 20–200 m of invaded plots along transects of a corresponding length and with a similar slope and aspect. In July 2013, May 2014, and August 2014, we identified to species and counted all vascular plants less than 1 m in height in each plot, following the species nomenclature in Flora Novae Angliae (Haines 2011). We also obtained estimates of Garlic Mustard seedling and adult densities in each plot. Although non-native taxa were sometimes present in both types of plots, we did not select our plots with regard to the presence of other non-native species, but rather considered their presence as a response variable to be generated from our floristic survey. Environmental data At each site we measured slope, aspect, percent slope, and canopy closure as well as soil texture (% sand, silt, and clay), moisture, and pH during the peak growing season for Garlic Mustard (June). We used a magnetic compass to determine aspect, a clinometer to measure slope, and a spherical densiometer to calculate canopy closure. We averaged 4 canopy-closure measurements at each plot (Strickler 1959). We employed a Theta Probe ML3 soil moisture sensor (Delta T Devices, LTD., Cambridege, UK) to quantify bulk soil moisture at 3 random points per plot and averaged those values to generate a seasonal estimate and then quantified moisture of the organic soil layer using the gravimetric water-content method (Jarrell et al. 1999). We determined the texture of mineral soil using the hydrometer method (Elliott et al. 1999) and measured pH of soil suspensions using a digital pH meter (Robertson et al. 1999). Based on known linkages between soil geological history and vegetation (De- Gasperis and Motzkin 2007, Motzkin et al. 1996, Schimel and Chadwick 2013, Soil Survey Staff 2017), we also mapped the spatial coordinates of our sites on local soil-surveys to identify soil-order classification. Soils at our sites included entisols with little to no horizon development in the relatively urban eastern-most site, inceptisols with moderate horizon development and weathering dominating the central portion of the study area in central Massachusetts and the Taconic Mountains, and spodosols with an accumulation of humus and several horizons dominating the western Berkshire Mountains (Soil Survey Staff 2017). Northeastern Naturalist Vol. 25, No. 3 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 403 Analysis Multivariate analyses. We utilized PC-ORD v. 5.10 (MjM Software, Gleneden Beach, OR) to conduct non-metric multidimensional scaling analysis (NMDS) of the full compositional data-set using averaged species densities for each of the 48 plots and Bray-Curtis dissimilarities. To reduce the effect of rare species, we removed species with an occurrence of ≤2 individuals (McCune and Grace 2002), resulting in a count of 61 species for analysis out of the 112 total observed. We found no major difference in species composition across sample dates, and thus created a single (averaged) compositional vector of community composition from our 3 census dates for each site. We considered Garlic Mustard invasion status as part of the sampling design and thus did not include Garlic Mustard density or occurrence as dependent variables. To reduce variation in species densities and to analyze plant communities in terms of relative abundance, we applied general relativization by species prior to analysis (McCune and Grace 2002). We employed the autopilot mode in PC-ORD for the NMDS analysis (medium thoroughness, 100 runs with real data, 0.000001 stability criterion, 10 iterations to evaluate stability, 500 maximum iterations). We visually inspected plot locations on a 2-dimensional plot of NMDS scores for grouping by site and invasion status, and by environmental variables (soil-texture category, soil order; Table 1) to determine if plant-community composition was ordered by these variables. We assigned each of the 61 species in the NMDS ordination a soil-order association based on occurrence within, or closest distance to, convex hulls around plot groupings by soil order. We calculated the proportions of plants within various disturbance and mycorrhizal categories, as described below, within each soil order. To test for the relative effects of Garlic Mustard presence and soil order on plant community composition, we used multi-response permutation procedures (MRPP) on a Bray-Curtis dissimilarity matrix weighted by groups, which was performed in PC-ORD. Univariate analyses. We used generalized linear mixed models (GLMMs) to test for effects of Garlic Mustard invasion, soil order, and their interaction on (1) density and species diversity (total species richness, Shannon diversity, and Pielou’s evenness) of the total plant community, (2) density and diversity of plant functional groups, (3) proportional density and species richness of plants according to their associations with mycorrhizal fungi, and (4) soil moisture. We also used GLMMs to test for soil-order effects on Garlic Mustard density, and for the effects of Garlic Mustard densities on mycorrhizal plant density in invaded plots only. The plant functional groups were forbs, tree seedlings, shrubs (including woody vines and ground cover), ferns and fern allies, graminoids (including sedges and rushes), and non-natives (all non-native plants). We assigned species’ tolerance to disturbance according to descriptions in Haines (2011); those identified as occurring in disturbed/anthropogenic habitats were designated “disturbance-tolerant”, and those without this identification were designated “disturban ce-intolerant” (see Supplementary Table 1, available online at http://www.eaglehill.us/NENAonline/ suppl-files/n25-3-N1636-Haines-s1, and, for BioOne subscribers, at https://dx.doi. Northeastern Naturalist 404 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 Vol. 25, No. 3 org/10.1656/N1636.s1). We categorized species as mycorrhizal or non-mycorrhizal (hereafter, mycorrhizal status), and further classified mycorrhizal plants by mycorrhizal type (endomycorrhizal or ectomycorrhizal), which we determined from published accounts (see Supplementary Analysis Methods and Supplementary Table 2 for additional information, available online at http://www.eaglehill.us/ NENAonline/suppl-files/n25-3-N1636-Haines-s1 and, for BioOne subscribers, at https://dx.doi.org/10.1656/N1636.s1). Only species for which we could find published information on mycorrhizal associations were included in these analyses. We analyzed proportions of plant density and species richness. Density and diversity were much higher for mycorrhizal than non-mycorrhizal plants, and proportions varied widely among mycorrhizal types; thus, all analyses were performed on each mycorrhizal category separately. In all GLMMs, we set site as the random effect. We conducted these analyses in the GLIMMIX procedure in SAS/STAT® software, Version 14.1 of the SAS System for Windows (© 2002–2012 by SAS Institute Inc., Cary, NC). To test for species-specific differences by Garlic Mustard invasion status, we analyzed species densities using a zero-inflated Poisson (ZIP) model and the GENMOD procedure in SAS/STAT software. We deemed P-values ≤0.05 significant for all analyses and adjusted P-values for multiple comparisons using Bonferroni corrections in the MRPP and Tukey’s HSD in the GLMMs. Results Community composition The 3-dimensional NMDS solution had a stress of 19.7, with 15.9%, 14.7%, and 17.3% of variation explained by axes 1, 2, and 3, respectively. Plant communities in non-invaded and invaded plots were separated in the ordination but varied by soil order in degree of separation (Fig. 1A). Plots on entisol soils were distinct from plots on all other soil orders, and communities on inceptisol and spodosol soils overlapped only slightly (Fig. 1A). Of the included environmental variables, soil pH had the strongest correlation with community composition in the ordination; pH was positively associated with axis 2 and was higher for inceptisols and spodosols compared to entisols (Pearson’s r2 = 0.21 for axis 2). Entisols, inceptisols, and spodosols were associated with progressively decreasing proportions of disturbance-tolerant and non-native plant species (Fig. 1B, C; also see Supplementary Table 3 and Supplementary Fig. 1, available online at http://www.eaglehill.us/ NENAonline/suppl-files/n25-3-N1636-Haines-s1 and, for BioOne subscribers, at https://dx.doi.org/10.1656/N1636.s1), but the proportion of mycorrhizal species did not vary with soil order (0.90, 0.94, and 0.93 mycorrhizal for entisol, inceptisol, and spodosol, respectively). Plant community composition, as determined by MRPP, varied significantly among soil orders, but not with Garlic Mustard invasion status (see Supplementary Table 4, available online at http://www.eaglehill.us/ NENAonline/suppl-files/n25-3-N1636-Haines-s1 and, for BioOne subscribers, at https://dx.doi.org/10.1656/N1636.s1). Based on these findings, we included soil order as a factor in the remainder of the analyses, but we excluded data from the single entisol site, due to a lack of Northeastern Naturalist Vol. 25, No. 3 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 405 replication. However, we provide supplementary figures that include the entisolsite data for qualitative comparison (see Supplementary Figs. 2–6, available online at http://www.eaglehill.us/NENAonline/suppl-files/n25-3-N1636-Haines-s1 and, for BioOne subscribers, at https://dx.doi.org/10.1656/N1636.s1). Total plant density and diversity, and soil characteristics There were no significant effects of Garlic Mustard presence on total plant community density or species diversity (See Supplementary Table 2, available online at http://www.eaglehill.us/NENAonline/suppl-files/n25-3-N1636-Haines-s1), and soil order did not significantly affect Garlic Mustard density (F1,35 = 1.66, P = 0.206). However, bulk soil moisture (F1,34 = 5.36, P = 0.0267) and organic soil pH (F1,34 = 17.37, P = 0.0002) were higher on average in invaded than non-invaded plots (Fig. 2A, B). Garlic Mustard density had a positive correlation with organic soil moisture in invaded plots (F1,13 = 4.92, P = 0.045; Fig. 2 C), but it was not correlated with organic soil pH (F1,13 = 0.82, P = 0.381) Figure 1. (A) Nonmetric multidimensional scaling (NMDS) ordination of plant species densities, separated by soil order; (B) proportions of species by disturbance tolerance; and (C) nativity associated with each soil order in the NMDS. In (A), convex hulls are shown for each soil order, and the arrow represents soil pH correlation with axis 2 (r2 = 0.21; all other environmental variables had r2 < 0.2 for both axes). Garlic Mustard was not included in these analyses. Different symbols in (A) represent plots in the following soils and invasion treatments: circles = entisol, triangles = inceptisol, squares = spodosol, open symbols = non-invaded, and closed symbols = invaded. Northeastern Naturalist 406 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 Vol. 25, No. 3 Density and diversity of functional groups Proportional densities and proportional species richness of functional groups varied by Garlic Mustard invasion status and soil order (Fig. 3). For all sites together, ferns, non-natives, and shrubs tended to have higher total proportional densities in non-invaded plots, but tree seedlings showed the opposite pattern, with higher proportional densities in invaded plots at all sites and overall. Total proportional density of forbs was also higher in non-invaded plots, apparently driven by a relationship with spodosol sites. Regarding species richness, ferns and trees trended toward higher richness in non-invaded than invaded plots, but forb richness was lower in non-invaded than in invaded plots. There was higher Shannon diversity (F1,33 = 8.42, P = 0.007), Pielou’s evenness (F1,31 = 14.2, P = 0.001), and forb richness (F1,33 = 6.05, P = 0.019), in invaded than non-invaded plots, but forb density did not vary significantly by invasion status (F1,33 = 0.03, P = 0.853) (Fig. 4). No other functional group was significantly affected by Garlic Mustard invasion (See Supplementary Table 2, available online at http://www.eaglehill.us/ NENAonline/suppl-files/nX25-3-N1636-Haines-s1 and, for BioOne subscribers, at https://dx.doi.org/10.1656/N1636.s1). Figure 2. Effects of Garlic Mustard presence on (A) soil moisture and (B) organic soil pH, and (C) Garlic Mustard density effects on organic soil moisture. Soil moisture and organic soil pH were significantly higher in invaded than non-invaded plots (P = 0.0267 and 0.0002, respectively). The dashed line in (C) represents the positive, significant effect of Garlic Mustard density on organic soil moisture (P = 0.045; analysis based on a GLMM, which has no r2 equivalent such as that supplied by linear regression analysis). Northeastern Naturalist Vol. 25, No. 3 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 407 Among the 40 most abundant species in our study, some native species were clearly affiliated with Garlic Mustard presence or absence, while others were not (Table 2). The forbs Maianthemum canadense (Canada Mayflower) and Erythronium americanum (Trout Lily) were significantly less abundant in invaded than non-invaded plots. Seedlings of the native trees Acer rubrum L. (Red Maple) and White Ash were also negatively associated with Garlic Mustard. Native species that were positively associated with Garlic Mustard included the forbs Tiarella cordifolia (Foamflower) and Impatiens capensis (Jewelweed), and Sugar Maple and Prunus serotina (Black Cherry) seedlings. Density and diversity of disturbance-tolerant and mycorrhizal dependent species There was a disturbance × soil × Garlic Mustard invasion effect on the density of species with regard to disturbance tolerance (Table 3); invaded inceptisols and Figure 3. Proportions of functional-group density and species richness in non-invaded and invaded plots for all data combined (Total) and by soil order. Plant functional groups occur in descending alphabetic order (ferns, forbs, graminoids, nonnatives, shrubs, trees) in each stacked bar. Garlic Mustard was not included in these analyses. Northeastern Naturalist 408 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 Vol. 25, No. 3 Table 2. The 40 species with highest density in the study by invasion association. Species negatively associated with Garlic Mustard have higher densities in non-invaded (N-inv.) than in invaded (Inv.) plots, positively associated species have the opposite association, and species that do not differ in densities by invasion status are neutral. Species in each category are sorted from highest to lowest maximum density. P-values and density estimates are from zero-inflated Poisson tests of density differences between invaded and non-invaded plots; P-values with an asterisk (*) are significant using the Bonferroni correction for multiple comparisons. Additional columns indicate number of sites and soil orders (I = inceptisol, S = spodosol) each species occurred in, and whether or not the species is considered disturbance tolerant. [Table continued on following page.] Functional Density No. of Soils Disturbance Species group N-Inv. Inv. P sites I S tolerant? Negative Garlic Mustard association Maianthemum canadense Desf. (Canda Mayflower) Forb 20.400 2.300 less than 0.001* 4 X X Yes Erythronium americanum Ker-Gawl. (Trout Lily) Forb 19.500 12.900 less than 0.001* 3 X X No Osmundastrum cinnamomea (L.) C. Presl (Cinnamon Fern) Fern 3.810 1.720 0.002* 1 X No Lysimachia borealis (Raf.) U. Manns & A. Anderb. (Starflower) Forb 2.960 0.608 0.003 2 X No Microstegium vimineum (Trin.) A. Camus (Japanese Stiltgrass) Non-native 2.290 1.020 0.048 1 X Yes Fraxinus americana L. (White Ash) Tree 2.120 0.600 less than 0.001* 7 X X No Euonymus alatus (Thunb.) Sieb. (Burning-Bush) Non-native 1.830 0.370 less than 0.001* 6 X X Yes Rubus hispidus L. (Bristly Blackberry) Shrub 1.330 0.083 0.030 2 X Yes Acer rubrum L. (Red Maple) Tree 1.180 0.097 0.004* 3 X X No Carex appalachica J. Webber & P.W. Ball (Appalachian Sedge) Graminoid 1.120 0.471 0.016 5 X X No Tussilago farfara L. (Coltsfoot) Non-native 0.936 0.028 0.009 3 X X Yes Dryopteris sp. (Wood Fern) Fern 0.875 0.279 0.012 4 X X No Parthenocissus quinquefolia (L.) Planch. (Virginia-Creeper) Shrub 0.726 0.341 0.041 5 X X Yes Positive Garlic Mustard association Tiarella cordifolia L. (Foam-Flower) Forb 16.900 59.900 less than 0.001* 1 X No Acer saccharum Marsh. (Sugar Maple) Tree 7.120 9.530 less than 0.001* 7 X X No Impatiens capensis Meerb. (Jewelweed) Forb 0.028 4.210 less than 0.001* 5 X X Yes Persicaria sp. (smartweed) Forb 0.028 2.750 less than 0.001* 1 X No Prunus serotina Herh. (Black Cherry) Tree 0.689 1.670 less than 0.001* 6 X X Yes Rubus sp. (blackberry) Shrub 0.009 1.380 0.002 3 X X Yes Circaea canadensis (L.) Hill (Broad-Leaved Enchanter’s-Nightshade) Forb 0.341 1.170 0.036 3 X X No Sanguinaria canadensis L. (Blood-Root) Forb 0.028 0.750 0.014 1 X X No Symphyotrichum sp. (American-Aster) Forb 0.028 0.608 0.030 2 X No Northeastern Naturalist Vol. 25, No. 3 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 409 Table 2, continued. Functional Density No. of Soils Disturbance Species group N-Inv. Inv. P sites I S tolerant? Neutral Garlic Mustard association Mitchella repens L. (Partridge-Berry) Shrub 7.250 0.000 1.000 1 X No Arisaema triphyllum (L.) Schott (Jack-In-The-Pulpit) Forb 3.030 2.590 0.156 6 X X No Onoclea sensibilis L. (Sensitive Fern) Fern 1.690 2.330 0.406 3 X X No Carex pensylvanica Lam. (Pennsylvania Sedge) Graminoid 1.580 2.120 0.352 3 X X No Carpinus caroliniana Walt. (American Hornbeam) Tree 1.370 1.640 0.633 1 X Yes Viburnum acerifolium L. (Maple-Leaved Viburnum) Shrub 1.240 0.000 1.000 1 X X No Eurybia divaricata (L.) Nesom (White Wood-Aster) Forb 0.583 1.210 0.091 3 X No Fagus grandifolia Ehrh. (American Beech) Tree 1.070 0.000 1.000 1 X No Viola pubescens Ait. (Yellow Forest Violet) Forb 0.000 1.070 1.000 1 X No Dryopteris carthusiana (Vill.) H.P. Fuchs (Spinulose Wood Fern) Fern 0.891 0.466 0.226 3 X X No Polygonatum biflorum (Walt.) Ell. (King Solomon’s-Seal) Forb 0.409 0.740 0.185 4 X X Yes Celastrus orbiculatus Thunb. (Asian Bittersweet) Non-native 0.734 0.516 0.349 4 X X Yes Lactuca canadensis L. (Tall Lettuce) Forb 0.000 0.705 1.000 1 X Yes Trillium sp. (wakerobin) Forb 0.517 0.687 0.433 4 X X No Vitis riparia Michx. (River Grape) Shrub 0.608 0.083 0.081 2 X X Yes Geum fragarioides (Michx.) Smedmark (Appalachian Barren-Strawberry) Forb 0.000 0.608 1.000 1 X Yes Toxicodendron radicans (L.) Kuntze (Poison-Ivy) Shrub 0.534 0.216 0.073 4 X Yes Maianthemum racemosum (L.) Link (Feathery False Solomon’s-Seal) Forb 0.479 0.440 0.864 4 X X No Northeastern Naturalist 410 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 Vol. 25, No. 3 non-invaded spodosols had lower densities than species that are not disturbance tolerant compared to disturbance-tolerant species in either soil order (Fig. 5). However, there were no correlations between Garlic Mustard densities and disturbancetolerant plant-density (F1,13 = 1.75, P = 0.209) or species richness (F1,13 = 0.89, P = 0.363), or non-disturbance tolerant plant density (F1,13 = 1.15, P = 0.303) or species richness (F1,13 = 0.04, P = 0.849). Total mycorrhizal plant proportional density and species richness did not vary significantly with Garlic Mustard invasion status or soil order (Table 3), nor were there correlations between Garlic Mustard densities and mycorrhizal plant density (F1,13 = 0.52, P = 0.484) or species richness (F1,13 = 0.52, P = 0.484). Discussion Few studies have examined regional variation in associations between Garlic Mustard invasion and forest understory plant communities, with prior research largely restricted to individual or a few adjacent local sites (e.g., Dávalos et al. 2015a, 2015b; Rodgers et al. 2008a; Stinson et al. 2007). Here, we demonstrate few direct effects of invasion on overall plant community patterns but some species- Figure 4. Effects of Garlic Mustard invasion and soil order on diversity, density, evenness, and richness of forbs. Different letters represent significantly different (P < 0.05) findings between invaded and non-invaded plots, and among invasion/soil order combinations; upper-case letters represent invasion effects, and lower-case letters represent soil order effects. Garlic Mustard was not included in these analyses. Northeastern Naturalist Vol. 25, No. 3 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 411 Table 3. Generalized linear mixed models output for effects of invasion on density and species richness of disturbance-tolerant and mycorrhizal plants (not including Garlic Mustard) with respect to soil order and disturbance. Characteristic Effect Density Species richness Disturbance tolerance Invasion F1,71 = 2.06, P = 0.155 F1,71 = 5.33, P = 0.024 Disturbance F1,71 = 77.30, P < 0.001 F1,71 = 30.50, P < 0.001 Soil F1,5 = 0.04, P = 0.855 F1,5 = 6.48, P = 0.052 Invasion × Disturbance F1,71 = 1.48, P = 0.227 F1,71 = 1.77, P = 0.187 Invasion × Soil F1,71 = 2.45, P = 0.122 F1,71 = 0.00, P = 0.962 Disturbance × Soil F1,71 = 1.68, P = 0.199 F1,71 = 12.40, P = 0.001 Invasion × Disturbance × Soil F1,71 = 5.27, P = 0.025 F1,71 = 1.72, P = 0.193 Mycorrhizal Invasion F1,33 = 0.10, P = 0.753 F1,32 = 1.74, P = 0.197 Soil F1,5 = 1.17, P = 0.330 F1,32 = 3.05, P = 0.090 Invasion × Soil F1,33 = 1.03, P = 0.316 F1,32 = 0.15, P = 0.703 Figure 5. Effects of Garlic Mustard presence and soil order on plant density and species richness of disturbance- tolerant and non-tolerant plants. Different letters represent significantly different (P < 0.05) invasion/ disturbance combinations, as determined by Tukey’s HSD tests. Garlic Mustard was not included in these analyses. Northeastern Naturalist 412 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 Vol. 25, No. 3 specific trends in negative or positive associations with Garlic Mustard across several temperate deciduous forests of the northeastern US. We also show that siteto- site variation in soil order exists regarding interactions between Garlic Mustard and native vegetation, and that local variation in soil moisture and pH may also affect densities of both Garlic Mustard and other species in the community. Species-specific responses to Garlic Mustard Our work provides a regional context to prior studies showing reduced abundances of spring ephemerals, forbs, and some tree seedlings with Garlic Mustard invasion at the local level (Nuzzo 1999, Rodgers et al. 2008a, Stinson et al. 2007, Waller et al. 2016). Of the individual taxa that were scarce under Garlic Mustard invasion, Canada Mayflower, Trout Lily, and Lysimachia borealis (Starflower), are known to be mycorrhizal (Brundrett and Kendrick 1990), as are Osmundastrum cinnamomea (Cinnamon Fern) (Berch and Kendrick 1982) and seedlings of Red maple and White Ash. The affiliation of the tree-seedling functional group with Garlic Mustard invasion appears to be driven largely by high abundances of Sugar Maple, and in spite of lower species-level abundances of White Ash and Red Maple within invaded patches in our study. The Sugar Maple seedlings were largely in the 0–2-y age group, and occurred in mesic sites with canopy Sugar Maples; thus, the co-occurrence of Sugar Maple seedlings may be an incidental consequence of local canopy structure or timing of seedling emergence, and of Garlic Mustard and Sugar Maple’s shared affinity for mesic sites (Rodgers et al. 2008a). All 3 tree-seedling species mentioned above are mycorrhizal obligates known to decline in growth in the presence of Garlic Mustard (Stinson et al. 2006), so it is unclear whether mycorrhizal suppression by Garlic Mustard is a major driver of all tree seedling abundances at our sites. Other mechanisms may interact with Garlic Mustard invasion to influence densities of other taxa, including higher soil moisture and pH in the invaded patches in our study. Elsewhere, soil moisture and pH have both been important at the local scale (Anderson and Kelley 1995, Meekins and McCarthy 2001), as have other factors not measured here, such as direct competition for nutrients (Poon and Maherali 2015), interactions with deer, which prefer native species to Garlic Mustard (Kalisz et al. 2014) and other exotic taxa (Boyce 2015). Regionally consistent effects of Garlic Mustard invasion on native plant assemblages Despite variation across soil orders and a general lack of Garlic Mustard effects on many of the measured community variables, 2 lines of evidence suggest that forest patches invaded by Garlic Mustard have different patterns of functionalgroup assemblages compared to uninvaded patches across the region. First, despite higher soil moisture at sites with Garlic Mustard, proportional densities and total densities of key fern-species that tend to prefer moist sites were lower at invaded sites. Lower fern densities could be due to competition or potential mycorrhizal suppression by Garlic Mustard (West et al. 2009), although allelopathy and/or high monospecific densities common among ferns themselves may have suppressive effects on Garlic Mustard itself (Stewart 1975). Second, and somewhat surprising, Northeastern Naturalist Vol. 25, No. 3 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 413 was a regionally consistent dip in non-native functional-group densities at sites with Garlic Mustard, which counters established theories of “invasional meltdown” and invasion “hotspots” which predict co-invasions in areas with recent disturbance and/or high resource availability (Stohlgren et al. 1999) and create feedbacks between disturbance and degradation of the native flora (Simberloff and Von Holle 1999). One explanation for reduced exotics in invaded plots in our study is our intentional sampling of intact forested sites, as opposed to forest edges or fragments where exotics are generally more common (Goldblum and Beatty 1999, Yates et al. 2004). Overall, our community-level data suggest that it is possible to categorize these functional groups as affiliates or non-affiliates with Garlic Mustard invasion at the regional scale. However, regional variability in environmental and vegetation patterns may counter predictions arising from smaller-scale experiments showing local effects of Garlic Mustard on native plants, or from broader geographic studies suggesting co-occurrences of exotics during the invasion process. Management implications Taken together, our results highlight the need to consider species-specific responses, landscape-level heterogeneity in soil conditions, and general plant community associations when evaluating ecological impacts and management priorities related to invasive plants. For certain individual species, a regional pattern of negative association with Garlic Mustard invasion suggests the need for coordinated management. These taxa include the mycorrhizal fungi-dependent herbaceous plants Trout Lily and Canada Mayflower, as well as White Ash and Red Maple seedlings. However, our data generally suggest against the practice of managing for high diversity (Roberts and Gilliam 1995), especially if abundances of native forest understory species are a priority, given higher overall diversity in sites with Garlic Mustard. While Garlic Mustard did not tend to occur with other non-natives at a regional scale, the 3-way interaction between disturbance tolerance, invasion status, and soil order indicates frequent co-occurrences and thus the need for co-management of disturbance-tolerant species with Garlic Mustard invasion, particularly on inceptisols. Finally, given Garlic Mustard’s affiliation with higher soil moisture, mesic sites may pose a greater management problem than drier sites. Ours is one of the first to capture landscape-level variation in vegetation patterns pertaining to Garlic Mustard invasion, and additional studies at this spatial scale could be highly informative for understanding broader geographic patterns of impact and management issues for Garlic Mustard. Acknowledgments The authors thank Billy DeVore and Mark Anthony for assistance with fieldwork; Dave Orwig, Julie Richburg, Bethany Bradley, and Christine Urbanowicz for comments on experimental design and earlier versions of the manuscript; and the following landowners for permission to conduct the research: The Trustees of Reservations, West Point Military Academy, Black Rock Forest, Audubon Society, MA State Forest, and M.A. Swedlund. This work was funded by a US Department of Defense Strategic Environmental Research and Northeastern Naturalist 414 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 Vol. 25, No. 3 Development Program (SERDP) grant (NRC2326) to K.A. Stinson and S.D. Frey. Views, opinions, and/or findings contained in this report are those of the authors and should not be construed as an official Department of Defense position or decision unless designated by other official documentation. Literature Cited Anderson, R.C., and T.M. Kelley. 1995. Growth of Garlic Mustard (Alliaria petiolata) in native soils of different acidity. Transactions of the Illinois State Academy of Science 88:91–96. Berch, S.M., and B. Kendrick. 1982. Vesicular-arbuscular mycorrhizae of southern Ontario ferns and fern-allies. Mycologia 74:769–776. Boyce, R.L. 2015. Recovery of native plant communities in southwest Ohio after Lonicera maackii removal. The Journal of the Torrey Botanical Society 142:193–204. Brundrett, M., and B. Kendrick. 1990. The roots and mycorrhizas of herbaceous woodland plants. I. Quantitative aspects of morphology. New Phytologist 114:457–468. Burke, D.J. 2008. Effects of Alliaria petiolata (Garlic Mustard; Brassicaceae) on mycorrhizal colonization and community structure in 3 herbaceous plants in a mixed deciduous forest. American Journal of Botany 95:1416–1425. Callaway, R.M., and W.M. Ridenour. 2004. Novel weapons: Invasive success and the evolution of increased competitive ability. Frontiers in Ecology and the Environment 2:436–443. Callaway, R.M., G.C. Thelen, A. Rodriguez, and W.E. Holben. 2004. Soil biota and exotic plant invasion. Nature 427:731–733. Cantor, A., A. Hale, J. Aaron, M. Traw, and S. Kalisz. 2011. Low allelochemical concentrations detected in Garlic Mustard-invaded forest soils inhibit fungal growth and AMF spore germination. Biological Invasions 13:3015–3025. Castellano, S.M., and D.L. Gorchov. 2012. Reduced ectomycorrhizae on oak near invasive Garlic Mustard. Northeastern Naturalist 19:1–24. Dávalos, A., V. Nuzzo, and B. Blossey. 2015a. Interactive effects of deer, earthworms, and non-native plants on rare forest-plant recruitment. Biological Conservation 187:173–181. Dávalos, A., V. Nuzzo, and B. Blossey. 2015b. Single and interactive effects of deer and earthworms on non-native plants. Forest Ecology and Management 351:28–35. Davis, M.A., C. MacMillen, M. LeFevre-Levy, C. Dallavalle, N. Kriegel, S. Tyndel, Y. Martinez, M.D. Anderson, and J.J. Dosch. 2014. Population and plant community dynamics involving Garlic Mustard (Alliaria petiolata) in a Minnesota Oak Woodland: A four-year study. The Journal of the Torrey Botanical Society 141:205–216. Dukes, J.S., and H.A. Mooney. 1999. Does global change increase the success of biological invaders? Trends in Ecology and Evolution 14:135–139. Elliott, E.T., J.W. Heil, E.F. Kelly, and H.C. Monger. 1999. Soil structural and other physical properties. Pp. 74–85, In G.P. Robertson, D.C. Coleman, C.S. Bledsoe, and P. Sollins (Eds.). Standard Soil Methods for Long-term Ecological Research. Oxford University Press New York, NY. 462 pp. Elton, C.S. 1958. The Ecology of Invasions by Animals and Plants. Chapman and Hall, New York, NY. 181 pp. Eschtruth, A.K., and J.J. Battles. 2009. Acceleration of exotic plant invasion in a forested ecosystem by a generalist herbivore. Conservation Biology 23:388–399. Goldblum, D., and S.W. Beatty. 1999. Influence of an old field/forest edge on a northeastern United States deciduous forest understory community. The Journal of the Torrey Botanical Society 126:335–343. Northeastern Naturalist Vol. 25, No. 3 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 415 Haines, A. 2011. New England Wild Flower Society’s Flora Novae Angliae: A Manual for the Identification of Native and Naturalized Higher Vascular Plants of New England. Yale University Press, New Haven, CT. 973 pp. Hale, A.N., L. Lapointe, and S. Kalisz. 2016. Invader disruption of belowground plant mutualisms reduces carbon acquisition and alters allocation patterns in a native forest herb. New Phytologist 209:542–549. Hall, B., G. Motzkin, D.R. Foster, M. Syfert, and J. Burk. 2002. Three hundred years of forest and land-use change in Massachusetts, USA. Journal of Biogeography 29:1319–1335. Jarrell, W.M., D.E. Armstrong, D.F. Grigal, E.F. Kelly, H.C. Monger, and D.A. Wedin. 1999. Soil water and temperature status. Pp. 55–73, In G.P. Robertson, D.C. Coleman, C.S. Bledsoe, and P. Sollins (Eds.). Standard Soil Methods for Long-term Ecological Research. Oxford University Press New York, NY. 480 pp. Kalisz, S., R.B. Spigler, and C.C. Horvitz. 2014. In a long-term experimental demography study, excluding ungulates reversed invader’s explosive population growth rate and restored natives. Proceedings of the National Academy of Sciences 111:4501–4506. Klironomos, J.N. 2002. Feedback with soil biota contributes to plant rarity and invasiveness in communities. Nature 6884:67–69. Knight, T.M., J.L. Dunn, L.A. Smith, J. Davis, and S. Kalisz. 2009. Deer facilitate invasive plant success in a Pennsylvania forest understory. Natural Areas Journal 29:110–116. Koch, A., P. Antunes, E. Kathryn Barto, D. Cipollini, D. Mummey, and J. Klironomos. 2011. The effects of arbuscular mycorrhizal (AM) fungal and Garlic Mustard introductions on native AM fungal diversity. Biological Invasions 13:1627–1639. Kueffer, C., P. Pyšek, and D.M. Richardson. 2013. Integrative invasion science: Model systems, multi-site studies, focused meta-analysis, and invasion syndromes. New Phytologist 200:615–633. Lankau, R.A., V. Nuzzo, G. Spyreas, and A.S. Davis. 2009. Evolutionary limits ameliorate the negative impact of an invasive plant. Proceedings of the National Academy of Sciences 106:15,362–15,367. McCune, B., and J.B. Grace. 2002. Analysis of Ecological Communities. MjM Software Design, Gleneden Beach, OR. 300 pp. Meekins, J.F., and B.C. McCarthy. 2001. Effect of environmental variation on the invasive success of a nonindigenous forest herb. Ecological Applications 11:1336–1348. Motzkin, G., and D. Foster. 2009. 1830 map of landcover and cultural features in Massachusetts. Harvard Forest Data Archive: HF122. Available online at http://harvardforest. fas.harvard.edu:8080/exist/apps/datasets/showData.html?id=hf122. Accessed 1 October 2017. Motzkin, G., D. Foster, A. Allen, J. Harrod, and R. Boone. 1996. Controlling site to evaluate history: Vegetation patterns of a New England sand plain. Ecological Monographs 66:345–365. Motzkin, G., P. Wilson, D.R. Foster, and A. Allen. 1999. Vegetation patterns in heterogeneous landscapes: The importance of history and environment. Journal of Vegetation Science 10:903–920. Murphy, G.E.P., and T.N. Romanuk. 2014. A meta-analysis of declines in local species richness from human disturbances. Ecology and Evolution 4:91–103. Myers, C.V., and R.C. Anderson. 2003. Seasonal variation in photosynthetic rates influences success of an invasive plant, Garlic Mustard (Alliaria petiolata). American Midland Naturalist 150:231–245. Northeastern Naturalist 416 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 Vol. 25, No. 3 Nuzzo, V. 1999. Invasion pattern of herb Garlic Mustard (Alliaria petiolata) in high quality forests. Biological Invasions 1:169–179. Nuzzo, V.A., J.C. Maerz, and B. Blossey. 2009. Earthworm invasion as the driving force behind plant invasion and community change in northeastern North American forests. Conservation Biology 23:966–974. Poon, G.T., and H. Maherali. 2015. Competitive interactions between a nonmycorrhizal invasive plant, Alliaria petiolata, and a suite of mycorrhizal grassland, old field, and forest species. Peerj 3 e1090. doi: 10.7717/peerj.1090. Prati, D., and O. Bossdorf. 2004. Allelopathic inhibition of germination by Alliaria petiolata (Brassicaceae). American Journal of Botany 91:285–288. Roberts, M.R., and F.S. Gilliam. 1995. Patterns and mechanisms of plant diversity in forested ecosystems: Implications for forest management. Ecological Applications 5:969–977. Robertson, G.P., P. Sollins, B.G. Ellis, and K. Lajtha. 1999. Exchangeable ions, pH, and cation-exchange capacity. Pp. 106–114, In G.P. Robertson, D.C. Coleman, C.S. Bledsoe, and P. Sollins (Eds.). Standard Soil Methods for Long-term Ecological Research. Oxford University Press, New York, NY. 480 pp. Rodgers, V.L., K.A. Stinson, and A.C. Finzi. 2008a. Ready or not, Garlic Mustard is moving in: Alliaria petiolata as a member of eastern North American forests. Bioscience 58:426–436. Rodgers, V.L., B.E. Wolfe, L.K. Werden, and A.C. Finzi. 2008b. The invasive species Alliaria petiolata (Garlic Mustard) increases soil-nutrient availability in northern hardwood– conifer forests. Oecologia 157:459–471. Rooney, T.P., and D.A. Rogers. 2011. Colonization and effects of Garlic Mustard (Alliaria petiolata), European Buckthorn (Rhamnus cathartica), and Bell’s Honeysuckle (Lonicera × bella) on understory plants after five decades in southern Wisconsin forests. Invasive Plant Science and Management 4:317–325. Simberloff, D., and B. Von Holle. 1999. Positive interactions of nonindigenous species: Invasional meltdown? Biological Invasions 1:21–32. Soil Survey Staff. 2017. US General Soil Map (STATSGO2). Natural Resources Conservation Service, US Department of Agriculture. Available online at https://sdmdataaccess. sc.egov.usda.gov. Accessed 15 February 2017. Stewart, R.E. 1975. Allelopathic potential of Western Bracken. Journal of Chemical Ecology 1:161–169. Stinson, K.A., S.A. Campbell, J.R. Powell, B.E. Wolfe, R.M. Callaway, G.C. Thelen, S.G. Hallett, D. Prati, and J.N. Klironomos. 2006. Invasive plant suppresses the growth of native tree seedlings by disrupting belowground mutualisms. PLoS Biology 4:e140. Stinson, K., S. Kaufman, L. Durbin, and F. Lowenstein. 2007. Impacts of Garlic Mustard invasion on a forest understory community. Northeastern Naturalist 14:73–88. Stohlgren, T.J., D. Binkley, G.W. Chong, M.A. Kalkhan, L.D. Schell, K.A. Bull, Y. Otsuki, G. Newman, M. Bashkin, and Y. Son. 1999. Exotic plant species invade hot spots of native plant diversity. Ecological Monographs 69:25-46. Strickler, G.S. 1959. Use of the densiometer to estimate density of forest canopy on permanent sample plots. USDA Forest Service Research Note No. 180. USDA Forest Service, Pacific Northwest Research Station, Portland, OR. 5 pp. Thompson, J.R., D.N. Carpenter, C.V. Cogbill, and D.R. Foster. 2013. Four centuries of change in northeastern United States forests. Plos One 8:15. Northeastern Naturalist Vol. 25, No. 3 D.F. Haines, J.A. Aylward, S.D. Frey, and K.A. Stinson 2018 417 Waller, D.M., E.L. Mudrak, K.L. Amatangelo, S.M. Klionsky, and D.A. Rogers. 2016. Do associations between native and invasive plants provide signals of invasive impacts? Biological Invasions 18:3465–3480. West, B., J. Brandt, K. Holstein, A. Hill, and M. Hill. 2009. Fern-associated arbuscular mycorrhizal fungi are represented by multiple Glomus spp.: Do environmental factors influence partner identity? Mycorrhiza 19:295–304. Whigham, D.F. 2004. Ecology of woodland herbs in temperate deciduous forests. Annual Review of Ecology, Evolution, and Systematics 35:583–621. Wolfe, B.E., V.L. Rodgers, K.A. Stinson, and A. Pringle. 2008. The invasive plant Alliaria petiolata (Garlic Mustard) inhibits ectomycorrhizal fungi in its introduced range. Journal of Ecology 96:777–783. Yates, E.D., D.F. Levia, and C.L. Williams. 2004. Recruitment of three non-native invasive plants into a fragmented forest in southern Illinois. Forest Ecology and Management 190:119–130.