Printer Friendly

Composition, structure, and trajectories of Great Lakes coastal pine forests in relation to historical baselines and disturbance history.


Historically, native pine species (eastern white pine, Pinus strobus; red pine, P. resinosa; and jack pine, P. banksiana) were an important component of the coastal ecosystem in much of the Great Lakes region (Albert et al., 2008b). These forests provided habitat for many wildlife species, including birds using the Great Lakes as a migration corridor (Albert, 2006). Pine dominance in these shoreline ecosystems was likely maintained by a combination of disturbance regime and environmental conditions (Lichter, 1998; Loope and Anderton, 1998). Some coastal pine forests were situated in recently formed dune systems (either dune and swale or parabolic dunes) where pines became established following sand stabilization (Lichter, 1998). Once these sites were vegetated, short return interval, low intensity fire regimes (partially related to Native American burning) maintained open conditions and inhibited succession to shade-tolerant species (Loope and Anderton, 1998). Other coastal pine forests were situated on more stable outwash plains, with similarly short fire return intervals. Wind disturbance is also likely to be important in coastal habitats, where trees situated on sandy soils, wet sites, or dune features can be especially prone to windthrow (Albert, 2006). Coastal areas also commonly have either excessively drained sandy soils or waterlogged organic soils, such conditions likely promoted consistently open canopy conditions that can favor pine persistence (Fahey and Lorimer, 2014).

The ecosystems of the Great Lakes region have been severely altered by human influences over the past 150 to 200 y (Schulte et al., 2007), and coastal habitats are no exception. Urban and agricultural development has removed ~25% of the area originally occupied by coastal pine forests (Wyse, 2004). Intense timber harvesting occurred across the region during the late 19th century (the 'pine logging era'), with long-lasting legacies in forest composition and structure (Rhemtulla et al., 2009). During this era (roughly 1830-1920), forests were selectively harvested for large pines and subsequent intense slash fires burned large swaths of the landscape (Whitney, 1987). Therefore, although many coastal areas are now protected natural areas, forest composition has generally been altered relative to pre-European colonization, often with reductions in pine dominance (Lichter, 1998; Wyse, 2004). For example, 25% of formerly pine-dominated coastal habitats now support other forest types, including hardwood-dominated and wetland forests (Wyse, 2004). Even where pines remain, in situ regeneration is not occurring in many pine stands (Menges and Armentano, 1984; Carleton et al., 1996). These changes can be partly attributed to natural successional processes, but human alteration of natural disturbance regimes (e.g., fire suppression) and the legacies of the original logging/slash fire disturbance have also likely had an effect (Cowles, 1899; Olson, 1958; Lichter, 1998). Restoration of coastal forest ecosystems has become a priority for managers in the region, in part because of their significant habitat value but also because of their role in hydrologic cycles and stabilization of shorelines (Albert, 2006). A better understanding of the historical development, current status, and possible future trajectories of coastal Great Lakes forests is necessary to guide these efforts.

The legacies of logging-era disturbance could strongly influence the need for pine restoration in coastal ecosystems and affect potential restoration strategies. Site-specific disturbance intensity during the pine logging era has likely had a strong impact on modern pine dominance in remnant or re- established forests (Lichter, 1998). Higher severity disturbance (such as clear-cutting or intense slash fires) may have produced post-disturbance conditions amenable to pine regeneration (Ahlgren, 1976), although elsewhere in the region very high severity or repeated disturbances often lead to dominance by sprouting species {e.g., aspens--Populus spp.; Whitney, 1987; Rhemtulla et al., 2009). Sites that experienced moderate intensity harvesting during the pine logging era (i.e., selective logging of the pine component) may be more likely to be dominated by shade-tolerant species derived from advance regeneration, such as red maple (Acer rubrum), sugar maple (A. saccharum), and American beech (Fagus grandifolia). Finally, sites that escaped pine logging might be expected to have a significant remnant component of pines but could also have a significant mid-story or understory of other species due to fire suppression (Nowacki and Abrams, 2008; Fahey, 2011).

The goal of this study was to relate current conditions in coastal Great Lakes pine forests to historical baselines and to assess the potential for restoration of pines in these ecosystems. Specific objectives were: (1) to compare the composition and structure of coastal pine forest stands using current and historical data; (2) to relate establishment patterns in remnant coastal pine stands to disturbance history, under the hypothesis that stands exhibiting moderate levels of disturbance during the pine logging era would have lower modern pine dominance than those that underwent very low or very high levels of disturbance; and (3) to project future trajectories in stand composition based on comparison of canopy and subcanopy composition.



The general study area for analysis of current and historical coastal pine forests was the shorelines of Lakes Michigan and Superior in Michigan, U.S.A. (Fig. 1). U.S. General Land Office Public Land Survey (PLS) records (collected from 1816 to 1856 for the Michigan study area) were used to evaluate presettlement composition of coastal pine forests (and to aid in selection of study areas). In the PLS surveyors recorded bearing trees at intersections of the township/range/section land grid system and also recorded trees along survey lines. Bearing trees were also recorded where survey lines intersected water features (often termed "meander" corners). For this study, bearing, meander, and line tree data were collected from maps with original survey notes transcribed onto U.S. Geological Survey (USGS) topographic quadrangles prepared by the Michigan Natural Features Inventory (MNFI).

Data used in this study were recorded within presettlement coastal pine forests delineated in a previous publication based on the MNFI maps (Albert et al., 2008b). PLS data were collected for legal and economic rather than ecological purposes and; therefore, have some limitations, such as surveyor biases in tree selection, species misidentification, and data fabrication (Schulte and Mladenoff, 2001). Despite these shortcomings, PLS records have proved useful for reconstructing forest composition, structure, and disturbance regimes prior to significant alteration of the landscape by Euro-American settlers (Bourdo, 1956; Schulte and Mladenoff, 2001; Rhemtulla et al., 2009).

Field study sites were located in areas currently or previously occupied by pine-dominated communities, based on presettlement maps (Albert et al., 2008b), modern vegetation data (Albert et al., 2008a), and aerial imagery in: Pictured Rocks National Lakeshore (Latitude: 46.57, Longitude: -86.35), Wilderness State Park (45.72, -84.94), Ludington State Park (44.02, -86.49), Warren Dunes State Park (41.92, -86.60), Grand Mere State Park (41.99, -86.56), Big Knob State Forest (45.99, -85.65), the Huron Mountain Club (46.88, -87.86), and Hiawatha National Forest (Pointe aux Chenes RNA, 45.92, -84.88; Fig. 1). These sites spanned a spectrum of coastal landforms that support pine forests including parabolic dunes of various ages (Ludington, Warren, Grand Mere, and Big Knob), dune and swale complexes (Wilderness and Pointe aux Chenes), and coastal plains (Huron Mountains and Pictured Rocks). Survey plots (0.05 ha, 25 X 20 m) were established at each site based on randomly located GIS coordinates within delineated stand areas. In each plot all trees ([greater than or equal to] 10 cm dbh) were measured and mapped and saplings (>1 m in height, <10 cm dbh) were tallied in two size classes (0-5 and 5-10 cm dbh). Snags, stumps (<2 m in height, with evidence for harvest origin), and downed woody debris were measured (length or height, diameter, and decay class). Increment cores were collected at 0.8 m height from all canopy trees (trees in dominant, codominant, or intermediate canopy positions) to evaluate the disturbance history and age structure of the canopy. Cores were also collected (at 0.8 m) from a subset (50%) of sub-canopy trees and large saplings (>5 cm dbh) to estimate the range of ages present in the understory.

Stand data collected in this study were supplemented with compositional data from literature sources that matched other modern or presettlement coastal pine forest locations (Albert et al., 2008a, b). These studies included data from: Wyse (2004) who sampled 100 plots (400 [m.sup.2]) across the Good Harbor Plains and Platte Plains areas of Sleeping Bear Dunes National Lakeshore; National Park Service vegetation surveys at Sleeping Bear Dunes (18 of 120 plots; Hop et al., 2011) and Pictured Rocks National Lakeshores (42 of 158 plots; Hop et al., 2010); and a survey of two coastal pine stands in Warren Dunes State Park (Smith and Woodland, 2007).


Increment cores were mounted on grooved wood blocks and sanded with progressively finer sand paper to help distinguish rings. Annual growth increments were measured to 0.001 mm using a Velmex stage micrometer and Metronics Quick-Check 4100. For the principal species in the data set (pines), master ring-width chronologies were developed from 20-25 cores with no periods of suppression. Dating of all other pine core series was evaluated against master chronologies using COFECHA (Holmes, 1983) and checked using visual cross-dating (Yamaguchi, 1991). For cores that did not directly intercept the pith, distance to pith was evaluated with pith locator templates (Applequist, 1958). The number of missed rings was estimated based on this distance and the average growth rate in the innermost 5 cm of the core. For all cores that passed within 5 cm of the pith, the estimated number of missed rings was added to the total ring count to estimate the core height recruitment date and these dates were then tabulated by species for each stand. An estimate of age required to reach core height was not included because such information is not available for these coastal ecosystems, and destructive sampling to determine sapling height growth rates was not possible in these reference sites (Palik and Pregitzer, 1995). Ages of some trees are likely to have been underestimated, but disturbance chronologies (the primary focus of the core analysis - see below) only included recruitments with rapid initial growth. Rapidly growing saplings would likely have taken only a few years to reach 0.8 m in height following canopy removal (Fahey and Lorimer, 2014).

Releases from suppression were determined for each measured core as an additional indicator of canopy removal events. For every date in the ring-width series, the average growth rate was determined for the prior 10 y period and the percent increase in growth for the following 10 y period was calculated (Nowacki and Abrams, 1997). Dates counted as prospective growth releases were those for which this 10 y increase was [greater than or equal to] 100%, with average post-release growth of [greater than or equal to] 1 mm/y, or [greater than or equal to] 150% with post-release growth of [greater than or equal to] 0.75 mm/y. Releases were only counted for trees that were deemed to have been understory trees at the time of release. For each species, we used the diameter threshold at which >95% of trees of that species in present-day stands are in dominant-codominant crown classes to indicate overstory status (Lorimer and Frelich, 1989).

Disturbance chronologies were developed from recruitment and release data for each site, based on an estimate of the proportion of canopy disturbed in each decade (Lorimer and Frelich, 1989). Disturbance chronologies were produced using only recruitment events followed by rapid initial growth (maximum annual growth in the first 10 y of the record at least 1 mm) and release events for trees in the understory at the time of release. Canopy turnover was then estimated based on the proportion of current basal area that had a canopy accession (understory release or recruitment with rapid initial growth) in each decade.


To evaluate the composition and structure of contemporary coastal pine forests relative to historical data (Objective 1), relative basal area (dominance) for all species was calculated for modern stands, literature data, and presettlement data. Both modern and presettlement data were separated into northern and southern geographical subsets based on the location of the climatic tension zone (Curtis, 1959), as indicated by the shift from oak or beech-maple to hemlock-northern hardwood dominance in the presettlement landscape (Albert et al., 2008b): South (Warren, Grand Mere) and North (Ludington, Wilderness, Sleeping Bear Dunes, Pointe aux Chenes, Big Knob, Pictured Rocks, and Huron Mountains). Species dominance data were compared between time periods using two-way contingency table analyses (Agresti, 2007). Relative density in modem stand data was also calculated for all species and by 5 cm diameter classes (0-5, 5-10 cm, etc.) to create diameter distributions. Relative density was calculated only for modern data because many presettlement trees were line trees that lacked distance information.

To assess the relationship between disturbance history and modern pine abundance (Objective 2), the effect that canopy turnover during the pine logging era had on current pine dominance was evaluated. Logging-era canopy turnover was estimated as the proportion of basal area (from cores with a growth ring record reaching back to the logging era, i.e., at least 1919) that had canopy accession (recruitment with rapid initial growth or release leading to eventual canopy position) during the decades designated as the 'pine logging era' (defined as 1830-1889 for southern sites, 1860-1919 for Ludington, 1870-1919 for northern Lake Michigan sites, and 1880-1919 for Lake Superior sites; Karamanski, 1989; NPS, 2010). Pine dominance was then compared among stands with low (<40%), moderate (40-80%), or high (>80%) logging-era canopy turnover using one-way analysis of variance in Sigmaplot v. 12 (2011 Systat Software Inc.). Groups were constructed based on estimated background disturbance rates for forests in the region (~1% canopy turnover per year for periods without major disturbance = ~40% turnover during period of the logging era; Frelich, 2002) to represent unlogged, selectively logged, or heavily logged/ burned conditions.

To investigate potential environmental drivers of establishment patterns, correlation analysis was conducted between decadal recruitment, climatic records (Palmer Drought Severity Index reconstruction--http://iridl.ldeo.columhia.edU/SOURCES/.LDEO/.TRL/. NADA2004/.pdsi-atlas.html--PDSI), and historical lake levels (NOAA Great Lakes Environmental Research Center-- html). Decadal recruitment was regressed against PDSI and lake levels both as raw recruitment and residuals from a fits with uniform, normal, negative exponential, and power functions (Szeicz and Macdonald, 1995).

To illustrate compositional differences between overstory and understory layers and potential future trajectories (Objective 3), an ordination was conducted on combined modern stand overstory and understory composition (Fralish et al., 1993). A 'plots x species' data matrix was constructed in which the overstory ([greater than or equal to] 10 cm dbh) and understory (<10 cm dbh) in each plot were included as separate sample units. Nonmetric multidimensional scaling ordination was conducted on this matrix with PC-ORD v.5 (McCune and Mefford, 2006) using the "Slow and thorough" autopilot setting. To assess the robustness of the ordination, this setting makes a comparison with 250 iterations of randomized data using Monte Carlo analysis. To assess potential successional trajectories, the overstory and understory samples for each plot were connected with successional vectors in the ordination space (McCune and Grace, 2002). Blocked Multi-response Permutation Procedure (MRBP) was used to test whether species composition differed significantly between time periods (blocked by site) and was conducted with PC-ORD using Euclidean distance measure (McCune and Grace, 2002).



Modern stands varied greatly in pine dominance and species composition, especially between northern and southern areas. Pine dominance was >50% in all northern sites except Ludington (southernmost of the northern sites; Tables 1 & 2). National Lakeshore survey plots at Pictured Rocks and Sleeping Bear Dunes also had high pine dominance (61.1 and 50.4% respectively), while one study area from Wyse (2004) had high pine dominance (Platte Plains; 54.5%) and the other more moderate levels (Good Harbor Plains; 24.5%). The southern sites all had much lower pine dominance ([less than or equal to] 30%; Table 1). White pine occurred in all modern sites but was especially abundant in the two southern Upper Peninsula sites (Pointe aux Chenes and Big Knob). Red pine was present in all northern sites and most abundant in the Lake Superior sites (Huron Mountains and Pictured Rocks), while jack pine was only present in two sites that were distant from each other (Ludington and Huron Mountains; Fig. 1). Associated species varied greatly across the northern sites, with red maple, red oak (Quercus rubra), hemlock (Tsuga canadensis), and paper birch (.Betula papyrifera) all representing the most abundant non pine species in at least one site (Table 2). White pine was the only pine species present in the southern sites and these sites were otherwise generally dominated by oaks, with the three major species (red, white--Q. alba, and black--Q. velutina) represented in varying proportions (Table 2).

Stands in both the north and south had negative exponential diameter distributions, with a high-density of small stems composed primarily of subcanopy tree or tall shrub species (Fig. 2). The ordination of overstory and understory composition had a two-dimensional solution that explained 70.3% of the variation in the original data matrix and was highly significant based on Monte Carlo tests (P = 0.012, Stress = 19.4; Fig. 3). The ordination solution illustrated a variety of compositional trajectories in these coastal pine forests but overstory and understory composition did not differ significantly based on MRBP analysis (blocked by site; A = 0.047, P = 0.122). Two stands, Huron Mountains and Ludington, appeared likely to maintain or even increase in pine dominance in the future. In the other northern stands, the understory was dominated by spruces (white--Picea glauca and black--Picea mariana) and balsam fir (Abies balsamea), which exist largely as subcanopy trees in those forests (Fig. 2). The most abundant canopy tree species in the understory of the northern stands were white pine and red maple, with other pine species present but relatively rare. In the southern stands, pines were almost entirely absent from the small size classes and the understory was dominated by sugar maple, basswood (Tilia americana), and sassafras (Sassafras albidum). Noncanopy species were also common in the understory of these stands, especially witch-hazel (Hamamelis virginiana) and serviceberry (Amelanchier spp.).


In the north, presetdement and modern composition (Fig. 4a) had a few major distinctions resulting in a significantly different compositions (presettlement vs. modern stands: [X.sup.2] = 21.19, df = 9, P = 0.012; presettlement vs. modern/literature data: [X.sup.2] = 18.21, df = 9, P = 0.033). The primary differences were more oak and maple and less hemlock, cedar (Thuja occidentalis), and beech in modern stands relative to presettlement stands. Pine species had generally slightly increased or had stable dominance since the presettlement era.

In the south modern composition was highly differentiated from presettlement conditions (presettlement vs. modern stands - [X.sup.2] = 42.88, df = 9, P < 0.001; presettlement vs. modern/literature data - [X.sup.2] = 42.98, df = 9, P < 0.001; Fig. 4b). Much of the difference was associated with a decline in pine dominance from presettlement to the modern era. However, there had also been a large decrease in beech and an increase in maple, basswood, and ironwood (Ostrya virginiana). A shift in oak species composition has also occurred, with white oak replaced by red oak and black oak in the modern landscape.


All sites appeared to have multiple age classes of pines originating across the decades and spanning the core record (Fig. 5). The northern stands were all broadly uneven-aged and all had complex pine age structures, with multiple large pine cohorts. Even the southern stands had pines of various ages, although Warren had no pine established since the 1880s. All of the northern sites, except Big Knob, had some pines that had established since 1950, especially at the Ludington and Pictured Rocks sites. The oldest trees in all of the northern stands were pines and there were very few old trees of other species, while in the southern stands, other species predated the oldest pines.

The majority of sites also exhibited a complex disturbance chronology with low-severity disturbances occurring in most decades in the record (Fig. 5). Most stand basal area was recruited/released between 1900 and 1950, which followed the most intense pine logging but coincided with hardwood logging. Accession of pines in northern stands happened during decades with a range of disturbance intensities not just those with high canopy turnover. Some of the northern stands (Pointe aux Chenes, Wilderness, Ludington, and especially Big Knob) underwent many pine release events. Recent nonpine accessions dominated the chronologies in the southern stands. Although there were multiple older age classes of pines, only one pine was released in the southern stands (Warren). Establishment patterns were not strongly correlated with either PDSI (uniform r = 0.05, normal r = 0.14, negative exponential r = 0.41, power r = 0.03) or lake levels (uniform r = 0.14, normal r = 0.22, negative exponential r = 0.20, power r = 0.14).

Pine dominance was generally lower with moderate levels (40-80%) of pine-logging-era canopy turnover (mean = 46.2% of total basal area), while pines were highly dominant in plots with both high (>80%) and low (<40%) levels of turnover (means = 76.0%, and 80.3%, respectively; Fig. 6). Pine dominance was marginally lower in the moderately disturbed group than in either the high or low disturbance groups (F2i]4 = 3.14, P = 0.07, Fig. 6), although the sample size of stands was low. Individual comparisons of low and high vs. moderate canopy turnover were both also marginally significant (P = 0.052, and 0.054, respectively).



Establishment and disturbance patterns in the remnant coastal pine forests sampled in this study were similar in many respects to those of other pine forests in the region on marginal habitats (Fahey and Lorimer, 2014). Although many Great Lakes region pine forests are largely even-aged and associated with past severe stand-replacing disturbances (Frelich, 2002), others have a more complex age-structure with multiple cohorts of pines (Fraver and Palik, 2011; Fahey and Lorimer, 2014). Several of the coastal pine forests had multiple age classes of pine and many trees appeared to have established after relatively low-severity disturbances (i.e., released from suppression by nonstand replacing disturbances; Fahey and Lorimer, 2014). It is possible that some of the pine recruitment in these forests was related to more intense disturbances than is apparent from the data (because of removal of evidence by subsequent logging). However, four of eight sites had multiple age classes of pines that predated the logging-era; these establishment patterns are likely related to either frequent low-severity fire or wind disturbance (or their interaction) that established pine regeneration and occasionally created openings in the canopy (Loope and Anderton, 1998; Frelich, 2002). Some pine establishments were probably related to selective logging and the response of pine regeneration to post-logging stand environments (Orwig and Abrams, 1999). For example there were many pine recruitment/release events during the 1920s to 1950s (e.g., Big Knob, Pictured Rocks, Ludington; Fig. 5), which likely coincided with hardwood logging in these areas (Karamanski, 1989; NPS, 2010). Nonetheless, the high frequency of pine releases in the core data indicated the ability of pines to survive in suppressed positions and be released into the canopy by moderate disturbances (Fraver and Palik, 2011). Further, the canopy in these coastal habitats is likely open enough to allow for advance regeneration, especially of mid-tolerant white pines (Fahey and Lorimer, 2014).

Magnitude of disturbance during the pine-logging-era appeared to have some influence on current pine dominance in these sites. Heavy pine-logging-era disturbance, which included both logging and fires, was often related to relatively high levels of current pine dominance, presumably because these disturbances allowed sufficient growing space, light, and germination sites for the pines to regenerate (Ahlgren, 1976). As hypothesized, moderate levels of disturbance during the pine logging era were negatively associated with modern pine dominance. This pattern likely reflects the removal of pines from these areas by selective logging, which both removed the seed sources and likely led to disturbance-mediated succession (Abrams and Scott, 1989). The few stands that had little canopy turnover during the pine-logging-era exhibited very high pine dominance, which could reflect continued dominance by old, remnant pines not removed by logging (e.g., Wilderness) or more recent establishment from seed. A potential limitation of this analysis is that recently established trees could have replaced trees established either before or during the logging-era, and the analysis cannot account for potential differences among plots in the proportion of each. Despite this potential shortcoming, the analysis appears to illustrate the long-term impacts of specific disturbance histories, including moderate severity disturbances, such as selective logging, on composition and structure of forest stands (Orwig and Abrams, 1999).


Coastal pine forest abundance has declined at a regional-scale, matching the overall decline in pine-dominated forests across the Lake States (Rhemtulla et al., 2009). The specific decline of coastal pine forests is partly related to conversion of coastal forests to other land uses (Wyse, 2004). However, other regional and stand-level processes, such as disease, browsing, succession, and logging history (discussed in more depth below), may also have resulted in declines in pine abundance within remaining coastal forests. There appears to be significant variability across the region and among specific habitat features in the degree to which composition has changed from presettlement conditions. For example northern stands appear to have undergone less of a change in pine dominance than southern stands (Fig. 4). Specific habitat features may also have played a role in whether or not pines have remained dominant (Fahey, 2011), but there was not a straightforward pattern. The two stands that had the highest modern pine dominance (Hurons and Pictured Rocks; Table 1) were on flat coastal plains. However, some dune systems also had high pine dominance (e.g., Pointe aux Chenes and Wilderness; Table 1) and Wyse (2004) showed higher pine dominance on dune features than plains. Specific site-scale historical trajectories have also likely affected pine dominance in individual stands (Steen-Adams et al., 2011). The various parks and preserves included in this study underwent logging and preservation at different times and their forests may have had very different transitions as a result. For example the state parks in this study were established in the 1920s and may have been spared some of the hardwood logging era (~1910s-1940s) and the Huron Mountain Club was established in the 1880s and was never extensively logged (Christy, 1929).

Although the stand transitions illustrated here suggested regional patterns and variation in those patterns, issues with comparisons between PLS data and the modern stand data preclude strong conclusions about absolute transitions in species composition or dominance. The most significant issue is the difference in sampling strategy between the time periods (i.e., dispersed vs. directed sampling, plot vs. point-centered quarter). For example PLS coastal pine areas were not necessarily entirely dominated by pine forests and include a much greater variety of habitat features and stand types than the small plots used for the modern data (or even the larger stand areas that they sampled, which were chosen based on the presence of pines). There are also issues with the PLS data itself, such as surveyor biases. The surveyors rarely recorded stems less than ~15 cm dbh and also may have been biased against large older trees that were seen as unlikely to persist (Bourdo, 1956). Surveyors were also likely biased for and against certain species due to ease of marking (e.g., paper birch) or perceived longevity. However, practical concerns such as locating the nearest tree easily identifiable as being in a given quadrant (i.e., N-E) would have severely limited species and diameter choice and may have outweighed these biases in many cases (Liu et al., 2011).


Although pines continue to be a dominant component of the canopy on sites selected for this study, especially in the northern part of the study region, dominance in the future may not continue without management intervention. The lack of pine saplings (Table 1) or recent canopy accession (Fig. 5), suggested that current conditions are not favorable to the establishment and recruitment of pines in most locations (Ludington and Pictured Rocks sites and Wyse, 2004 provided exceptions). There are likely multiple interacting causes for lack of pine recruitment, including high levels of browsing damage (Rooney and Waller, 2003; Zenner and Peck, 2009), control of fire and changes in precipitation (and resulting ecosystem "mesophication"; Nowacki and Abrams, 2008), and diseases such as white pine blister rust (Cronartium ribicola, which can be especially damaging in moist conditions in close proximity to the Great Lakes; Van Arsdel et al., 1961).

Successional trajectories in a majority of stands suggested a shift away from pine dominance, even in stands with high current pine canopy dominance (Figs. 2, 3). Stands in the southern part of the region are transitioning toward dominance by mesic hardwood species, such as sugar maple and basswood, much like other formerly fire-maintained forest types in the region (Nowacki and Abrams, 2008). In the north shade-tolerant conifer and hardwood species commonly associated with mesic or wet-mesic sites (such as spruce, balsam fir, and red maple) are prevalent in the understory. These patterns may illustrate strong successional pressure toward mesophytic canopy tree species in the absence of frequent fires (Olson, 1958). High stem densities of understory trees (such as ironwood in the north and sassafras, serviceberry, and witch-hazel in the south) will also likely have a negative effect on regeneration of pines (Smidt and Puettmann, 1998). There also may be a shift in dominance patterns within the pine species, as mid-tolerant white pine is well-represented in the understory of the northern stands (although not as much in the south; Fig. 2), while the intolerant red and jack pines are much less common (see also Wyse, 2004).

Restoration of historical disturbance regimes may be necessary to maintain or reintroduce some degree of pine canopy dominance (and attendant habitat value) in these coastal forests. Management strategies would need to create more open canopy and understory conditions as well as promoting groundlayer conditions amenable to pine establishment (e.g., bare soil for red pine; Ahlgren, 1976) and could include prescribed fire, canopy gap creation, and understory thinning (Zenner and Peck, 2009). In stands that have lost their pine component completely, underplanting of pines into canopy gaps (natural or created) may be necessary (Fahey and Lorimer, 2013). Deer browsing and disease may limit the effectiveness of underplanting, but the strength of these effects likely differs with geographic region and stand conditions (Zenner and Peck, 2009). If they are effectively restored, coastal pine forests may be well adapted to projected future climatic changes, as increased drought frequency and reduced lake levels (Angel and Kunkel, 2010) may produce conditions amenable to pines.

Acknowledgments.--Funding for this project was provided by the Illinois-Indiana Sea Grant. Access to field research sites was granted by the Michigan Department of Natural Resources, USDA National Forest Service, USDI National Park Service, and Huron Mountain Wildlife Foundation. Field and lab assistance was provided by David Carter. Comments on previous versions of the manuscript were provided by Dennis Albert and Marlin Bowles and two anonymous reviewers.


ABRAMS, M. D. AND M. L. SCOTT. 1989. Disturbance-mediated accelerated succession in two Michigan forest types. Forest Science, 35:42-49.

AGRESTI, A. 2007. An introduction to categorical data analysis. Wiley, New York, NY. 744 p.

AHLGREN, C. E. 1976. Regeneration of red pine and white pine following wildfire and logging. Journal of Forestry, 74:135-140.

ALBERT, D. 2006. Borne of the wind: an introduction to the ecology of Michigan Sand Dunes. University of Michigan Press, Ann Arbor, MI. 63 p.

--, J. COHEN, M. KOST, B. SLAUGHTER, AND H. ENANDER. 2008a. Distribution maps of Michigan's Natural Communities. Report no. 2008-1. Michigan Natural Features Inventory, Lansing, MI. 174 p.

ALBERT, D. A., P. J. COMER, AND H. ENANDER. 2008b. Atlas of early Michigan's Forests, Grasslands, and Wetlands: an interpretation of the 1816-1856 General Land Office Surveys. Michigan State University Press, Lansing, MI. 136 p.

ANGEL, J. R. AND K. E. KUNKEL. 2010. The response of Great Lakes water levels to future climate scenarios with an emphasis on Lake Michigan-Huron .Journal of Great Lakes Research, 36:51-58.

APPLEQUIST, M. B. 1958. A simple pith locator for use with off-center increment cores. Journal of Forestry, 56:141.

BOURDO, E. A. 1956. A review of the General Land Office survey and of its use in quantitative studies of former forests. Ecology, 37:754-768.

CARLETON, T. J., P. F. MAYCOCK, R. ARNUP, AND A. M. GORDON. 1996. In situ regeneration of Pinus strobus and Pinus resinosa in the Great Lakes forest communities of Canada. Journal of Vegetation Science, 7:431-444.

CHRISTY, B. H. 1929. The book of Huron Mountain. Huron Mountain Club, Big Bay, MI. 216 p.

COWLES, H. C. 1899. The ecological relations of the vegetation on the sand dunes of lake Michigan. Botanical Gazette, 27:95-117, 167-202, 281-308, 361-391.

CURTIS, J. T. 1959. The vegetation of Wisconsin: an ordination of plant communities. Univ of Wisconsin Press, Madison, WI, U.S. 657 p.

FAHEY, R. T. 2011. Establishment and persistence of early successional pine species in late-successional landscapes of the Great Lakes region. Ph.D. Dissertation. University of Wisconsin, Madison, WI. 212 p.

--AND C. G. LORIMER. 2013. Restoring a midtolerant pine species as a component of late-successional forests: results of gap-based planting trials. Forest Ecology and Management, 292:139-149.

--AND--. 2014. Persistence of pine species in late-successional forests: evidence from habitat-related variation in stand age structure. Journal of Vegetation Science, 25:584-600.

FRALISH, J. S., S. B. FRANKLIN, P. A. ROBERTSON, S. M. KETTLER, AND F. B. CROOKS. 1993. An ordination of compositionally stable and unstable forest communities at Land Between the Lakes, Kentucky and Tennessee, p. 247-267. In: J. S. Fralish, R. P. McIntosh, and O. L. Loucks (eds.), John T. Curtis: Fifty years of Wisconsin plant ecology, The Wisconsin Academy of Sciences, Arts and Letters, Madison, WI.

FRAVER, S. AND B. J. PALIK. 2011. Stand and cohort structures of old-growth Pinus resinosa-dominated forests of northern Minnesota, USA. Journal of Vegetation Science, 23:249-259.

FRELICH, L. E. 2002. Forest dynamics and disturbance regimes: studies from temperate evergreen-deciduous forests. Cambridge University Press, Cambridge, UK. 280 p.

HOLMES, R. L. 1983. Computer-assisted quality control in tree-ring dating and measurement. Tree-Ring Bulletin, 43:69-78.

HOP, K., S. MENARD, J. DRAKE, S. LUBINSKI, AND J. DIECK. 2010. National Park Service Vegetation Inventory Program: Pictured Rocks National Lakeshore. Report no. NPS/GLKN/NRR 2010-201. National Park Service, Fort Collins, CO.

--, J. DRAKE, S. LUBINSKI, S. MF.NARD, AND J. DIECK. 2011. National Park Service Vegetation Inventory Program. Sleeping Bear Dunes National Lakeshore. Report no. NPS/GLKN/NRR--2011/395. National Park Service, Fort Collins, CO.

KARAMANSKI, T. J. 1989. Deep woods frontier: a history of logging in northern Michigan. Wayne State University Press, Detroit, MI. 308 p.

LICHTER, J. 1998. Primary succession and forest development on coastal Lake Michigan sand dunes. Ecological Monographs, 68:487-510.

LIU, F., D. J. MLADENOFF, N. S. KEULER, AND L. SCHULTE MOORE. 2011. Broadscale variability in tree data of the historical Public Land Survey and its consequences for ecological studies. Ecological Monographs, 81:259-275.

LOOPE, W. L. AND J. B. ANDERTON. 1998. Human vs. lightning ignition of presettlement surface fires in coastal pine forests of the upper Great Lakes. American Midland Naturalist, 140:206-218.

LORIMER, C. G. AND L. E. FRELICH. 1989. A methodology for estimating canopy disturbance frequency and intensity in dense temperate forests. Canadian Journal of Forest Research, 19:651-663.

MCCUNE, B. AND J. B. GRACE. 2002. Analysis of ecological communities. MjM Software Design, Gleneden Beach, OR. 300 p.

--AND M.J. MEFFORD. 2006. PC-ORD. Multivariate analysis of ecological data. MjM Software Design, Gleneden Beach, Oregon, USAVersion 5.2.

MENGES, E. AND T. ARMENTANO. 1984. Successional relationships of pine stands at Indiana Dunes. Proceedings of the Indiana Academy of Science, 94:269-287.

NOWACKI, G. J. AND M. D. ABRAMS. 1997. Radial-growth averaging criteria for reconstructing disturbance histories from pre-settlement origin oaks. Ecological Monographs, 67:225-249.

--AND--. 2008. The demise of fire and "mesophication" of forests in the eastern United States. BioScience, 58:123-138.

NPS (NATIONAL PARK SERVICE). 2010. Logging history. Bulletin. Pictured Rocks National Lakeshore, USDI National Park Sendee, Grand Marais, MI. 2 p.

OLSON, J. S. 1958. Rates of succession and soil changes on southern Lake Michigan sand dunes. Botanical Gazette, 125-170.

ORWIG, D. A. AND M. D. ABRAMS. 1999. Impacts of early selective logging on the dendroecology of an old-growth, bottomland hemlock-white pine-northern hardwood forest on the Allegheny Plateau. Journal of the Torrey Botanical Society, 126:234-244.

PALIK, B. J. AND K. S. PREGITZER. 1995. Variability in early height growth rate of forest trees: implications for retrospective studies of stand dynamics. Canadian Journal of Forest Research, 25:767-776.

RHEMTULLA, J. M., D. J. MLADENOFF, AND M. K. CLAYTON. 2009. Legacies of historical land use on regional forest composition and structure in Wisconsin, USA (mid-1800s-1930s-2000s). Ecological Applications, 19:1061-1078.

ROONEY, T. P. AND D. M. WALLER. 2003. Direct and indirect effects of white-tailed deer in forest ecosystems. Forest Ecology and Management, 181:165-176.

SCHULTE, L. AND D. MLADENOFF. 2001. The original U.S. public land survey records: their use and limitations in reconstructing presettlement vegetation. Journal of Forestry, 99:5-10.

SCHULTE, L. A., D. J. MLADENOFF, T. R. CROW, L. C. MERRICK, AND D. T. CLELAND. 2007. Homogenization of northern U.S. Great Lakes forests due to land use. Landscape Ecology, 22:1089-1103.

SMIDT, M. F. AND K. J. PUETTMANN. 1998. Overstory and understory competition affect under-planted eastern white pine. Forest Ecology and Management, 105:137-150.

SMITH, P. F. AND D. W. WOODLAND. 2007. Forest composition study of the Great Lakes coastal forest at Warren Dunes State Park, Berrien County, Michigan. The Michigan Botanist, 46:33-62.

STEEN-ADAMS, M. M., D. J. MLADENOFF, N. E. LANGSTON, F. LIU, AND J. ZHU. 2011. Influence of biophysical factors and differences in Ojibwe reservation versus Euro-American social histories on forest landscape change in northern Wisconsin, USA. Landscape Ecology, 26:1165-1178.

SZEICZ, J. M. AND G. M. MACDONALD. 1995. Recent white spruce dynamics at the subarctic alpine treeline of north-western Canada. Journal of Ecology, 83:873-885.

VAN ARSDEL, E. P., A. J. RIKER, AND T. F. KOUBA. 1961. The Climatic Distribution of Blister Rust on White Pine in Wisconsin. Report no. LS-87. U.S. Department of Agriculture, Forest Service, Lake States Forest Experiment Station, St. Paul, MN. 34 p.

WHITNEY, G. G. 1987. An ecological history of the Great Lakes forest of Michigan. Journal of Ecology, 75:667-684.

WYSE, T. C. 2004. Vegetation and disturbance history of coastal pine forests in northern lower Michigan, USA. MS Thesis. Ohio State University, Columbus, OH. 74 p.

YAMAGUCHI, D. K. 1991. A simple method for cross-dating increment cores from living trees. Canadian Journal of Forest Research, 21:414-416.

ZENNER, E. K. AND J. L. E. PECK. 2009. Maintaining a pine legacy in Itasca State Park. Natural Areas Journal, 29:157-166.



The Morton Arboretum, 4100 Illinois Route 53, Lisle, Illinois 60532-1293

(1) Telephone: (630) 719-2419;

TABLE 1.--Characteristics of sample plots and vegetation in
modern coastal pine forest study areas. Stands listed from most
southern to most northern (i.e., in order of increasing latitude)

                      Area     Basal area                 density #
                     sampled   ([m.sup.2]      Pine         (stems
Stand                 (ha)     [ha.sup.-1)   dominance   [ha.sup.-1])

Warren                 0.1        36.0         22.5          510
Grand Mere             0.1        25.0         14.1          500
Ludington             0.15        34.3         40.3          587
Wilderness             0.4        28.6         64.3          495
Pointe aux Chenes     0.15        37.3         71.8          553
Big Knob              0.05        50.7         65.5          840
Pictured Rocks         0.1        31.6         84.2          690
Huron Mtns.            0.1        28.3         97.7          440
Total                 1.15         32          58.9          543

                       Sapling        Pine        Snag         Stump
                      density #     sapling    Basal area   Basal area%
                        (stems      relative   ([m.sup.2]   ([m.sup.2]
Stand                [ha.sup.-1])   density    [ha.sup.-1)  [ha.sup.-1)

Warren                   1380          0.0        3.8           3.6
Grand Mere               1300          1.5        3.3           0.0
Ludington                 480        100.0        5.1           8.4
Wilderness               1523         24.5        4.0           8.8
Pointe aux Chenes        1100          6.1        1.1           9.4
Big Knob                  420          0.0        3.5           7.1
Pictured Rocks           1360         22.8        3.2          21.2
Huron Mtns.               420        100.0        5.7           3.2
Total                    1142         24.4        3.8           8.1

                        DWD *
Stand                [ha.sup.-1)   Cores

Warren                  51.7         36
Grand Mere               5.1         44
Ludington                9.2         87
Wilderness               8.6        181
Pointe aux Chenes        9.6         51
Big Knob                 6.9         27
Pictured Rocks           3.2         69
Huron Mtns.             25.7         50
Total                   13.2        545

# Overstory includes stems [greater than or equal to]10 cm dbh,
saplings includes stems [greater than or equal to]1 m height and
<10 cm dbh

% Stumps defined as dead stems <2 m in height with evidence for
harvest origin

* DWD--down coarse woody debris >10 cm in diameter at the large

TABLE 2.--Overstory basal area ([m.sup.2] [ha.sup.-1] for stems
[greater than or equal to]10 cm dbh) by species in modern coastal
pine forest study areas. Stands are listed left to right from
most southern to most northern (i.ein order of increasing

Species                                    South

                           Acronym   Warren   Grand Mere

Pinus strobus               PIST       8.1        3.5
Pinus resinosa              PIRE       0.0        0.0
Quercus rubra               QURU      11.6        6.9
Tsuga canadensis            TSCA       8.7        0.0
Pinus banksiana             PIBA       0.0        0.0
Acer rubrum                 ACRU       0.0        0.3
Thuja occidentalis          THOC       0.0        0.0
Quercus velutina            QUVE       0.0        7.1
Betula papyrifera           BEPA       0.0        0.0
Picea glauca                PIGL       0.0        0.0
Acer saccharum              ACSA       3.3        0.0
Picea mariana               PIMA       0.0        0.0
Quercus alba                QUAJL      0.0        2.7
Tilia americana             TIAM       2.7        0.0
Liriodendron tulipifera     LITU       0.0        2.4
Sassafras albidum           SAAL       0.3        1.8
Total                                 36.0       25.0

Species                                  North
                           Ludington   Wilderness   aux Chenes

Pinus strobus                 6.9         8.1         26.1
Pinus resinosa                1.5        10.3          0.6
Quercus rubra                16.2         5.4          0.2
Tsuga canadensis              1.2         0.0          2.5
Pinus banksiana               5.4         0.0          0.0
Acer rubrum                   2.9         1.2          2.2
Thuja occidentalis            0.0         1.8          2.4
Quercus velutina              0.0         0.0          0.0
Betula papyrifera             0.0         0.3          1.3
Picea glauca                  0.0         0.8          0.3
Acer saccharum                0.0         0.0          0.0
Picea mariana                 0.0         0.2          1.3
Quercus alba                  0.0         0.0          0.0
Tilia americana               0.0         0.0          0.0
Liriodendron tulipifera       0.0         0.0          0.0
Sassafras albidum             0.0         0.0          0.0
Total                        34.3        28.6         37.3

Species                                   North

                           Big Knob   Pictured Rocks   Huron Mtns.

Pinus strobus               26.1          13.9            2.0
Pinus resinosa               7.0          12.7           15.8
Quercus rubra                0.0           0.0            0.0
Tsuga canadensis            12.0           0.0            0.7
Pinus banksiana              0.0           0.0            9.9
Acer rubrum                  3.4           1.4            0.0
Thuja occidentalis           0.8           0.0            0.0
Quercus velutina             0.0           0.0            0.0
Betula papyrifera            1.3           3.0            0.0
Picea glauca                 0.0           0.6            0.0
Acer saccharum               0.0           0.0            0.0
Picea mariana                0.0           0.0            0.0
Quercus alba                 0.0           0.0            0.0
Tilia americana              0.0           0.0            0.0
Liriodendron tulipifera      0.0           0.0            0.0
Sassafras albidum            0.0           0.0            0.0
Total                       50.7          31.6           28.3
COPYRIGHT 2014 University of Notre Dame, Department of Biological Sciences
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2014 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Fahey, Robert T.
Publication:The American Midland Naturalist
Article Type:Report
Geographic Code:1U3MI
Date:Oct 1, 2014
Previous Article:Growth, biomass, and allometry of resprouting shrubs after fire in scrubby flatwoods.
Next Article:Effects of population size, forest fragmentation, and urbanization on seed production and gene flow in an endangered maple (Acer miyabei).

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |