How does understory vegetation diversity and composition differ between monocultures and mixed plantations of hybrid poplar and spruce?

Although monocultures are important for timber production, they are often associated with lower biological diversity than mixtures. Thus, mixed plantations have been suggested as a way to enhance biodiversity because of their inherent compositional diversity. However, the effects of monocultures versus mixtures on understory diversity and composition can vary in different ecosystems. The objective of this study was to assess how monocultures and mixed plantations influence understory vegetation diversity and composition in the boreal forest region of southern Quebec. We sampled plantations established with deciduous Populus trichocarpa Torrey & A. Gray × balsamifera L. and P. maximowiczii Henry × balsamifera L. and coniferous Picea abies (L.) Karst. and Picea glauca (Moench) species planted in monocultures and in mixed plantations on abandoned farmlands and a forest site. We assessed understory vegetation diversity and composition in each canopy type (coniferous, deciduous, mixed) and in each plantation type. We evaluated bryophyte and lichen diversity and composition specifically in tree microhabitats: soil, tree bases, and tree trunks. We found that vascular plant and lichen species richness was similar in all plantation types, while bryophyte species richness was higher in spruce monocultures and in mixed plantations compared to poplar monocultures. Our results also highlight how land-use history influenced vascular plant composition as abandoned farmland sites were composed of more ruderal vascular plants, while the previously forested site was composed of species found in natural forests. In the context of reforestation and plantations, our study suggests mixing spruce with poplars to maximize understory vegetation diversity as the addition of spruce in mixed plantations promoted the establishment of terrestrial bryophytes, while poplars favored the establishment of epiphytic lichens.


Introduction
Understory vegetation communities including vascular plants, bryophytes, and lichens contribute to ecological services in forests, such as carbon sequestration, nutrient cycling, and organic matter decomposition (Nilsson and Wardle, 2005;Fenton and Bergeron, 2006;Grau-Andrés et al., 2022).Establishment of these species in a forest or plantation depends on microclimatic conditions and microhabitat availability.Microclimatic conditions are modified by factors such as soil nutrient availability, pH, and light availability (Anderson et al., 1969;Porté et al., 2004;Reich et al., 2012;Augusto et al., 2015), which are influenced by the overstory composition (Loreau, 2000;Gamfeldt et al., 2013;Stefańska-Krzaczek et al., 2022;Wang et al., 2022).For example, deciduous forests usually have a nutrient-rich broadleaf litter, greater light availability, and warmer soils than in coniferous forests (Barbier et al., 2008;Augusto et al., 2015).Greater light availability and warmer soils found under deciduous trees can stimulate organic matter decomposition (Laganière et al., 2010).These conditions are favorable for the establishment of vascular plants (Saetre et al., 1997;Barbier et al., 2008;Hart and Chen, 2008), while bryophyte establishment is constrained by the fall of broadleaf litter (Jean et al., 2017;Bartels et al., 2018).Unlike deciduous forests, coniferous forests have nutrient-poor needleleaf litter that can acidify the soil and have lower light availability and soil temperature (Vance and Chapin, 2001;Laganière et al., 2010) that can further limit organic matter decomposition rates which accumulates and increases soil insulation and limit moisture evaporation (Facelli and Pickett, 1991;Millar, 2012).These conditions are more favorable to bryophyte establishment (Saetre et al., 1997;Fenton et al., 2005;Hart and Chen, 2008) than that of vascular plants.The occurrence of lichen species depends on light availability and on the physical and chemical properties of the bark of trees (Jüriado et al., 2009;Randlane et al., 2017).In addition to microclimatic conditions, forest bryophytes and lichens establishment also depends on the availability of different types of microhabitats (Cole et al., 2008;Rubio-Salcedo et al., 2015), such as soil, deadwood, or trees (During, 1992;Haughian, 2018;Barbé et al., 2020).Some species are associated with microhabitats created by coniferous forests while others are associated with those of deciduous forests (Caners et al., 2013;Barbé et al., 2020).Thus, the distinct environmental conditions and microhabitats created by different tree species composing the overstory are key to understanding the biodiversity patterns of understory vegetation diversity and composition in forests.
Because the overstory composition acts as a filter for understory vegetation communities (Kutnar et al., 2023), intensively managed forest plantations of a single tree species (monocultures) provide little to no variability in understory microclimate or microhabitats, and therefore do not allow establishment of diverse understory communities, making them "biological deserts" (Brockerhoff et al., 2008).Monoculture plantations have lower levels of biodiversity than surrounding native forests (Bremer and Farley, 2010;Brockerhoff et al., 2013).Generally, forest plantations are dominated by a single tree species, and they are managed as even-aged monocultures with an equal spacing between trees (Puettmann et al., 2015;Liu et al., 2018).This management practice results in structurally simplified stands that fail to provide suitable habitats for forest species (Bauhus et al., 2010;Liu et al., 2018;Messier et al., 2022).Consequently, intensively managed forest plantations are perceived as a threat to biodiversity because their use is predicted to grow 37% by 2050 (FAO, 2022).
In contrast, mixed plantations enhance levels of biodiversity compared to monoculture plantations (Hartley, 2002;Nichols et al., 2006;Jonsson et al., 2019) due to their compositional and structural diversity (Saetre et al., 1997;Messier et al., 2022).A higher diversity of tree species provides more heterogeneous habitats (Carnus et al., 2006;Brockerhoff et al., 2008), that in turn can support a larger number of associated species (Thompson et al., 2009;Nadrowski et al., 2010;Feng et al., 2022) such as the understory vegetation (Ampoorter et al., 2020;Guo et al., 2021;Messier et al., 2022).Tree species differ in their growth patterns, canopy structures, leaf litter and bark characteristics.This diversity results in formation of various microhabitats within the forest, including different levels of shade, nutrient availability, and moisture (Tews et al., 2004;Puettmann, 2011).Consequently, mixed plantations can provide diverse microhabitats and appropriate abiotic conditions leading to the presence and coexistence of more species (Thompson et al., 2009).
However, for both mixed forests and mixed plantations, there is conflicting evidence on their benefits to understory vegetation diversity compared to monospecific forests.Some authors found that mixed forests had a similar understory vegetation diversity as monospecific forests (Augusto et al., 2003;Barsoum et al., 2016), while others reported that they had higher understory diversity than monospecific forests (Gamfeldt et al., 2013;Wang et al., 2022).Other studies reported that mixed forest understories were compositionally intermediate between deciduous and coniferous stands (Hart and Chen, 2008).This lack of evidence is related to the difficulty in disentangling overstory composition from confounding factors such as the age, canopy openness, soil, or topography (Pitkänen, 2000;Gärtner and Reif, 2004;Nadrowski et al., 2010).Furthermore, studies on understory vegetation biodiversity in plantations did not consider the complete understory vegetation (vascular plants, bryophytes, and lichens) as they often focused on only one taxonomic group such as vascular plants (Soo et al., 2009), bryophytes (Pharo and Lindenmayer, 2009) or lichens (Hilmo et al., 2009).Despite the well-documented effects of mixed plantations on the understory vegetation diversity (Messier et al., 2022;Wang et al., 2022), the effects of tree species identity and combinations on the overall forest species diversity and composition are not fully understood, specifically, in the boreal forest region of southern Quebec where mixed plantations are less developed than monoculture plantations.
It is widely recognized that microhabitat availability has an important influence over bryophyte and lichen establishment (Cole et al., 2008;Király and Ódor, 2010;Haughian, 2018;Hämäläinen et al., 2023).Microhabitats offer a stable substrate that can provide variations in temperature and moisture levels for bryophytes and lichens.The various species composing mixed plantations could provide heterogenous microhabitats (e.g., soils, tree bases, and tree trunks) that may enhance the diversity of bryophyte and lichen specific to each overstory tree species.However, bryophyte and lichen diversity and composition in different microhabitats in monoculture and mixed plantations have not previously been explored in the boreal forest region of southern Quebec.Thus, a better understanding of how tree composition affects understory vegetation diversity and composition related to the microhabitat availability should help guide plantation management to favor plant diversity.
The main objective of this study was to evaluate the understory vegetation (including vascular plants, bryophytes and lichens) diversity and composition in mixed and monoculture plantations of hybrid poplar (Populus spp.) and spruce (Picea spp.) species.We also evaluated whether the effect of monocultures or mixed plantations on bryophyte and lichen richness and composition varied among three microhabitats: soil, tree bases and tree trunks.We expected to have more vascular plant and epiphytic lichen species in hybrid poplar and mixed plantations compared to spruce plantations because mixed plantations provide more light availability for understory vascular plants and many lichen species are specialized to be epiphytes on deciduous trees only.Conversely, we expected bryophyte diversity to be higher in spruce plantations compared to hybrid poplar and mixed plantations.We expected that the richness of bryophytes, and lichens on soil, tree bases and tree trunks would be greater in mixed plantations compared to monoculture plantations.In terms of composition, we expected that vascular plants would be more dominant in hybrid poplar than in spruce plantations and that bryophytes would be more present in spruce than in hybrid poplar plantations because deciduous forests usually have a nutrient-rich deciduous broadleaf litter, greater light availability, and warmer soils than in coniferous forests which are favorable to the establishment of vascular plants.We expected that lichens would be more dominant in hybrid poplar than in spruce plantations because Populus species provide a habitat for a wide range of epiphytic lichens due to their non-acidic bark and nutrient-rich substrate conditions.However, coniferous forests have a low-nutrient needleleaf litter that acidifies the soil and have lower light availability and soil temperature which are more favorable to bryophyte establishment than that of vascular plants.Lastly, we hypothesized that mixed plantations would be compositionally intermediate between spruce and hybrid poplar plantations.

Study sites
Three plantations were established in 2003 near the localities of Amos (48 • 36′N, 78 • 04′W), Rivière-Héva (48 • 1′N, 78 • 16′W) and Nédelec (47 • 45′N, 79 • 22′W) (Abitibi-Témiscamingue, QC, Canada) (Fig. 1a).Prior to planting, our studied sites had different land-use histories: the Amos and Rivière-Héva sites were abandoned farmland, while the Nédelec site was a forest site previously dominated by trembling aspen (Populus tremuloides), pin cherry (Prunus pensylvanica L.) and paper birch (Betula papyrifera Marsh.)forests that had been harvested in 2000.The Amos and Rivère-Héva plantations were in a matrix of abandoned farmland, mostly reforested with conifers.The Nédelec plantations were in a forested matrix but surrounded by clearcuts reforested with conifers and hybrid poplars.The soils of the Amos and Rivière-Héva sites were classified as heavy clay soils while soils at the Nédelec site had a sandyloam soil texture.Based on the 30-year climate average , mean annual precipitation and temperature at the Amos and Rivière-Héva sites were respectively 929 mm and 1.5 • C, while, they were 904 mm and 2.2 • C at the Nédelec site (Environnement Canada, 2021).We used the same mean annual precipitation and temperature values in the Rivière-Héva site as those in the Amos site because there was no meteorological station specific to Rivière-Héva.
Site preparation at the three sites was carried out in 2002 to remove pre-existing vegetation and tree stumps.At the Amos site, chains and a farm tractor were used to remove shrubs and scattered tree stumps, while a brush shredder was used to remove shrubby vegetation at the Rivière-Héva site.At the Nédelec site, a bulldozer was used to remove the tree stumps and large debris.Soils of the three sites were then plowed to a depth of 30 cm in the fall of 2002 and harrowed in the spring of 2003 to level the soil just before planting rooted hybrid poplar cuttings and spruce seedlings.During the first five years, competitive vegetation was removed annually mechanically by cultivating between trees and rows with a small tractor and discs.

Vegetation surveys
During the spring of 2021, we sampled six or nine 1 m 2 quadrats within each treatment (n = 6 for monoculture treatment; n = 9 for mixed plantation treatment).One plantation type (e.g., poplar1) could be adjacent to one or two other types (e.g., poplar2 and NS:poplar1).This condition might act as potential edge effects that our studied plantations could experience.Consequently, we systematically placed three 1 m 2 quadrats (no tree boles included) spaced 4.5 m apart in a linear transect established in the center of each plantation type to avoid edge effects (Fig. 1b).Within each monoculture plantation, we placed three other 1 m 2 quadrats under trees along a diagonal line.Within mixed plantations, we placed three 1 m 2 quadrats under each species of tree along diagonal lines (total = six).In each quadrat, we inventoried all woody plants<1.3m tall (shrubs and tree seedlings), herbaceous plants (grasses, forbs and ferns), bryophytes and lichens.We expressed the abundance of each vascular plant species as a percent cover per quadrat.We collected bryophytes and lichens in three microhabitats: the soil, the base of trees (0-50 cm) and the trunk of trees (50-150 cm).We did not sample bryophytes and lichens on woody debris or large shrubs because they were absent or minimal in plantations because of the initial mechanical weed control that removed them.We placed all sampled bryophyte and lichen specimens in individual bags and brought to the laboratory for identification under a binocular microscope.We based vascular plant nomenclature on the VASCAN database (Brouillet et al., 2010).We followed Faubert (2012) for bryophyte nomenclature and Brodo et al. (2001), Hinds et al. (2007) and Brodo (2016) for lichen nomenclature.

Light measurement
Light is one of abiotic variables that depends on the overstory composition and may affect the abundance and presence of understory species (Anderson et al., 1969;Porté et al., 2004).The distinct variability in light created by tree species composing the overstory is likely key to understanding the biodiversity patterns of the understory vegetation.
We measured percent incident light 50 cm above the forest floor using a LAI-2200C plant canopy analyzer (LI-COR Biosciences, Lincoln, NE).We measured incident light in each monoculture above the three quadrats along a diagonal line and above six quadrats in mixed plantations (three under poplar trees and three under spruce trees).We also took reference full light measures in an open field near each treatment.We calculated incident light within the plantations as the ratio of the light measured at 50 cm above the forest floor and the reference light.

Taxonomic diversity
We conducted all statistical analyses using R 4.1.0(2021-05-18) (R Core Team, 2021).We assessed at two scales: first, we focused on the effect of canopy type, i.e., deciduous, mixed, or coniferous to describe general patterns and to better link to the literature.Secondly, to examine the effect of specific clones and species in the plantations, we examined the effect of plantation type.
To evaluate the effect of canopy type (i.e., deciduous, coniferous, and mixed) on species diversity of the understory vegetation, we used Hill numbers which represent a statistically rigorous alternative to other diversity indices as they are expressed in units of effective numbers of species (Chao et al., 2014).We quantified three Hill numbers which were the effective number or number of species (q0), exponential of Shannon or number of frequent species (q1), and Simpson's inverse or number of dominant species (q2) using the «iNEXT» R package (Hsieh et al., 2016).To examine the accumulative and rarefaction of species diversity, we used accumulation curves to represent the cumulative number of species as a function of the number of quadrats sampled.We then compared Hill numbers between canopies using sample-size-based rarefaction and extrapolation (R/E) curves with 95% confidence intervals (Chao et al., 2014).By standardizing samples depending on sample size, integrated curves allow to compare biodiversity data (Chao et al., 2014).
To evaluate the effect of plantation type on species richness, we compared taxonomic diversity between plantation types (i.e., poplar1, poplar2, NS, WS, NS:poplar1, WS:poplar1, NS:poplar2, and WS:poplar2) using GLMMs from the lme4 package (Bates et al., 2014) for all taxonomic groups combined (i.e., vascular plants, bryophytes, and lichens) and for each taxonomic group separately.The richness referred to the mean richness across all the quadrats sampled in each plantation type (i.e., before the extrapolation curve).We fitted plantation type as a fixed effect for all models while treatment was nested as a random effect within site which was also a random effect.We checked the estimates of residual deviance and residual degrees of freedom using the function «overdisp_fun» and if it fitted the null hypothesis of equidispersion, we used a Poisson distribution for GLMM; if not, we applied Negative-Binomial distribution to fit over-dispersion.We analyzed all models (except for vascular plant richness) with a negative binomial distribution instead of Poisson distribution because we detected overdispersion.We tested significance of the predictors with a type II Wald chi-square (χ2) tests with the «Anova» function in the car package (Fox and Weisberg, 2019).Where categorical explanatory variables had a significant effect, we applied Tukey's multiple comparison tests with the package emmeans (Russell, 2022).We considered a p value ≤ 0.05 as significant.
We also examined bryophyte and lichen richness in each microhabitat in each plantation type.We expressed richness as the mean number of species found in each microhabitat (soil, base of trees, trunk of trees) within the quadrats of each plantation.We followed the same procedures as in the previous description to analyze the richness by microhabitat.

Community composition
We tested the differences in community composition among plantation types using permutational multivariate analysis of variance (PERMANOVA) with 999 permutations with the «adonis» function from the vegan package (Oksanen et al., 2015).We performed a principal coordinates analysis (PCoA) (Borcard et al., 2011) with Jaccard's dissimilarity for all 3 taxonomic groups combined.We then evaluated the composition of each group separately, and we performed a PCoA with Bray-Curtis dissimilarity matrices as a distance measure for vascular plants (abundance data) and a PCoA with Jaccard's dissimilarity for bryophytes and lichens combined (presence/absence data).We used the PCoA function in the package vegan 2.6-2 (Oksanen et al., 2015) with the Cailliez correction (Gower and Legendre, 1986) to correct negative eigenvalues.We used the vegan «envfit» function to fit species vectors to ordinations from 999 permutations.We used the function «ordiselect» from the goeveg package (Goral et al., 2018) to show the most relevant and influential species in ordination diagrams and we only represented significant species in all ordinations (p value ≤ 0.05).We also examined differences in overall community composition between the land-use histories of sites using PERMANOVA to explain our results.We also analyzed the bryophyte and lichen community composition in each microhabitat type per plantation type.We assessed the composition of the bryophyte and lichen communities together because lichens had low occurrence and abundance in our plantations.We followed the same procedures as in the previous description to analyze the composition by microhabitat.

Effect of plantation type on incident light
We used Linear mixed models (LMM) from the package nlme (Pinheiro et al., 2021) to test the effect of plantation type on incident light at 50 cm above the forest floor.We took the average of light measurements within each 15 × 15 m plantation as the response variable.We verified the independence of residuals, homogeneity of variances, and normality of residuals by diagnostic graphs.We applied log-transformations of incident light to achieve normality.We used the post hoc emmeans method to make pairwise comparisons between plantation types with the package emmeans (Russell, 2022).
Sample-size-based rarefaction and extrapolation (R/E) curves showed differences among the canopy types and deciduous plantations had the lowest diversity (Hill numbers) (Fig. 2).Coniferous and mixed plantations were more diverse, and their confidence interval curves overlapped for all Hill numbers (Fig. 2) indicating that the number of species (q0), the number of frequent species (q1), and the number of dominant species (q2) were similar between the two canopy types.Although the confidence intervals overlapped, coniferous plantations were more diverse for Simpson diversity (q = 2) indicating that they had a higher number of dominant species than mixed plantations (Fig. 2).
We did not find low slopes towards the end for q0 in mixed plantations indicating that they could still harbor a higher species richness with an increasing number of quadrats.In contrast, plateaus were reached for coniferous and deciduous plantations in q0.Shannon (q1) and Simpson (q2) diversity curves leveled off for all canopy types.

Effects of plantation type on understory vegetation richness
Mean total species (i.e., all taxonomic groups) richness differed between poplar1 compared to WS:poplar1, WS:poplar2, NS and WS (Fig. 3a).The mean total species richness in Poplar1 plantations increased when this clone was mixed with white spruce.Vascular plant species richness did not vary among plantation types (Fig. 3b) (ANOVA, p = 0.22; see supplementary information Table S1).However, bryophyte species richness differed between plantation types (ANOVA, p < 0.001; see supplementary information Table S1) as it was higher in spruce monocultures (NS and WS) than in hybrid poplar monocultures (poplar1 and poplar2; Fig. 3c).All mixed plantations had similar bryophyte species richness as spruce monocultures except for NS:poplar1 which was lower than WS (Fig. 3c).There was a difference between NS and WS for bryophytes in mixed plantations, where WS:poplar appeared to be more effective than NS:poplar at harbouring bryophyte species.Lichen species richness was similar in all plantation types (Fig. 3d) (ANOVA, p = 0.68; see supplementary information Table S1).

Understory vegetation community composition
The understory community composition varied with plantation type and with land-use history (PERMANOVA; p = 0.001; Table 1a and  Table 1b).The PCoA on the total understory vegetation community (all taxonomic groups) and on the vascular plant community showed that there was a gradient from coniferous to deciduous plantations along the second axis (7.39 and 6.24% of variance respectively) with overlapping ellipses (Fig. 4a and Fig. 4b).Mixed plantations were compositionally intermediate between coniferous and deciduous plantations (Fig. 4a and Fig. 4b).The first axis of the PCoA (10.91% and 13.40% of variance respectively) was related to the land-use history of sites (Fig. 5a and Fig. 5b).The land-use history of sites separated along axis 1 showing that abandoned farmland sites (Amos and Rivière-Héva) tended to be associated with species that classified as ruderal vascular plants i.e., that Fig. 2. Species diversity (including all 3 taxonomic groups combined: vascular plants, bryophytes, and lichens) for Hill numbers of species (q = 0), exponential of Shannon (q = 1) and Simpson's inverse (q = 2).Solid lines: interpolation/rarefaction; Dashed curves: Extrapolation.The symbols indicate the observed Hill numbers (q = 0, 1, 2) for each canopy type.Canopy types: coniferous in orange, deciduous in blue and mixed in green.can colonize and grow rapidly after the plantation establishment such as Hieracium spp.(HIESPP), Symphyotrichum cordifolium (SYMCOR) and Vicia cracca (VICCRA) (on the left side of the ordination (Fig. 5a and Fig. 5b)).In contrast, the plantations established on the previously forest site (Nédelec) were composed of forest vascular plant species such as Diervilla lonicera (DIELON) Maianthemum canadense (MAICAN) and Vaccinium angustifolium (VACANG) (on the right side of the ordination, Fig. 5a and Fig. 5b), ruderal mosses such as Ceratodon purpureus (CER-PUR) and Pohlia nutans (POHNUT) (on the right side of the ordination, (Fig. 5a)) and lichens such as Evernia mesomorpha (EVEMES), Hypogymnia physodes (HYPPHY) and Parmelia sulcata (PARSUL) (on the right side of the ordination, (Fig. 5a)).
The bryophyte and lichen community composition, as represented by PCoA, shifted from coniferous to deciduous plantations along the first PCoA axis (10.84% of variance) and significant differences were found for plantation type (PERMANOVA; p = 0.001; Table 1A), although there was significant overlap of the ellipses (Fig. 4c).The part of the variance explained by the second axis was difficult to interpret (7.52% of variance).There was dissimilarity among canopy types in bryophyte and lichen composition reflected by the presence of mosses in coniferous plantations while epiphytic lichens were present in deciduous plantations.Indeed, coniferous plantations were associated with many mosses such as Bracythecium acutum (BRAACT), B. campestre (BRACAM), B. curtum (BRACUR).B.falcatum (BRAFAL), Pleurozium schreberi (PLESCH) and Sanionia uncinata (SANUNC) (on the left side of the ordination (Fig. 4c)) while deciduous plantations (especially poplar2), were associated with lichen species such as Evernia mesomorpha (EVEMES), Bryoria fuscescens (BRYFUS), Physcia aipolia (PHYAIP) and Usnea spp.(USNSPP) (on the right side of the ordination (Fig. 4c)).In mixed plantations, the bryophyte and lichen community composition was generally intermediate between coniferous and deciduous plantations (Fig. 4c).

Effect of plantation type on bryophyte and lichen richness and composition according to microhabitats
Bryophyte and lichen richness was affected by plantation type only for the soil microhabitat where deciduous plantations had less species than in almost all other plantation types (except in NS:poplar1) (Fig. 6).Bryophyte and lichen community composition on the soil varied by plantation type (PERMANOVA; p = 0.001; Table 1c).There was a gradient from coniferous plantations to mixed plantations along the first axis (10.60% of variance) with overlapped ellipses while the deciduous plantations were explained by the second axis (9.20% of variance) (Fig. 7a).Soil in coniferous plantations was dominated by pleurocarpous mosses such as Bracythecium acutum (BRAACT), B. curtum (BRACUR), Pleurozium schreberi (PLESCH) (on the left side of the ordination (Fig. 7a)) while deciduous plantations, especially poplar2 were dominated by acrocarpous mosses such as Ceratodon purpureus (CERPUR) (on the right side of the ordination (Fig. 7a)).Composition of the bryophyte community on the soil in mixed plantations was generally more similar to that of coniferous than to deciduous plantations (Fig. 7a).Although there was significant overlap of the ellipses and non-significant effect of plantation type (PERMANOVA; p = 0.35; Table 1c) on the bryophyte and lichen community composition at the base of trees, Platygyrium repens (PLAREP) was associated with the base of hybrid poplar trees while Callicladium haldanianum (CALHAL) was associated with the base of spruce trees (Fig. 7b).In mixed plantations, bryophytes and lichens appeared to be compositionally intermediate between coniferous and deciduous plantations at the base of trees (Fig. 7b).The bryophyte and lichen community composition on the trunk of trees was significantly different (PERMANOVA; p = 0.02; Table 1c) between forest types in the first axis of the PCoA ordination (28.98% of the variance).Trunks of trees were mostly composed of lichens which were more present in hybrid poplar and mixed plantations (Fig. 7c).

Incident light
Incident light was different between plantation types (ANOVA, p < 0.001; see supplementary information Table S2) as it was significantly greater in hybrid poplar compared to spruce plantations (Fig. 8).NS and WS had the lowest incident light values but were not significantly different from the mixed plantations (Fig. 8).Incident light was similar between mixed and deciduous plantations except for WS:poplar1 which was significantly lower than poplar1(Fig.8).

Diversity responses to canopy and plantation types
Mixed and coniferous plantations contained similar understory vegetation diversity and they had greater Hill numbers than deciduous plantations.These results partially support our hypothesis and indicate that the co-occurrence of hybrid poplar and spruce could increase the diversity of the understory vegetation compared to hybrid poplar monocultures.The most important factors affecting the diversity of forest-floor species, especially bryophytes, are the availability of potential substrates and microhabitats (Mills and Macdonald, 2005;Haughian, 2018).The possible explanation for the similarity of the diversity in coniferous and mixed plantations is that the addition of spruce in mixed plantations created microhabitats and conditions which were favorable for the establishment of species found in coniferous plantations.However, our data suggests that mixed plantation richness could increase with the number of quadrats (Fig. 2), indicating that mixed plantations could harbor more species than monocultures.
Conditions created by deciduous forests are normally more favorable to the establishment of vascular plants than coniferous forests (Barbier et al., 2008;Augusto et al., 2015), as deciduous forests have warmer, more nutrient-rich soils and greater light availability, which increase the rate of litter decomposition (Barbier et al., 2008;Laganière et al., 2010;Cavard et al., 2011).However, our study showed that vascular plant richness was similar in all plantation types which did not confirm our hypothesis.This finding could be due to the presence of seeds of some species that can persist and remain dormant for up to 100 years in the soil (Whittle et al., 1997) allowing for a fast recolonization, even after a heavy site-preparation and plantation establishment (De Grandpre and Bergeron, 1997;Greene et al., 1999;Ramovs and Roberts, 2005;Soo et al., 2009).The understory layer created by ruderal species can delay the rate of successional change by creating conditions on the ground that inhibit the growth and emergence of many species (Mallik, 2003;Royo and Carson, 2006).Thus, our sites previously dominated by ruderal vascular plants prior to the plantation were probably rapidly recolonized by species present before the plantation that may have buffered the effect of plantation type on vascular plant diversity.
Bryophyte richness was plantation type sensitive as it was higher in spruce and mixed plantations and lower in hybrid poplar plantations, which partially supports our hypothesis since we expected to have more

Table 1
Effect of a) plantation type and b) site on the composition of total vegetation, vascular plants, bryophytes, and lichens using the permutational multivariate analysis of variance (PERMANOVA) based on Jaccard (total, bryophytes, and lichens) and Bray-Curtis (vascular plant) transformed community data.c) Effect of plantation type on the composition of bryophytes, and lichens on soil, at base of trees and on trunk of trees using the permutational multivariate analysis of variance (PERMANOVA) based on the Jaccard dissimilarity index.bryophyte species only in spruce plantations.The low richness of bryophyte species in hybrid poplar plantations could be explained by their inability to grow in the presence of broadleaf litter (Jean et al., 2017;Bartels et al., 2018) due to bryophytes specializing on the acidic and low-light conditions below coniferous canopies (Barbier et al., 2008;Tullus et al., 2015).Recalcitrant needleleaf litter associated with low decomposition rates in spruce plantations (Chomel et al., 2015) had probably produced favorable conditions for bryophytes.Moreover, coniferous plantations created overall more shaded conditions than those found in deciduous plantations which could decrease soil temperatures and decomposition rates of the organic matter leading to a high soil insulation and a cool ground surface typical of coniferous forests (Crawford et al., 2003;Laganière et al., 2010;Millar, 2012;Barbé et al., 2020).These conditions have created suitable microclimate conditions for bryophyte establishment in coniferous plantations.Combined with the microhabitat results, we found that bryophytes growing on the soil were sensitive to plantation type with less species in hybrid poplar plantations than in all other plantation types.Tree bases and tree trunks are in a vertical posture, which may make them less exposed to changes in abiotic factors such as light, litter accumulation, soil moisture, and pH.This indicates that soils are probably more exposed than tree bases and tree trunks to differing abiotic environments caused by canopy composition that mainly affect soil bryophyte communities.The addition of spruce in mixed plantations might have changed the physical  S3 for more details on species names.
properties of litter layers by diluting the quantity of hybrid poplar litter with a slower decomposition rate of spruce litter (Chomel et al., 2015;Chomel et al., 2016) allowing the establishment of bryophytes on the soil.The presence of spruce in mixed plantations also generated incident light levels similar to those found in coniferous plantations which also promoted shade-tolerant species such as bryophytes.
The similarity of lichen richness combined with the low presence of lichens on the soil in all plantation types are probably due to the absence of available free substrate (Randlane et al., 2017) as lichens can not generally compete with ruderal vascular plants and vigorous bryophytes growing on soil microhabitats (Gilbert, 1993).

Community composition
As hypothesized, we found that mixed plantations were compositionally intermediate between coniferous and deciduous plantation types.This indicates that understory vegetation communities in mixed plantations were composed of the same species that were found in hybrid poplar and spruce monocultures.In our study, spruce plantations were associated with bryophytes on the soil while some epiphytic lichens preferred hybrid poplar rather than spruce plantations.Several factors such as available understory light, water-holding capacity, pH and roughness of bark are important for the establishment and growth of  S3 for more details on species names.bryophytes and lichens in forests (Lõhmus and Lõhmus, 2008;Hilmo et al., 2009).Populus species are a substrate for a great number of epiphytic lichens as they possess non-acidic bark and nutrient-rich substrate conditions (Sheard and Jonescu, 1974;Boudreault et al., 2000;Jüriado et al., 2003;Hämäläinen et al., 2023).Thus, this probably favored the occurrence of epiphytic lichens in hybrid poplar (e.g., Evernia mesomorpha (EVEMES), Bryoria fuscescens (BRYFUS) and Physcia aipolia (PHYAIP)).Our results suggest that the dissimilarity in the microhabitats generated by hybrid poplar and spruce promoted the establishment of both terrestrial bryophytes (via spruce) and epiphytic lichens (via hybrid poplar) in mixed plantations.
The composition of understory vegetation within plantations is not only affected by the canopy type and composition, it can also be influenced by the silvicultural practices that are applied prior to the plantation establishment and the land-use history (Haeussler et al., 2002;Ross-Davis and Frego, 2002;Brudvig and Damschen, 2011).While both bryophytes and vascular plants can exhibit sensitivity to previous landuse (Ross-Davis and Frego, 2002;Soo et al., 2009;Bremer and Farley, 2010), bryophytes are generally considered to be more sensitive to abiotic changes because unlike vascular plants, bryophytes have no diverse life strategies or vascular tissues for water transport making them sensitive to changing environmental conditions (Bates et al., 2005;Becker Scarpitta et al., 2017).However, our study revealed that vascular plant composition was more sensitive than bryophyte composition to the land-use history.Abandoned farmland sites (Amos and Rivière-Héva) were dominated by ruderal and vascular plants that are disturbance tolerant, while the forest site (Nédelec) was composed of species usually found in natural forests.Our findings concur with those of Ramovs and Roberts (2005) who reported that the understory vegetation in plantations established on abandoned farmland were mostly species found in disturbed and non-forest habitats.This indicates that many plant species composition still reflected the agricultural environment despite the presence of a forest cover in the study area (Bellemare et al., 2002).Recruitment of forest species in the forest site and recolonization of the newly established habitats offered by plantations might be related to the presence of seed banks from the previous forest (Caners et al., 2009).The open canopy and the disturbed area after plantation establishment in forest sites (Nédelec) might have provided opportunities for ruderal and pioneer moss species such as Ceratodon purpureus (CERPUR) and Pohlia nutans (POHNUT) to establish.These species have been reported as widespread and ruderal moss species which are commonly found in harsh and disturbed habitats (Shaw and Beer, 1999;Rydgren et al., 2004;Ochyra et al., 2008) and they may be persistent for many decades after disturbance (Jonsson and Esseen, 1990).
Because the overstory changes affect the understory vegetation, we can expect that when plantations will become older, the lightdemanding and ruderal species will be replaced by shade-tolerant species (Baum et al., 2009).The older the trees will become, the more shade levels increase, leading to a high number of forest species throughout succession.

Tree identity matters
Though not significant, there was a trend that poplar2 appeared to contain greater total understory vegetation (Fig. 3) and bryophytes and lichens on soil and trees (Fig. 6) than in poplar1 plantations.The understory species richness of poplar1 was also relatively low and improved when mixed with coniferous species.Terminal bud phenology showed significant differences among poplar clones in terms of bud break time, and dormancy period as poplar2 had a longer growing season duration, broke buds earlier and had an earlier onset of dormancy than poplar1 (Elferjani et al., 2016).Thus, the seasonal  S3 for more details on species names.variations in the canopy associated with leaf budding and defoliation in hybrid poplar trees could affect understory vegetation by influencing abiotic environments such as incident light.Our study focused on current incident light, however, the light-vegetation relationship is likely to be better explained by the light during the whole growing season created by these phenological traits than by current light conditions (Thomas et al., 1999).This might explain a greater number of species as poplar2 plantations created a forest environment (canopy closure) longer and earlier in the year than poplar1.Poplar2 grew faster than poplar1 and leaf nitrogen (N) of poplar2 was greater in mixture than in monocultures (Benomar et al., 2013), suggesting that the presence of poplar2 increased N mineralization at the soil surface and thus, may increase nutrient-loving species richness by promoting a habitat with greater Navailability compared to mixed plantings with poplar1 (e.g., NS:pop-lar1).However, nitrification can eventually result in the invasion of nutrient-loving species and competition exclusion which would decrease species diversity (Gilliam, 2006;McClean et al., 2011).The effect of these clone dissimilarities on the biodiversity of understory vegetation is not fully understood and future investigations will be required to evaluate their impacts on the understory vegetation diversity and composition.
Although this was a non-significant trend, WS appeared to harbour greater total bryophyte diversity (Fig. 3c) and bryophytes richness on soil and trees (Fig. 6).In a previous study, WS:poplar2 produced greater aboveground biomass than the other mixture and the basal diameter of NS was slightly smaller when combined with hybrid poplars (Benomar et al., 2013).This suggests that the high aboveground biomass may lead to taller trees with larger canopies, and lower light availability during the whole growing season in WS than in NS plantations.The low light availability may have limited moisture evaporation and created more suitable microclimate conditions for bryophytes in mixed plantations with WS similar to those in coniferous plantations leading to nonsignificant differences in bryophyte richness (see supplementary information Table S1).

Conclusion
While the effect of plantations on understory vegetation had been assessed, our study is among the first to compare the effect of monoculture and mixed plantations on all taxonomic groups of the understory vegetation (vascular plants, bryophytes, and lichens) in the boreal forest.This study also focused on the impact of different plantation types on bryophyte and lichen richness and composition on three microhabitats which are the soil, the base of trees, and the trunk of trees.We provide evidence that mixed plantations could harbor more bryophyte species compared to hybrid poplar plantations.The presence of multiple tree species in mixed plantations results in greater resource and habitat availability, promoting the growth of different species and increasing overall biodiversity.Short-rotation hybrid poplar plantations in mixed plantations could also provide temporary habitats for lichens and thus, contribute in part to enhance biodiversity in mixed plantations.
One of the limitations of the study is the small size of plantations with relatively few trees.The few trees sampled for epiphytic species may not have captured the full range of species diversity present in larger areas with a greater number of trees.Our studied plantations were also near boundaries, such as forest edges that could potentially act as edge effects.Our results show that land-use histories which are abandoned farmland sites (Amos and Rivière-Héva) and forest site (Nédelec) are an important determinant of future compositional differences in plantations and must be considered when assessing the impact of plantations on the understory vegetation.It could have been better if we had replicate sites for each land use history type to evaluate the overstory effects on the understory vegetation diversity and composition regardless of site origins or land-use histories considering that overstoryunderstory relationship is not always causal as it may react in parallel with site origins (Berger and Puettmann, 2000).
The lower presence of bryophytes in hybrid poplar plantations compared to spruce plantations, and the well-defined intermediate response of mixed plantations, provide justification for introducing spruce into hybrid poplar plantations to increase understory diversity when planning reforestation and plantations.

Fig. 1
Fig. 1. a) Location of the plantation sites within the province of Québec, eastern Canada.Inset the region of Abitibi-Témiscamingue.b) Schematic representation of the experimental design of monoculture and mixed plantations in the three sites (Amos, Rivière-Héva and Nédelec).Each site had eight 15x15m plantation plots (represented by squares of different colors): 4 monoculture and 4 mixed plantations.In each plantation plot, we placed six 1 m 2 quadrats within monoculture or nine within mixed plantations (as red border squares).Trees in monoculture are in light green and in mixed plantations in light and dark greens.

Fig. 3 .
Fig. 3. Mean (±SEM) richness of understory vegetation for a) total (all taxonomic groups); b) vascular plants; c) bryophytes; and d) lichens by plantation type.Different letters indicate that mean richness differed significantly between plantation types.

Fig. 4 .
Fig. 4. Principal coordinate analysis (PCoA) plots based on a) total vegetation community composition; b) ascular plant community composition; and c) bryophyte and lichen community composition in each canopy (ellipses) and plantation type (symbols).The ellipses are the 95% confidence intervals of the mean positions of canopy types (coniferous as dashed lines in orange; deciduous as dashed lines in blue and mixed as dashed lines in green).See supplementary information TableS3for more details on species names.

Fig. 5 .
Fig. 5. Principal coordinates analysis (PCoA) plots on a) total vegetation community composition; b) vascular plant community composition; and c) bryophyte and lichen community composition in each site represented by symbols (Amos, Nédelec and Rivière-Hiva).The ellipses are the 95% confidence intervals of the mean positioned according to canopy types (Coniferous as dashed lines in orange; deciduous as dashed lines in blue and mixed as dashed lines in green).See supplementary information TableS3for more details on species names.

Fig. 6 .
Fig. 6.Mean richness (±SEM) of bryophytes and lichens a) On soil; b) At the base of trees; c) On the trunk of trees.

Fig. 7 .
Fig. 7. Composition of bryophytes and lichens on each microhabitat for each plantation type.a) Bryophyte and lichen composition on soil; b) Bryophyte and lichen composition at the base of trees; c) Bryophyte and lichen composition on trunk of trees.See supplementary information TableS3for more details on species names.