The effect of species diversity on tree growth varies during forest succession in the boreal forest of central Canada

The effect of species diversity on tree growth varies during forest succession in the boreal forest of central Canada

Forest Ecology and Management 455 (2020) 117641 Contents lists available at ScienceDirect Forest Ecology and Management journal homepage: www.elsevi...

2MB Sizes 0 Downloads 30 Views

Forest Ecology and Management 455 (2020) 117641

Contents lists available at ScienceDirect

Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco

The effect of species diversity on tree growth varies during forest succession in the boreal forest of central Canada Anthony R. Taylora,b, Bilei Gaoa, Han Y.H. Chena,c,

T



a

Faculty of Natural Resources Management, Lakehead University, 955 Oliver Road, Thunder Bay, ON P7B 5E1, Canada Natural Resources Canada, Canadian Forest Service – Atlantic Forestry Centre, 1350 Regent Street, P.O. Box 4000, Fredericton, New Brunswick E3B 5P7, Canada c Key Laboratory for Humid Subtropical Eco-Geographical Processes of the Ministry of Education, School of Geographical Sciences, Fujian Normal University, Fuzhou, China b

A R T I C LE I N FO

A B S T R A C T

Keywords: Boreal forest Complementarity Diversity Facilitation Fire Productivity Succession Tree growth

Although major advances have demonstrated that species diversity has a general positive effect on forest ecosystem productivity, some studies report negligible or even negative effects, highlighting remaining uncertainty in our knowledge of the ecological mechanisms that influence diversity–productivity relationships. In particular, ecological succession is postulated to drive temporal shifts in the strength and direction of diversity–productivity relationships, but few studies have explicitly tested this hypothesis because long-term succession data (from forest initiation to eventual climax) are rare. Using a detailed, replicated chronosequence (space-for-time substitution) study design of 53 natural forest stands (ages 8 to 210 years) in the boreal forests of central Canada, we investigated the relationship between neighbourhood species diversity and tree growth of five dominant boreal tree species, covering entire, long-term secondary successional sequences following stand-replacing wildfire. We found compelling evidence that the strength of the relationship between species diversity and tree growth changes over the course of secondary succession, following a general “hump-shaped” pattern, with mid-succession stages of higher functional diversity exhibiting the strongest growth–diversity relationships. However, tree species exhibited individualistic responses to succession-driven changes in species diversity, with broadleaf species (e.g., Populus tremuloides) generally showing negative responses, whereas conifers (e.g., Pinus banksiana) responded more favorably to higher neighbourhood diversity. Furthermore, our results show the effect of individual tree size on the relationship between species diversity and tree growth to be highly variable, contradicting the hypothesis that larger trees benefit more from complementarity due to size-asymmetric competitive ability. These results contribute to disentangling the mechanisms that link species diversity to forest growth and function, which is important to sustainable forest management planning and for predicting the consequences of global biodiversity loss, especially for the boreal forest, which plays a critical role in controlling global carbon flux and climate.

1. Introduction The past several decades have seen a sharp increase in the number of studies investigating the relationship between species diversity and forest ecosystem productivity (Hooper et al., 2005; Zhang et al., 2012). Although major advances have demonstrated that species diversity generally, positively affects forest ecosystem productivity (Duffy et al., 2017), some studies report negligible or even negative effects (e.g., Edgar & Burk, 2001; Vilà et al., 2003; Laganière et al., 2015). A number of possible explanations for these divergent responses have been postulated, including the potential role of ecological succession in driving



temporal shifts in the strength and direction of diversity–productivity relationships (Paquette & Messier, 2011; Barrufol et al., 2013; Lasky et al., 2014); and recognition that at finer biotic scales, individual species may exhibit differential responses to local, neighbourhood diversity depending on site conditions, distance between trees, and relative tree position within the forest canopy (Chamagne et al., 2017; Fichtner et al., 2017; Williams et al., 2017). Accordingly, closer examination of fine-scale processes from which ecosystem-level responses emerge is critical to disentangling the mechanisms that link species diversity to forest ecosystem productivity for predicting the consequences of global biodiversity loss (Hooper et al., 2012).

Corresponding author at: Faculty of Natural Resources Management, Lakehead University, 955 Oliver Road, Thunder Bay, ON P7B 5E1, Canada. E-mail address: [email protected] (A.R. Taylor).

https://doi.org/10.1016/j.foreco.2019.117641 Received 22 July 2019; Received in revised form 16 September 2019; Accepted 17 September 2019 0378-1127/ Crown Copyright © 2019 Published by Elsevier B.V. All rights reserved.

Forest Ecology and Management 455 (2020) 117641

A.R. Taylor, et al.

light when interacting with smaller ones. As a result, the effect of complementarity on growth may be stronger for trees of relatively larger size. This is indirectly supported by Zhang et al. (2016), who showed that overstorey tree species diversity had a positive effect on upper canopy trees but its influence on smaller, understorey trees was negligible. Other studies report conflicting results, showing stronger complementarity benefits among smaller, rather than larger trees, depending on regional climate (Madrigal-González et al., 2016), or no size-dependent effect at all (Báez & Homeier, 2018). Such differences in results may be because past studies have not included a wide enough range of tree sizes, with most focusing on trees greater than 10 cm in diameter at breast height (DBH). Moreover, directly using DBH to represent tree size neglects the important influence of forest age on tree size, as trees of the same size may constitute different forest strata (e.g., understorey versus overstorey) depending on stage of succession. The circumpolar boreal forest is the largest intact terrestrial biome in the world. It is critical to regulating global carbon flux, but its productivity is considered sensitive to changes in species diversity (Paquette & Messier, 2011; Jucker et al. 2016; Liang et al., 2016). A deeper understanding of the mechanisms that drive boreal productivity is critical to addressing global climate change. In this study, we used a detailed chronosequence (space-for-time substitution) design to investigate the relationship between neighbourhood species diversity and tree growth of five dominant tree species across 53 natural stands in the boreal forests of central Canada, covering a wide range of forest stand ages (from 8 to 210 years old) and species compositions following stand-replacing wildfire. These forests provide a unique and novel opportunity to study diversity–productivity relationships across entire, long-term succession sequences (from forest initiation to climax) as their dynamics are largely driven by wildfire and dominant post-fire succession pathways are well documented (e.g., Carleton & Maycock, 1978; Bergeron and Dubue, 1988; Taylor & Chen, 2011). Furthermore, relative to temperate and tropical forests, the boreal contains fewer tree species, which potentially simplifies disentangling individual species contributions to diversity effects. We hypothesized that (1) the effect of neighbourhood species diversity on tree species growth would change as forests underwent secondary succession and become stronger when competition for resources intensifies and as functional diversity increases when early and late-succession species codominate (Fig. 1); and (2) the effect of species diversity would increase with relative tree size because trees have size-asymmetric competitive ability for resources.

The positive effect of species diversity on forest ecosystem productivity is thought to be primarily driven by complementarity, which encompasses the more specific mechanisms of niche (resource) partitioning and interspecific facilitation (Hooper et al., 2005; Williams et al., 2017; Mina et al., 2018). Therefore, it is not unreasonable to expect that the nature of the relationship between species diversity and individual tree species growth will vary as forests undergo secondary succession because community structure and site conditions are known to vary as forests age (Whittaker, 1970; White, 1979; Fridley, 2001; Forrester, 2014). Following severe disturbance, resources such as light and soil nutrients are often plentiful, and the establishment and abundance of species exhibiting overlapping functional traits is high (e.g., photophilia), reducing the effectiveness of niche partitioning (Pacala & Rees, 1998). However, as communities develop, competition intensifies, and site resources become increasingly limited (Odum, 1969; Huston & Smith, 1987; Glenn-Lewin et al., 1992). This can potentially amplify the effectiveness of complementarity (e.g., facilitation), as positive biotic interactions are expected to increase under harsher, resource-limited conditions (Bertness & Callaway, 1994; Maestre et al., 2009). Furthermore, as communities transition from fast-growing, early succession colonizers to slower-growing, shade-tolerant species, overlapping mixtures of early and late-succession species may promote stronger niche partitioning due to greater diversity of contrasting functional traits (Huston & Smith, 1987; Forrester, 2014; Reich, 2014). However, in the long-term absence of severe disturbance, forests may become dominated by climax communities of late-succession species, exhibiting high functional redundancy, e.g., high abundance of shadetolerant species (White, 1979; Chen & Popadiouk, 2002). Concurrent increases in the availability of site resources, related to senescence of overstorey trees, reduces competition and may diminish the benefits of complementarity (see Fig. 1, conceptual diagram). Despite previous efforts to investigate the relationship between species diversity and tree species growth (e.g., Cavard et al., 2010; Chamagne et al., 2017; Fichtner et al., 2017), few studies have explicitly tested how diversity–productivity relationships vary with secondary forest succession over the long-term because this has been constrained by the availability of succession sequences (e.g., Guo, 2003; Barrufol et al., 2013; Lasky et al., 2014). The relationship between species diversity and tree species growth may also be influenced by individual tree size as this strongly affects the ability of individuals to compete for site resources (Coomes et al., 2011). For example, competition for light among trees is size asymmetric, in that larger trees capture disproportionally greater amounts of

2. Methods 2.1. Study area Our study was conducted in the boreal forest, approximately 150 km north of Thunder Bay, Ontario, Canada (49°40′ N and 89°50′ W, Fig. 2). This area is characterized by warm summers and cold, snowy winters. Mean annual temperature is 1.9 °C and mean annual precipitation is 824.8 mm (582.7 mm as rainfall and 238.2 cm as snowfall) as measured by the closest meteorological station in Cameron Falls, Ontario, Canada (Environment Canada, 2019). Soils in our study area were largely deposited by the Wisconsinan glaciation, which ended approximately 9500 years ago in this region. Stand-replacing wildfire is the most common natural disturbance in the study area, with an average fire-return interval of approximately 100 years during the past century, resulting in a mosaic of stand ages across the landscape (Senici et al., 2010). Dominant tree species in our study area, in order from least to most shade tolerant, include jack pine (Pinus banksiana Lamb.), trembling aspen (Populus tremuloides Michx.), white birch (Betula papyrifera Marsh.), black spruce (Picea mariana Mill. BSP), and balsam fir (Abies balsamea [L.] Mill.).

Fig. 1. Conceptual diagram of hypothetical change in strength of the diversity–productivity relationship following post-fire boreal forest succession with changes in functional diversity. 2

Forest Ecology and Management 455 (2020) 117641

A.R. Taylor, et al.

Canadian system of soil classification (Soil Classification Working Group, 1998). To ensure that each sample stand met the selection criteria, soil pits were dug in each candidate stand to verify whether the site was mesic, following the procedures described in Taylor et al. (2000). Soil attribute data for all sites used in this study have been previously reported by Hume et al. (2016). Stand age for sample stands less than 90 years old was determined from detailed fire history records for our study area (Hart & Chen, 2008; Senici et al., 2010). For stands greater than 90 years old, tree ages were used to estimate minimum stand age following the procedures described in Senici et al. (2010). Of all the stands sampled, we selected either jack pine or trembling aspen trees to determine minimum stand age as both species are shade intolerant and regenerate immediately following fire (Ilisson & Chen, 2009). In each stand, three canopy stems were sampled by extracting a core or stem disc at breast height (1.3 m above root collar). The cores and discs were transported to our laboratory, where the cores were mounted on constructed core strips and sanded to make rings visible. Stem discs were cut transversely, then mounted on constructed core strips and sanded to make rings visible. Rings were counted using a handheld magnifier or a microscope until the same count was obtained three successive times. Based on a locally derived age-correction model developed by Vasiliauskas and Chen (2002), 7 years were added to ring counts to determine minimum stand age.

Fig. 2. Map showing location of study area (white square) north of Thunder Bay, Ontario, Canada. White square indicates general area where sample stands were selected based on road access, stand inventory availability, and local fire history.

2.2. Sampling design We used a detailed, chronosequence sampling design to test whether the effect of neighbourhood species diversity on tree growth varies as forests undergo secondary succession. Although the use of the chronosequence method has been criticized due to its assumption that sample stands along the temporal sequence have followed the same developmental history (Johnson & Miyanishi, 2008), given careful site selection, replication, and demonstration of developmental links, the chronosequence method is appropriate for studying forest succession patterns over decadal to centennial time scales (Walker et al., 2010). Based on local fire history and the availability of different aged stands in the study area, we were able to sample six different stand ages (i.e., time since last stand-replacing fire). This covered early, mid, and late boreal forest secondary succession, including 8, 16, 34, 99, 147, and 210 years since fire, which represent the stand initiation (8–16 years old), stem exclusion (34 years old), canopy transition (99–147 years old), and gap dynamics (210 years old) stages of boreal forest stand dynamics, respectively (Chen & Popadiouk, 2002). Boreal forest stands may undergo multiple succession pathways following stand-replacing disturbance (Taylor & Chen, 2011). To account for this, we sampled a variety of stands of different overstorey species compositions (i.e., broadleaf, conifer, and mixedwood overstorey stand types) for each stand age (Table 1). In the central Canadian boreal forest, as stands regenerate following stand-replacing fire, they become dominated by either trembling aspen (broadleaf type), jack pine (conifer type), or their mixture (mixedwood type). As post-fire stands age into late succession, they transition into stands dominated by mixed trembling aspen and white birch (broadleaf type), mixed black spruce and balsam fir (conifer type), and their mixture (mixedwood type) (e.g., Carleton & Maycock, 1978; Frelich & Reich, 1995; Bergeron, 2000; Taylor & Chen 2011). We selected post-fire stands that had not been managed (e.g., planted, sprayed, or thinned), including three replicates for each age class and overstorey type combination; however, one replicate mixedwood stand at age 147 years was accidentally damaged by harvesting activities between consecutive years of sampling, resulting in a total of 53 stands measured overall. Sample stands were positioned several kilometers away from each other and selected from different road accesses to minimize the impact of spatial autocorrelation. In order to minimize variability in site and soil conditions, all selected stands were located on mesic site types with flat mid-slope positions, with no slope exceeding 5%. All stands are underlain by moderately deep (≥50 cm) glacial tills, belonging to the Brunisolic soil order, according to the

2.3. Data collection Within each of the 53 sample stands, we randomly located and established a 0.04 ha (11.28 m radius) fixed-area circular plot, approximately 50 or more meters from the forest edge, to represent the stand. We recorded species identity and diameter at breast height (DBH, 1.3 m above root collar) for all trees larger than 1 cm DBH within each plot. Overstorey types were assigned based on the relative basal area of broadleaf and conifer tree species in a plot. Broadleaf and conifer stands were defined as having > 80% broadleaf or conifer tree species composition by stem density or basal area, and all other stands were classified as mixedwood stands (Table 1). To determine tree growth, we used stratified random sampling to select trees of varying DBH, from tree stems with a minimum DBH of 2 cm to the maximum DBH that could be found for each species in a given sample plot. Because species composition changes with succession, we sampled jack pine from the 8-, 16-, 34-, 99-, and 147-year-old stands; trembling aspen from all age stands; white birch, black spruce, and balsam fir from the 99-, 147-, and 210-year-old stands. For each species within each sample plot, tree size was grouped into 4 cm DBH interval classes, and we randomly sampled up to three trees from each DBH class if available. Trees with crooked stems, heart rot, or other forms of stem damage, such as stem abrasion, fungal infections, or major branch losses, were not sampled. In total, we sampled 163 trees for jack pine, 286 trees for trembling aspen, 229 trees for white birch, 167 trees for black spruce, and 135 trees for balsam fir across all overstorey types and stand ages. Appendix S1 (Table S1.1) shows the number of sample trees per species, age class, and tree size. For each sample tree, increment cores or disc samples were collected to estimate annual tree radial growth increment. For trees < 10 cm in DBH, we cut a stem disc at DBH, and for trees ≥10 cm in DBH, we used an increment core borer with a 5.15 mm diameter bit to extract stem core samples. All samples were sealed in plastic bags (for discs) or straws (for cores) and transported from the field to the laboratory. In the laboratory, all stem disc and core samples were sanded and annual increments were measured using a WinDENDRO measuring system. Tree annual growth rate was calculated as the average annual basal area increment of the last 5 years (2009–2014). Because tree-ring measurements were conducted for growth within tree bark, whereas tree basal area growth is calculated for DBH outside the bark, similar to 3

Forest Ecology and Management 455 (2020) 117641

A.R. Taylor, et al.

Table 1 Characteristics of the 53 stands sampled for this study, located in the boreal forest of northern Ontario, Canada. Values are means with 1 standard error in parentheses. Age (years)

8

Broadleaf Conifer Mixedwood Broadleaf Conifer Mixedwood Broadleaf Conifer Mixedwood Broadleaf Conifer Mixedwood Broadleaf Conifer Mixedwood Broadleaf Conifer Mixedwood

16

34

99

147

b

210

a b

Overstorey

Basal area (m2 ha−1)

Stand composition (%)

3 (1) 4 (1) 5 (1) 20 (2) 9 (1) 9 (1) 25 (1) 28 (2) 17 (3) 51 (7) 52 (2) 43 (5) 58 (8) 51 (9) 36 (3) 41 (3) 40 (8) 46 (3)

a

Trembling aspen

White birch

Jack pine

92 (3)

3 (3)

30 (6) 89 (6) 1 (1) 30 (3) 94 (3) 4 (2) 50 (4) 91 (2) 3 (2) 40 (12) 85 (3) 1 (1) 38 (2) 54 (22) 5 (5) 11 (4)

4 (2) 9 (5) 1 (1)

3 (3) 100 (0) 66 (4) 1 (1) 97 (2) 67 (1) 1 (1) 94 (3) 41 (5)

4. (4) 1 (1) 3 (2) 16 (11) 7 (4) 2 (2) 30 (1) 24 (18) 7 (4) 39 (5)

43 (12) 9 (6) 53 (27)

5 (3)

Black spruce

Balsam fir

Others 2 (1)

2 (2) 1 (1) 1 (1) 8 (5) 1 (1) 50 (17) 15 (8) 5 (1) 37 (26) 12 (5) 10 (6) 36 (18) 38 (7)

1 (1) 1 (1) 1 (1)

1 (1) 4 (3) 18 (3) 2 (1) 7 (1) 19 (4) 10 (4) 50 (17) 7 (3)

1 (1) 4 (1) 2 (1) 1 (1) 1 (1) 2 (1) 2 (1)

The ‘Others’ category includes Salix spp., Acer spicatum, Alnus viridis, Sorbus decora, Corylus cornuta, Prunus pensylvanica, and Larix laricina Each age–overstorey combination has three replications, except the 147-yr-old stands

Chen and Klinka (2003), we first developed species-specific relationships between DBH with and without bark. We then used these relationships to calculate DBH with bark from measured stemwood DBH in 2009 and 2014 for each sample. We calculated Shannon’s diversity index (H) as a measure of species diversity for each sample plot, which accounts for both species richness and evenness (Zhang et al., 2012): S

H = −∑

i=1

Pi ln (Pi )

represents stand age (years) as a categorical variable corresponding to stage of secondary succession; RS is relative tree size (a continuous variable); BA is plot-level basal area (m2 ha−1; a continuous variable), which accounts for plot stand density; πplot is the random effect of sample plots, which accounts for autocorrelation among trees sampled within each plot; and ε is the sampling error. All continuous variables were centred prior to analysis, and we conducted each linear mixedeffect model using restricted maximum likelihood estimation with the R “lme4” package (Bates et al., 2017). To evaluate the necessity of adding a random regression coefficient for plot, naive models (without a random effect) were compared with mixed-effect models by performing likelihood ratio tests and comparing AIC values between models. For all species’ models, adding a random coefficient significantly reduced error variance. Scatter plots of model residuals versus fitted values and each explanatory variable, Normal QQ plots, Levene's test, and the Breusch-Pagan test were conducted to assess heteroskedasticity and normality of each model’s residuals. Significant heteroskedasticity was detected for each model and was corrected by natural log transformation of the response variable (Appendix S2, Fig. S2.1-S2.4). Model fit was further evaluated using marginal and conditional R2 as calculated by the R “MuMIn” package and the methods of Nakagawa and Schielzeth (2013). The marginal R2 estimates variation explained by only the fixed effect portion of the model, while the conditional R2 also includes variation accounted for by the random effect, which approximates a ‘traditional’ R2 provided from standard linear regression. Significance of regression terms was evaluated using F-tests based on the Satterthwaite approximation of degrees of freedom (Schaalje et al., 2002). For models in which the 3way interaction term was not significant (p > 0.05), it was removed to reduce loss of degrees of freedom and ease model interpretation. However, non-significant two-way interaction terms were retained as these were useful in testing and illustrating our main hypotheses (i.e., whether neighbourhood species diversity effects on tree growth significantly varied by successional stage and relative tree size). Post hoc simple linear regressions of the fitted values produced from the linear mixed-effect models were conducted to assist in interpreting the partial effects of each explanatory variable. To help interpret the partial effects of relative tree size on growth–diversity relationships, relative tree size was grouped into “small” and “large” tree categories for each species using the mean relative tree size for each species as the split point. Separate post hoc simple linear regressions of the small and

(1)

where S is the species richness, Pi is the relative abundance of species i based on basal area of the ith species. To examine whether the effect of species diversity on tree growth varied according to the size of the sample tree, relative to all other trees in the sample plot, we also calculated relative tree size for all sample trees by using basal area (BA) of each sample tree divided by the mean BA of all trees within the sample plot. We used relative tree size instead of the absolute tree size because our study covered a wide range of stand ages, and relative size better reflects the competitive position of individual trees when encountering other individuals in a forest stand (Luo & Chen, 2015). 2.4. Data analysis To address our hypotheses concerning the influence of secondary forest succession and relative tree size on the relationship between neighbourhood species diversity and tree growth, we used linear mixedeffect models to test the effects of Shannon’s diversity index, stand age, relative tree size, and their interactions on tree growth rate for each of the five tree species in our study. We initially included tree species as a fixed factor in a single linear mixed-effect model to account for differential species effects; however, although tree species was highly significant, we opted to conduct separate, species-specific linear mixedeffect models to ease interpretation of models and evaluate unique species-specific responses. The following model structure was used to analyze each species:

Y = β0 + β1 H + β2 A + β3 RS + β4 H × A + β5 A × RS + β6 H × RS + β7 H × A × RS + β8 BA + πplot + ε

(2) 2

−1

where Y is annual basal area growth rate (cm year ) for a given tree species; H is Shannon’s diversity index (a continuous variable); A 4

Forest Ecology and Management 455 (2020) 117641

A.R. Taylor, et al.

detected for small jack pine trees (Fig. 4a, b). During the early canopy transition stage (i.e., 99-year-old stands), the growth rate of trembling aspen, again, showed significant decreases with increasing diversity (Fig. 3d) for both small and large trees (Fig. 4c, d). Although no overall growth–diversity relationship for white birch was detected in the 99-year-old stands (Fig. 3d), a significant, positive trend was observed for large white birch trees during this time (Fig. 4f). Interestingly, all of the conifer species sampled (i.e., jack pine, black spruce, and balsam fir) showed significant, positive tree growth trends with increasing species diversity during the canopy transition stage (Fig. 3d) for all relative tree sizes, expect for large balsam fir trees (Fig. 4j), for which the sample size was quite low (only 13 trees). Further, its worth noting that although white birch and black spruce did not show significant age x diversity interactions, this does not negate the fact that significant (non-zero) diversity-growth slopes at one or more levels of age and/or relative tree size may still occur and provide relevant, biological insight, as demonstrated here by the significant growth-diversity trends for large white birch trees and black spruce in the 99-year-old stands. During the late canopy transition and gap dynamics stages of boreal forest succession (i.e., 147 and 210-year-old stands), the strength of relationships between tree growth rate and species diversity appeared to diminish. Despite an unexpected, but strong, positive growth–diversity trend observed for trembling aspen in the 147-year-old stands (Fig. 3e), balsam fir was the only other species to demonstrate a significant growth–diversity relationship during late succession, with an overall positive trend observed in the 147-year-old stands (Fig. 3e) and a positive trend for small trees in the 210-year-old stands (Fig. 4i). Also worth noting is that the strong, positive growth–diversity trend observed for trembling aspen in the 147-year-old stands was primarily driven by relatively large trees, as the smaller trees showed a significant, but opposite, negative trend (Fig. 4c, d). Relative tree size had a significant effect on growth–diversity relationships (Table 2; Fig. 4), but no clear pattern could be discerned as the direction and size of this effect varied idiosyncratically among species. Furthermore, aside from the significant interaction effect observed in four of the species’ models, the main effect of relative tree size was highly significant for all five species (Table 2), which is well demonstrated by the strong, positive, linear relationships for tree growth rate versus relative tree size across all stand ages (Appendix S3, Fig. S3.1). It is interesting to note that the strong, positive growth–size relationships we observed (Fig. S3.1a, b, c, e, f) appeared to weaken in mid-succession (99-year-old stands), as indicated by the shallower slopes and insignificant growth–size trend for white birch (Fig. S3.1d). This, coincidentally, is also the stage of succession during which we observed the strongest growth–diversity relationships and may suggest the effect of species diversity on tree growth overrides the influence of

large tree data for each species and stand age were then conducted using the fitted values from the linear mixed-effect models. To help interpret whether the effect of secondary succession on the relationship between species diversity and tree growth was related to changes in functional diversity over time, we calculated functional dispersion (Laliberté and Legendre, 2010) using the R “FD” package (Laliberté et al., 2015) for each sample plot. Functional dispersion is a multidimensional functional diversity index that provides mean distance in multidimensional trait space of individual species to the centroid of all species and takes into consideration species abundance. Furthermore, unlike other commonly used functional diversity indices (e.g., Villéger et al., 2008), functional dispersion is less sensitive to small communities (< 3 species) such as encountered in our dataset. To calculate functional dispersion we used shade tolerance and drought tolerance, provided by Niinemets and Valladares (2006), as functional traits. Together, both traits provide a measure of species adaptations to both above and below ground site resources (light and soil moisture). We used analysis of variance to test whether functional dispersion significantly varied with stand age and visually assessed how temporal patterns in functional dispersion compared with growth-diversity trends uncovered by our linear mixed-effect models. All data analyses were performed in R 3.5.1 (R Development Core Team, 2019). 3. Results 3.1. Effects of species diversity, stand age, and relative tree size on growth Significant (p < 0.05) interaction terms were detected in each species’ model except for the white birch model (Table 2). Significant three-way interactions were observed for the jack pine and balsam fir models, but not for trembling aspen or black spruce. The strong, but differing, interaction effects observed among models suggest the relationship between tree growth rate and diversity is not only species specific, but is also influenced by stand age (i.e., succession) and relative tree size. Furthermore, the high conditional R2 values (≥0.75) indicate a substantial level of variation in tree growth rate was accounted for by each species’ model, especially for trembling aspen (Table 2). During the stand initiation stages of boreal forest succession, from age 8 to 16 years, the relationship between tree growth and diversity was not significant for trembling aspen or jack pine (Fig. 3a, b), regardless of relative tree size (Fig. 4a–d). However, as stands entered into the stem exclusion stage of succession (i.e., 34-year-old stands), a significant (p < 0.05), negative growth–diversity trend was observed for trembling aspen (Fig. 3c), for both small and large trees (Fig. 4c, d). Although no overall significant trend was observed for jack pine during this stage (Fig. 3c), a significant positive growth–diversity trend was

Table 2 Model fit (R2) and the effects (p values) of stand age (A), diversity (H), relative tree size (RS) and their interactions, with stand basal area (BA) as a covariate, on annual tree growth rate for five major tree species. Statistically significant terms (p < 0.05) are in bold. NA indicates interaction term was not significant and removed from model. Source

Jack pine

Trembling aspen

White birch

Black spruce

Balsam fir

H A RS BA A×H A × RS H × RS A × H × RS

0.877 0.033 < 0.001 0.952 0.616 0.001 < 0.001 0.003

0.650 0.068 < 0.001 0.353 0.031 < 0.001 < 0.001 NA

0.357 0.410 < 0.001 0.514 0.333 0.108 0.667 NA

0.357 < 0.001 < 0.001 0.059 0.127 0.058 0.032 NA

0.815 0.332 < 0.001 0.437 0.085 0.002 < 0.001 0.003

Marginal R2 Conditional R2

0.72 0.87

0.84 0.93

0.50 0.75

0.75 0.75

0.64 0.79

Linear mixed effect model fit tests used Satterthwaite approximations of degrees of freedom. Marginal and conditional R2 calculated using the R MuMIn package based on the method of Nakagawa & Schielzeth (2013). 5

Forest Ecology and Management 455 (2020) 117641

A.R. Taylor, et al.

Fig. 3. Effect of diversity (Shannon’s index) on the annual basal area growth rate of five tree species in relation to stand age after accounting for relative size and stand-level basal area. Colours indicate species. Lines are simple linear regressions of fitted values derived from the linear mixed-effect models. Solid lines indicate the slope is significantly (p < 0.05) different from zero. BF = balsam fir, BS = black spruce, JP = jack pine, TA = trembling aspen, WB = white birch.

jack pine being the least shade-tolerant and most drought-tolerant of our study species, and balsam fir being the opposite.

relative tree size during this stage of succession. 3.2. Role of functional diversity on species diversity effects

4. Discussion Functional diversity (i.e., dispersion) did not significantly vary with stand age (F = 1.22, p = 0.32). Nonetheless, marginal differences in mean functional dispersion were visually apparent across secondary succession (Fig. 5) with functional diversity appearing lowest in early succession, peaking during mid-succession (99-year-old stands), then slightly decreasing in late-succession. The marginal peak in functional diversity in the 99-year-old stands is largely driven by the higher cooccurrence of jack pine and balsam fir (Table 1), which vary considerably from one another in both shade- and drought-tolerance, with

Our results provide compelling evidence that the effect of neighbourhood species diversity on tree species growth changes as boreal forests undergo secondary succession and is strongest during mid-succession, supporting our first hypothesis and corroborating previous reports that ecological succession drives temporal shifts in the strength and direction of diversity–productivity relationships (Guo, 2003; Paquette & Messier, 2011; Barrufol et al., 2013; Lasky et al., 2014; Mori et al., 2017). 6

Forest Ecology and Management 455 (2020) 117641

A.R. Taylor, et al.

Fig. 5. Mean functional diversity, calculated as functional dispersion, of sample plots for each stand age. Higher dispersion values indicate greater functional diversity. Error bars show standard error.

(Madrigal-González et al., 2016; Báez & Homeier, 2018). 4.1. Hump-shaped pattern in tree growth–diversity relationship during succession During early succession, we found no strong evidence of growth–diversity relationships, which is consistent with our hypothesis that high resource availability following fire may lessen the effect of complementarity on tree growth, affirming previous reports (e.g., Zhang et al. 2012; Forrester, 2014). However, as stands entered the competitive stem exclusion stage (34-year-old stands), significant positive and negative growth–diversity trends emerged, indicating the effect of species diversity on growth strengthens as competition for resources intensifies. Indeed, as stands aged through mid- and late-succession (99-, 147-, and 210-year-old stands), a general “hump-shaped” pattern in the overall strength of community growth–diversity relationships could be discerned, as conceptualized in Fig. 1. This hump-shaped pattern approximately coincides with observed functional diversity with stand age (Fig. 5) and supports our hypothesis that species diversity effects on tree growth increase during key transitional stages of forest succession in the boreal forest (Huston & Smith, 1987; Forrester, 2014; Reich, 2014), such as when communities shift from fast-growing, early succession colonizers to slower-growing, shade-tolerant species of contrasting functional traits. Further, the observed hump-shaped pattern also coincides with previous work in our study area (e.g., Taylor et al., 2014; Laganière et al., 2015; Gao et al., 2018) and elsewhere (Rothstein et al., 2004; Goulden et al., 2011) that forest carbon sequestration tends to be highest during mid-succession, suggesting succession-driven species diversity may contribute to temporal shifts in boreal forest productivity (Taylor et al., 2014; Laganière et al., 2015; Gao et al., 2018). Interestingly, the observed hump-shaped pattern contradicts studies of secondary succession in tropical forests. For instance, Lasky et al. (2014) reported stronger, positive growth–diversity trends during early succession, followed by weak, negative growth–diversity trends in older stands, possibly attributed to the saturating effect of greater species diversity on ecosystem function as stands aged (Tilman et al., 1997; Hooper et al., 2005). However, saturating effects on growth–diversity relationships may be less important in boreal systems given species diversity (and functional diversity) is generally much smaller (Pan et al., 2013). Also, it is challenging to draw direct comparisons of growth–diversity relations over succession between (and within) forest biomes as the effects of diversity on forest productivity are known to

Fig. 4. Effect of diversity (Shannon’s index) on annual basal area growth rate of five tree species in relation to relative size and stand age. Values (simple linear regression slope coefficients and 95% confidence interval error bars) represent the diversity effect for given age class and relative tree size (small and large). Values whose error bars do not overlap with zero indicate a significant simple linear regression slope. BF = balsam fir, BS = black spruce, JP = jack pine, TA = trembling aspen, WB = white birch.

Although our results showed that relative tree size had a significant effect on the relationship between neighbourhood species diversity and tree species growth, no clear pattern could be discerned, as the strength and direction of this effect varied considerably between species and over time (Fig. 4). This contradicts our second hypothesis that the effect of species diversity on tree growth increases with relative tree size, due to size-asymmetric competitive ability (Coomes et al., 2011), but does support previous reports that size-dependent effects are highly variable

7

Forest Ecology and Management 455 (2020) 117641

A.R. Taylor, et al.

response is unclear, but may be related to the fact that balsam fir was the most abundant understorey tree species found in the 210-year-old stands, as is typical during late succession in the study area, given fir’s high shade tolerance (Humbert et al., 2007; Taylor & Chen, 2011), and was better able to take advantage of greater resource availability when mixed with aspen and birch (e.g., enhanced understorey light and nutrient-rich broadleaf litter). However, the overall, negligible tree growth–diversity relationships we observed in late succession may be attributable to reduced complementarity caused by greater availability of light and soil resources during the gap dynamics stage compared with mid-succession stages (Hume et al., 2016; Kumar et al., 2018). Furthermore, during the gap dynamics stage, stands tend to be dominated by large, old trees, which allocate more resources to self-maintenance than radial growth (Mencuccini et al., 2005), which may have further limited detection of diversity effects on tree growth.

vary according to environmental context (Fridley, 2001; Jucker et al., 2016); and the nature of succession itself varies widely among forest types (Glenn-Lewin et al., 1992). For example, in western Canada the average fire return interval is much shorter (< 50 years in parts) than central and eastern boreal regions (Boulanger et al., 2014; Senici et al., 2010). Consequently, it is less likely western boreal forests would experience successional sequences “long enough” for hump-shaped patterns in growth–diversity relations to emerge (Johnstone & Chapin, 2006; Kurkowski et al., 2008). 4.2. Tree growth–diversity relationships are individualistic Despite the general hump-shaped pattern we uncovered, the strength and direction of individual species growth–diversity relationships varied considerably, supporting previous reports (e.g., Fichtner et al., 2017), but contrasting with others (e.g., Chamagne et al., 2017). As stands entered the stem exclusion and early canopy transition stages (34- and 99-year-old stands), trembling aspen exhibited a strong, negative growth–diversity trend, which coincided with an increase in the relative abundance of conifer tree neighbours (Table 1). Similarly, MacPherson et al. (2001) and Edgar and Burk (2001) both reported aspen was more productive in monospecific than in mixed boreal stands. This decrease in growth may be driven by reductions in light availability and alteration of soil chemistry caused by the presence of conifers in more diverse stand types (Prescott et al., 2000; Calder et al., 2011) and may be most noticeable when soil nutrients are already more limited during and following the stem exclusion stage (Calder et al., 2011). Indeed, Hume et al. (2016) reported a significant decrease in soil nitrogen concentration in the upper soil layers of the mixedwood stands during this stage of succession in our study plots. Unexpectedly, during late canopy transition (147-year-old stands), trembling aspen exhibited an overall positive growth–diversity trend, but this was driven by large aspen trees, as smaller trees still showed a negative trend. Explanations for this divergence are unclear, but it may be related to significant increases in soil nitrogen concentration in the mixedwoods from age 99 to 147 years, as observed by Hume et al. (2016), indicating soil limitations on aspen growth may subside as stands continue to age. Alternatively, it may be because in the 147-yearold stands, the large, old aspen trees present were substantially larger than all other trees measured and thus may have benefited from reduced overstorey competition for light when mixed with conifers, highlighting the positive contribution of crown complementarity on diversity-enhanced productivity (Williams et al., 2017). Although we did not observe as strong a diversity effect on the growth of white birch, birch trees in the 99-year-old stands exhibited a similar response as trembling aspen in the 147-year-old stands, in that mixing with conifers had a significant, negative effect on small birch, but a positive effect on large birch. The overall, weaker diversity effect on birch may be related to it being a more shade-tolerant, slower growing species (relative to aspen) that is less sensitive to changes in light and soil conditions caused by the presence of conifers. In contrast to the broadleaf species, strong positive growth–diversity relationships were observed for each of the conifer species throughout the stem exclusion and canopy transition stages (34- and147-year-old stands). These positive growth–diversity trends may have resulted from improved soil conditions (i.e., facilitation) from inputs of nutrient-rich broadleaf litter when conifers are mixed with aspen and birch (Cote et al., 2000; Calder et al., 2011; Laganière et al., 2015; Hume et al., 2016), enhanced light availability through canopy tree crown complementarity (Williams et al., 2017), or even below ground complementarity driven by niche partitioning of soil moisture resources by jack pine and balsam fir; but, further measurements and analyses are required to test these hypotheses. During the late-succession, gap dynamics stage (210-year-old stands), only small balsam fir trees (DBH range: 2.1–19.3 cm) exhibited a significant growth response to species diversity. The cause of this

5. Conclusion Using a detailed, replicated chronosequence design covering entire, long-term post-fire successional sequences common to Canada’s central boreal forest, we found evidence that the relationship between neighbourhood species diversity and tree species growth significantly varied during secondary succession. Our results showed the strength of this effect followed a general “hump-shaped” pattern over time, with midsuccessional stages of high functional diversity exhibiting the strongest growth–diversity trends; but at the species level, responses were individualistic, exhibiting positive, negative, and neutral trends. Our study yielded conflicting results for the effect of tree size on the relationship between species diversity and growth, contradicting the hypothesis that larger trees benefit more from complementarity due to size-asymmetric competitive ability, but supporting previous reports that size-dependent effects are highly variable. From a forest management perspective, our results indicate creating or maintaining mixed species stand structures of contrasting functional traits (e.g., broadleaf versus needle leaf tree species or combinations of species with varying shade and drought tolerances) may encourage greater growth of commercial conifer species (e.g., jack pine, black spruce and balsam fir). Such may be achieved, for example, by promoting shade-tolerant conifers in young post-fire stands through precommercial thinning and/or fill-planting. Alternatively, given the negative effect of species diversity on aspen growth we observed, if the management objective was to facilitate higher aspen growth (perhaps to support oriented strand board and rayon markets in Canada’s boreal forest), maintaining high aspen dominance throughout stand succession, by removing conifer competition, may be advantageous. Overall, our results contribute to disentangling the mechanisms that link species diversity to forest ecosystem productivity, which is important for understanding and predicting the consequences of global biodiversity loss, especially in the context of the boreal forest, which plays a critical role in controlling global carbon flux and climate change. However, continued study is warranted to further understand how and why growth-diversity trends change over time as successional dynamics vary widely among and within the world’s forest biomes.

Acknowledgements We thank Siyao Yang, Hua Liu, and Wanwen Yu for assistance with field work and Eric Searle and Masumi Hisano for their constructive input on data analysis. We also thank Caroline Simpson who provided useful comments on this manuscript. This study was funded by the Natural Sciences and Engineering Research Council of Canada Strategic Grant Project (STPGP428641) and Discovery Program (RGPIN-20140418), and Natural Resources Canada.

8

Forest Ecology and Management 455 (2020) 117641

A.R. Taylor, et al.

Authors’ contributions

diversity, and composition in boreal forest. Ecol. Monogr. 78, 123–140. Hooper, D.U., Chapin, F.S., Ewel, J.J., Hector, A., Inchausti, P., Lavorel, S., Wardle, D.A., 2005. Effects of biodiversity on ecosystem functioning: a consensus of current knowledge. Ecol. Monogr. 75, 3–35. Hooper, D.U., Adair, E.C., Cardinale, B.J., Byrnes, J.E., Hungate, B.A., Matulich, K.L., O’Connor, M.I., 2012. A global synthesis reveals biodiversity loss as a major driver of ecosystem change. Nature 486, 105. Humbert, L., Gagnon, D., Kneeshaw, D., Messier, C., 2007. A shade tolerance index for common understory species of northeastern North America. Ecol. Ind. 7, 195–207. Hume, A., Chen, H.Y.H., Taylor, A.R., Kayahara, G.J., Man, R.Z., 2016. Soil C:N:P dynamics during secondary succession following fire in the boreal forest of central Canada. For. Ecol. Manage. 369, 1–9. Huston, M., Smith, T., 1987. Plant succession: life history and competition. Am. Nat. 130, 168–198. Ilisson, T., Chen, H.Y.H., 2009. Response of six boreal tree species to stand replacing fire and clearcutting. Ecosystems 12, 820–829. Johnson, E.A., Miyanishi, K., 2008. Testing the assumptions of chronosequences in succession. Ecol. Lett. 11, 419–431. Johnstone, J.F., Chapin, F.S., 2006. Fire interval effects on successional trajectory in boreal forests of northwest Canada. Ecosystems 9, 268–277. Jucker, T., Avăcăriței, D., Bărnoaiea, I., Duduman, G., Bouriaud, O., Coomes, D.A., 2016. Climate modulates the effects of tree diversity on forest productivity. J. Ecol. 104, 388–398. Kumar, P., Chen, H.Y., Thomas, S.C., Shahi, C., 2018. Linking resource availability and heterogeneity to understorey species diversity through succession in boreal forest of Canada. J. Ecol. 106, 1266–1276. Kurkowski, T.A., Mann, D.H., Rupp, S.T., Verbyla, D.L., 2008. Relative importance of different secondary successional pathways in an Alaskan boreal forest. Can. J. For. Res. 38, 1911–1923. Laganière, J., Cavard, X., Brassard, B.W., Paré, D., Bergeron, Y., Chen, H.Y., 2015. The influence of boreal tree species mixtures on ecosystem carbon storage and fluxes. For. Ecol. Manage. 354, 119–129. Laliberté, Legendre, & Shipley, 2015. Package “FD”. https://cran.r-project.org/web/ packages/FD/FD.pdf (accessed on: 3 May 2019). Laliberté, E., Legendre, P., 2010. A distance based framework for measuring functional diversity from multiple traits. Ecology 91, 299–305. Lasky, J.R., Uriarte, M., Boukili, V.K., Erickson, D.L., John Kress, W., Chazdon, R.L., 2014. The relationship between tree biodiversity and biomass dynamics changes with tropical forest succession. Ecol. Lett. 17, 1158–1167. Liang, J., Crowther, T.W., Picard, N., Wiser, S., Zhou, M., Alberti, G., De–Miguel, S., 2016. Positive biodiversity–productivity relationship predominant in global forests. Science 354 (6309) aa8957. Luo, Y., Chen, H.Y.H., 2015. Climate change–associated tree mortality increases without decreasing water availability. Ecol. Lett. 18, 1207–1215. MacPherson, D.M., Lieffers, V.J., Blenis, P.V., 2001. Productivity of aspen stands with and without a spruce understory in Alberta's boreal mixedwood forests. For. Chronicle 77, 351–356. Madrigal-González, J., Ruiz-Benito, P., Ratcliffe, S., Calatayud, J., Kändler, G., Lehtonen, A., et al., 2016. Complementarity effects on tree growth are contingent on tree size and climatic conditions across Europe. Sci. Rep. 6, 32233. Maestre, F.T., Callaway, R.M., Valladares, F., Lortie, C.J., 2009. Refining the stress-gradient hypothesis for competition and facilitation in plant communities. J. Ecol. 97, 199–205. Mencuccini, M., Martinez-Vilalta, J., Vanderklein, D., Hamid, H.A., Korakaki, E., Lee, S., Michiels, B., 2005. Size–mediated ageing reduces vigour in trees. Ecol. Lett. 8, 1183–1190. Mina, M., Huber, M.O., Forrester, D.I., Thürig, E., Rohner, B., Hector, A., 2018. Multiple factors modulate tree growth complementarity in Central European mixed forests. J. Ecol. 106, 1106–1119. Mori, A.S., Osono, T., Cornelissen, H.C., Craine, J., Uchida, M., 2017. Biodiversity–ecosystem function relationships change through primary succession. Oikos 126, 1637–1649. Nakagawa, S., Schielzeth, H., 2013. A general and simple method for obtaining R2 from generalized linear mixed effects models. Methods Ecol. Evol. 4, 133–142. Niinemets, U., Valladares, F., 2006. Tolerance to shade, drought, and waterlogging of temperate Northern Hemisphere trees and shrubs. Ecol. Monogr. 76, 521–547. Odum, E.P., 1969. The strategy of ecosystem development. Science 164, 262–270. Pacala, S.W., Rees, M., 1998. Models suggesting field experiments to test two hypotheses explaining successional diversity. Am. Nat. 152, 729–737. Pan, Y., Birdsay, R.A., Phillips, O.L., Jackson, R.B., 2013. The structure, distribution, and biomass of the world’s forests. Annu. Rev. Ecol. Evol. Syst. 44, 593–622. Paquette, A., Messier, C., 2011. The effect of biodiversity on tree productivity: from temperate to boreal forests. Glob. Ecol. Biogeogr. 20, 170–180. Prescott, C.E., Zabek, L.M., Staley, C.L., Kabzems, R., 2000. Decomposition of broadleaf and needle litter in forests of British Columbia: influences of litter type, forest type, and litter mixtures. Can. J. For. Res. 30, 1742–1750. R Development Core Team, 2019. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. Reich, P.B., 2014. The world–wide ‘fast–slow’ plant economics spectrum: a traits manifesto. J. Ecol. 102, 275–301. Rothstein, D.E., Yermakov, Z., Buell, A.L., 2004. Loss and recovery of ecosystem carbon pools following stand-replacing wildfire in Michigan jack pine forests. Can. J. For. Res. 34, 1908–1918. Schaalje, G.B., McBride, J.B., Fellingham, G.W., 2002. Adequacy of approximations to distributions of test statistics in complex mixed linear models. J. Agric., Biol. Environ. Statist. 7, 512.

ART, BG and HYHC conceived the ideas and designed methodology; BG and HYHC collected the data; ART and BG analysed the data; ART led the writing of the manuscript. All authors contributed critically to the drafts and gave final approval for publication. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.foreco.2019.117641. References Báez, S., Homeier, J., 2018. Functional traits determine tree growth and ecosystem productivity of a tropical montane forest: insights from a long–term nutrient manipulation experiment. Glob. Change Biol. 24, 399–409. Barrufol, M., Schmid, B., Bruelheide, H., Chi, X., Hector, A., Ma, K., et al., 2013. Biodiversity promotes tree growth during succession in subtropical forest. PLoS One 8 (11), e81246. Bates, D., Maecler, M., Bolker, B., Walker, S., Christensen, R., Singmann, H., et al., 2017. Package “lme4”. https://cran.r-project.org/web/packages/lme4/lme4.pdf (accessed on: 3 March 2019). Bergeron, Y., 2000. Species and stand dynamics in the mixed woods of Quebec's southern boreal forest. Ecology 81, 1500–1516. Bergeron, Y., Dubue, M., 1988. Succession in the southern part of the Canadian boreal forest. Vegetatio 79, 51–63. Bertness, M.D., Callaway, R., 1994. Positive interactions in communities. Trends Ecol. Evol. 9, 191–193. Boulanger, Y., Gauthier, S., Burton, P.J., 2014. A refinement of models projecting future Canadian fire regimes using homogeneous fire regime zones. Can. J. Forest Res. 44 (4), 365–376. Calder, W.J., Horn, K.J., St. Clair, S.B., 2011. Conifer expansion reduces the competitive ability and herbivore defense of aspen by modifying light environment and soil chemistry. Tree Physiol. 31, 582–591. Carleton, T.J., Maycock, P.F., 1978. Dynamics of the boreal forest south of James Bay. Can. J. Bot. 56, 1157–1173. Cavard, X., Bergeron, Y., Chen, H.Y.H., Paré, D., 2010. Mixed–species effect on tree aboveground carbon pools in the east–central boreal forests. Can. J. For. Res. 40, 37–47. Chamagne, J., Tanadini, M., Frank, D., Matula, R., Paine, C.E.T., Philipson, C.D., Hector, A., 2017. Forest diversity promotes individual tree growth in central European forest stands. J. Appl. Ecol. 54, 71–79. Chen, H.Y.H., Klinka, K., 2003. Aboveground productivity of western hemlock and western redcedar mixed–species stands in southern coastal British Columbia. For. Ecol. Manage. 184, 55–64. Chen, H.Y.H., Popadiouk, R.V., 2002. Dynamics of North American boreal mixedwoods. Environ. Rev. 10, 137–166. Coomes, D.A., Lines, E.R., Allen, R.B., 2011. Moving on from metabolic scaling theory: hierarchical models of tree growth and asymmetric competition for light. J. Ecol. 99, 748–756. Cote, L., Brown, S., Pare, D., Fyles, J., Bauhus, J., 2000. Dynamics of carbon acid nitrogen mineralization in relation to stand type, stand age and soil texture in the boreal mixedwood. Soil Biol. Biochem. 32, 1079–1090. Duffy, J.E., Godwin, C.M., Cardinale, B.J., 2017. Biodiversity effects in the wild are common and as strong as key drivers of productivity. Nature 549, 261. Edgar, C.B., Burk, T.E., 2001. Productivity of aspen forests in northeastern Minnesota, USA, as related to stand composition and canopy structure. Can. J. For. Res. 31, 1019–1029. Environment Canada, 2019. Canadian Climate Normals 1981–2010. http://climate. weather.gc.ca/climate_normals/index_e.html (accessed on: 6 Mar 2019). Fichtner, A., Härdtle, W., Li, Y., Bruelheide, H., Kunz, M., von Oheimb, G., 2017. From competition to facilitation: how tree species respond to neighbourhood diversity. Ecol. Lett. 20, 892–900. Forrester, D.I., 2014. The spatial and temporal dynamics of species interactions in mixed–species forests: from pattern to process. For. Ecol. Manage. 312, 282–292. Frelich, L.E., Reich, P.B., 1995. Spatial patterns and succession in a Minnesota southern boreal forest. Ecol. Monogr. 65, 325–346. Fridley, J.D., 2001. The influence of species diversity on ecosystem productivity: how, where, and why? Oikos 93, 514–526. Gao, B., Taylor, A.R., Searle, E.B., Kumar, P., Ma, Z., Hume, A.M., Chen, H.Y.H., 2018. Carbon Storage Declines in Old Boreal Forests Irrespective of Succession Pathway. Ecosystems 21, 1168–1182. Glenn-Lewin, D.C., Peet, R.K., Veblen, T.T., 1992. Plant Succession: Theory and Prediction. Chapman & Hall, New York, NY. Goulden, M.L., McMillan, A.M.S., Winston, G.C., Rocha, A.V., Manies, K.L., Harden, J.W., Bond-Lamberty, B.P., 2011. Patterns of NPP, GPP, respiration, and NEP during boreal forest succession. Glob. Change Biol. 17, 855–871. Guo, Q., 2003. Temporal species richness biomass relationships along successional gradients. J. Veg. Sci. 14, 121–128. Hart, S.A., Chen, H.Y., 2008. Fire, logging, and overstory affect understory abundance,

9

Forest Ecology and Management 455 (2020) 117641

A.R. Taylor, et al. Senici, D., Chen, H.Y.H., Bergeron, Y., Cyr, D., 2010. Spatiotemporal variations of fire frequency in central boreal forest. Ecosystems 13, 1227–1238. Soil Classification Working Group, 1998. The Canadian System of Soil Classification. NRC Research Press, Ottawa, Ontario, Canada. Taylor, K.C., Arnup, R.W., Meredith, M.P., Parton, W.J., Nieppola, J., 2000. A Field Guide to Forest Ecosystems of Northeastern Ontario. NEST Field Guide FG–01. Northeast Science and Technology, Ontario Ministry of Natural Resources, Timmins, Ontario, Canada. Taylor, A.R., Seedre, M., Brassard, B., Chen, H.Y.H., 2014. Decline in net ecosystem productivity following canopy transition to late-succession forests. Ecosystems 17, 778–791. Taylor, A.R., Chen, H.Y.H., 2011. Multiple successional pathways of boreal forest stands in central Canada. Ecography 34, 208–219. Tilman, D., Lehman, C.L., Thomson, K.T., 1997. Plant diversity and ecosystem productivity: theoretical considerations. Proc. Natl. Acad. Sci. 94, 1857–1861. Vasiliauskas, S., Chen, H.Y.H., 2002. How long do trees take to reach breast height after fire in northeastern Ontario? Can. J. For. Res. 32, 1889–1892. Vilà, M., Vayreda, J., Gracia, C., Ibáñez, J.J., 2003. Does tree diversity increase wood

production in pine forests? Oecologia 135, 299–303. Villéger, S., Mason, N.W., Mouillot, D., 2008. New multidimensional functional diversity indices for a multifaceted framework in functional ecology. Ecology 89, 2290–2301. Walker, L.R., Wardle, D.A., Bardgett, R.D., Clarkson, B.D., 2010. The use of chronosequences in studies of ecological succession and soil development. J. Ecol. 98, 725–736. White, P.S., 1979. Pattern, process, and natural disturbance in vegetation. Botanical Rev. 45, 229–299. Whittaker, R.H., 1970. Communities and Ecosystems, second ed. MacMillan Publishing Co., Inc., New York, NY. Williams, L.J., Paquette, A., Cavender-Bares, J., Messier, C., Reich, P.B., 2017. Spatial complementarity in tree crowns explains overyielding in species mixtures. Nat. Ecol. Evol. 1, 63. Zhang, Y., Chen, H.Y.H., Reich, P.B., 2012. Forest productivity increases with evenness, species richness and trait variation: a global meta–analysis. J. Ecol. 100, 742–749. Zhang, Y., Chen, H.Y.H., Taylor, A.R., 2016. Aboveground biomass of understorey vegetation has a negligible or negative association with overstorey tree species diversity in natural forests. Glob. Ecol. Biogeogr. 25, 141–150.

10