Changes in Species Composition and Growth in a Restored Maple-Basswood Forest after 30 Years

As the benefits of forests for environmental and human health are increasingly recognized, it becomes more important to better understand how to restore and manage them. Because forests provide key ecosystem services such carbon storage, soil erosion control, water conservation, and wood production, many forest restoration projects are being planned as part of the UN Decade on Ecosystem Restoration, 2021–2030 (Hua et al. 2022, Pearce 2022). Because of the long time period required to document forest restoration patterns, few studies have examined changes in growth and mortality of trees over decades and if species present early in succession persist to become part of a mature forest. Goals of most restoration projects are to produce self-sustaining ecosystems (Hobbs et al. 2007) that restore biodiversity and ecosystem services. Studies of forest restoration over time contribute to our knowledge of restoration and ecological processes as well as provide the ecological understanding to set appropriate goals for specific restoration projects.

We report on the results of a 30-year-old deciduous forest restoration in southeastern Minnesota, USA on former agricultural fields. The goals of the project were to return the land to the Big Woods (maple-basswood) forest type currently found in this area (Daubenmire 1936, Bray 1956) and to better understand successional patterns in these forests. Although Acer saccharum (sugar maple) and Tilia americana (basswood) are the dominant species in studies of the Big Woods, data on bearing trees (section corner trees) from the United States survey of public lands in the mid-1800’s showed Ulmus americana (American elm) and Ulmus rubra (slippery elm) were the most common species, followed by A. saccharum and T. americana, Quercus (oak) spp., Fraxinus (ash) spp. and others (Grimm 1984). The purpose of the current study was to compare tree growth, mortality, and colonization patterns among tree species after 30 years, noting similarities and changes to the growth patterns after 23 years described by Shea and Helgeson (2018). We used these results to determine the status of the restoration and make recommendations for future projects.

Eleven tree species commonly found in local maple-basswood forests were planted as two-year-old bare root nursery stock in 1990 at a density of approximately 1,000 seedlings per hectare. The original species are listed in Table 1. Data were collected from tagged trees in sixteen 0.1 ha transects in two fields adjacent to the St. Olaf College campus in Northfield, Minnesota. A detailed description of the methods and growth patterns in the restoration are reported in Shea and Helgeson (2018). The height of trees was measured every two to four years from 1990 to 2020, using a clinometer when trees were more than 3 m tall. DBH (diameter at breast height, 1.37 m) was also measured on trees > 1.37 m tall. Starting in 2013 new colonizing trees with a DBH > 2.5 cm were measured for diameter and height. A tree was considered dead if no leaves were growing or if the tree was missing. Canopy cover was estimated with a densiometer in six 25 m2 plots in each field section in 2015 and 2020 according to methods in Kilgore and Dolan (2012).

Comparisons of mean tree height and diameter at breast height (DBH) showed significant differences among species (one-way ANOVA, F_14,1557 = 30.7, p < 0.001 for height, F_{14, 1555} = 36.7, p < 0.001 for DBH) (R Core Team 2021). Quercus rubra (red oak), Fraxinus americana (white ash), T. americana, and Quercus alba (white oak) were the tallest species and also the species with the largest DBH measurements. Mean heights in these species ranged from 12–17 m and mean DBH ranged from 15–25 cm. Smaller potential future canopy trees were A. saccharum, Acer rubrum (red maple), Juglans nigra (black walnut), and Quercus macrocarpa (bur oak). Heights in this cohort ranged from 8–9 m and DBH ranged from of 8–11 cm. Although all tree species increased in diameter and height over time, since 2013 the growth patterns of the five more common species showed relative changes (Figure 1). The growth of Q. macrocarpa and Q. alba slowed compared to the growth of F. americana, A. saccharum, and J. nigra. Quercus macrocarpa went from third in height in 2013 to the shortest of the five species in 2020. Growth patterns of Q. rubra and T. americana were similar to those of F. americana, but were not included on the figure for clarity and because of small sample sizes.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Height growth curves of five common originally planted tree species: Fraxinus americana, Quercus alba, Acer saccharum, Juglans nigra and Quercus macrocarpa from 1990–2020.

We compared tree mortality from 1990–2020 and 2013–2020 to that reported in Shea and Helgeson (2018) from 1990–2013 (Table 1). The long-term percent mortality for all the original trees planted in 1990 increased from 35.79% in 2013 to 41.77% in 2020. However, the annual mortality rate decreased slightly, from 1.91% to 1.79%. Over the 30-year study, species with the highest mortality rates were understory species, Prunus americana (American plum) and Ostrya virginiana (hophornbeam), 95% and 92% respectively, followed by A. saccharum (61.59%) and J. nigra (47.43%). Tilia americana (17.65%) and F. americana (21.22%) had the lowest total mortality rates. Examination of annual mortality rates based on a negative exponential formula (Lorimer et al. 2001) showed the following changes since 2013: the average annual mortality continued to be high and increased for the understory species P. americana (9.50%) and O. virginiana (8.07%). Species with the next highest mortality were A. saccharum, Q. alba, and Q. macrocarpa, with annual mortality rates of 3.14%, 2.13%, and 1.98% respectively. The annual mortality rate since 2013 decreased in A. rubrum (0.44%), T. americana (0.65%), and F. americana (0.79%).

Due to mortality, the mean density of the original trees planted in 1990 decreased from approximately 941 tree/ha to 548 trees/ha. However, the overall tree density in 2020 was 992 trees/ha as a result of colonization, likely from nearby forest fragments and seed production within the study site (Hewitt et al. 2019). The most common colonizing tree species (DBH > 2.5 cm) were F. americana (32.4%), P. americana (9.9%) and U. americana (9.6%) (Table 2). All original species had some colonizing individuals (previously unrecorded individuals with DBH > 2.5 cm). Nearby natural populations surveyed by Daubenmire (1936) had similar densities of trees > 2.5 cm DBH: 852 tree/ha at the Northfield site (8 km from this study site) and 968 trees/ha at the Minnetonka site (64 km north).

A notable change over time was the increase in species richness, from 11 planted species in 1990 to 15 species in 2013 and 19 species in 2020. The new species included Acer ginnala (Amur maple), Acer negundo (boxelder), Carya cordiformis (bitternut hickory), Gleditsia triacanthos (honeylocust), Juglans cinerea (butternut), Populus tremuloides (trembling aspen), U. americana, U. rubra, and Ulmus thomasii (rock elm). Because A. ginnala is considered an invasive species, it was removed from the site. The overall species diversity of the restoration was significantly greater in 2020 than in 1990 when Simpson indices were compared with a two-sample t-test (t_∞ = 2.432, p < 0.05). The Shannon and Simpson diversity indices (Brower et al. 1998) increased from 0.84 (Shannon) and 0.82 (Simpson) in 1990 to 0.99 (Shannon) and 0.84 (Simpson) in 2020.

Growth and mortality patterns showed that Q. rubra, Q. alba, T. americana, and F. americana were the dominant species in both height and diameter after 30 years, as shown in Figure 1. These four species, as well as A. saccharum and A. rubrum, showed a decrease in average annual mortality since 2013 (Table 1). The increase in mortality for other species is likely linked to additional shading as canopy cover increased. Mean canopy cover estimates increased from 90.6% in 2015 to 96.4% in 2020. In addition, some individuals that never thrived finally died after 30 years and some species have shorter life spans. Nearby mature forests sites are dominated by A. saccharum and T. americana, with Quercus spp. and Ulmus spp. (elm) also common before Ophiostoma spp. (Dutch elm disease) killed many of the large elms. (Daubenmire 1936, Hauer et al. 2020).

Through colonization, the total number of trees has remained fairly constant over time. Colonizing trees made up nearly half (45%) of the 2020 tree population of 992 trees/ha, an increase from the 34% reported in Shea and Helgeson (2018). Fraxinus americana was the most common species (33.1%) and the most common colonizer (Tables 1, 2). However, the next four most common species (J. nigra, Q. macrocarpa, Q. alba, A. saccharum in Table 2), represented 64% of the individuals in 1990 and 34% of the individuals in 2020, suggesting more species had intermediate densities. Potential canopy species, A. saccharum, T. americana, and Q. rubra increased to 5.2%, 4.0%, and 4.0% of the individuals, respectively, while Q. alba, Q. macrocarpa, and J. nigra decreased in frequency. Of the species added since 1990, Ulmus spp. became more common, making up 4.7% of the trees in 2020, and shade intolerant species P. tremuloides, P. americana, Prunus serotina (black cherry), and C. cordiformis, filled in forest gaps (Baker 1949, Bray 1956, Smith 2008). The most common colonizers were mainly wind-dispersed species (Fraxinus spp., Ulmus spp.), producing large numbers of seeds that disperse widely, especially in more open areas (Gardescu and Marks 2004, Singleton et al. 2001).

Tree growth, mortality, and colonization showed rapid initial changes among shade intolerant or intermediately shade tolerant species such as Fraxinus spp. and Quercus spp. (Baker 1949). As the canopy closed in parts of the restoration, shade tolerant and future canopy species such as A. saccharum and T. americana had decreased mortality rates and increased growth rates. Our data suggest that it would be better to initially plant species with low shade tolerance and higher growth rates such as Fraxinus spp., Prunus spp., and Quercus spp. Shade tolerant species, such as A. saccharum and T. americana, are likely to grow better after some canopy development.

Overall, the restoration has maintained its original tree density, and increased in species richness from 11 to 19 species. Having nearby sources for colonization was important in maintaining tree density, increasing tree diversity, and enhancing resilience to environmental changes (Timpane-Padgham et al. 2017). Future disturbances such as climate change and loss of Fraxinus spp. due to Agrilus planipennis (emerald ash borer) are likely to shift forest composition (Hufnagel and Garamvolgyi 2014, Herms et al. 2007) and create gaps where species with appropriate plant functional traits can establish alternative stable states (Fukami 2015, Perez-Hernandez and Gavilan 2021). The continued tree growth, low annual mortality and increase in diversity metrics suggest this restoration is on a trajectory to meet the goal of becoming a mature maple-basswood forest, able to adapt to environmental changes.