Skip to main content
NCBI home page
AoB Plants logoLink to AoB Plants
. 2021 Feb 23;13(2):plab011. doi: 10.1093/aobpla/plab011

Presence of a dominant native shrub is associated with minor shifts in the function and composition of grassland communities in a northern savannah

Isaac Peetoom Heida 1,✉, Charlotte Brown 1, Margarete A Dettlaff 1, Kenneth J Oppon 1, James F Cahill Jr 1
Editor: David Gorchov
PMCID: PMC8050699  PMID: 33889378

Abstract

Ecosystems are spatially heterogenous in plant community composition and function. Shrub occurrence in grasslands is a visually striking example of this, and much research has been conducted to understand the functional implications of this pattern. Within savannah ecosystems, the presence of tree and shrub overstories can have significant impacts on the understory herbaceous community. The exact outcomes, however, are likely a function of the spatial arrangement and traits of the overstory species. Here we test whether there are functional linkages between the spatial patterning of a native shrub and the standing biomass, community composition, and overall nutrient cycling of a neighbouring grassland understory communities within the Aspen Parkland of central Alberta, Canada. In a paired grassland-shrub stand study, we found the native shrub, Elaeagnus commutata, has relatively few stand-level impacts on the composition and standing biomass of the ecosystem. One factor contributing to these limited effects may be the overdispersion of shrub stems at fine spatial scales, preventing areas of deep shade. When we looked across a shrub density gradient and incorporated shrub architecture into our analyses, we found these shrub traits had significant associations with species abundance and root biomass in the understory community. These results suggest that stem dispersion patterns, as well as local stand architecture, are influential in determining how shrubs may affect their herbaceous plant understory. Thus, it is important to incorporate shrub spatial and architectural traits when assessing shrub-understory interactions.

Keywords: architecture, composition, grassland, parkland, savannah, shrub, structure, understory

Shrubs can have differential impacts on neighboring plants. Species diversity and ecosystem can have a large influence on the exact effects a shrub may have on its surroundings, making comparisons between studies difficult. Here, we test whether shrub spatial arrangement and architecture is associated with impacts on the surrounding plant community. Compared to binary shrub-grassland analyses, our trait-based approach was able to detect significant associations between shrub presence and plant community function. Our shrub architecture approach is likely applicable to other shrub-focussed research and may help to increase our understanding of patterns of shrub presence in a variety of ecosystems.

Introduction

Many ecosystems contain heterogenous subcomponents, each with distinct plant communities and functions. In savannah type ecosystems, the presence or absence of tall woody species can play a dramatic role in the structure and function of low-lying herbaceous plant communities. Understanding how shrubs influence plant communities has gained increased interest over the past decades, in part due to the dramatic negative effects often associated with increases in shrub density (Van Auken 2009). Reports of negative shrub impacts have traditionally arisen from ecosystems where there is little historic woody cover (Eldridge et al. 2011). In these scenarios, shrubs present a novel architecture and growth form that can alter the functioning of the ecosystem (Knapp et al. 2008), an outcome often viewed as negative in economically important grasslands. However, even without encroachment, understanding the structure and function of shrub dominated communities is important as these communities can have distinct ecological functionality within the broader landscape. In areas where shrublands, grasslands and forests all coexist, shrublands can act as intermediate successional stages in transitions between grasslands and forests (Archer et al. 1988; Aerts et al. 2006). The complexity and heterogeneity of these grassland–shrubland–forest ecotones also results in temporal instability, where small changes to disturbance or climatic regimes can result in dramatic shifts in the relative abundance of each component of the system (Camill et al. 2003). In these systems, it is important to understand the function of each component type, so that we can better understand the implications of dramatic community-cover shifts on landscape level functionality.

Shrublands can develop from grasslands in response to shifts in the disturbance regime such as reductions in fire frequency and overgrazing (Van Auken 2009). Anthropogenic activity has perpetuated both of these changes, increasing the likelihood for grasslands to change into shrublands. As shrubs expand into new areas, they can alter soil nutrient cycles (Jackson et al. 2002; Baer et al. 2006; Zhou et al. 2019), change light and moisture availability (Scott et al. 2006; Brantley and Young 2007), influence plant community composition (Báez and Collins 2008), and alter the physical characteristics (such as topography) of the habitat (Parizek et al. 2002). Shrubs can also interact with trees, and there is evidence both of facilitation of tree establishment (Smit et al. 2008) as well as suppression of tree growth (Berkowitz et al. 1995). The exact outcomes of shrub presence in a given system are hard to predict (Eldridge et al. 2011) as they are contingent on multiple local abiotic and biotic factors such as precipitation (Knapp et al. 2008), grazing pressure (Eldridge and Soliveres 2015) and the traits of the shrub itself (Eldridge et al. 2011). Shrub traits likely have consequences for plant community assembly as they determine the exact effects of the shrub on resource availability, creating assembly filters for understory communities. Shrub canopies overtop low-lying grassland vegetation, intercepting light and reducing light availability below the canopy (Brantley and Young 2007), sometimes favouring the establishment of shade-tolerant species (Yang et al. 2010). Shrubs’ height and rigid stems also allow for the creation of refugia under their canopies, as they can ameliorate stressors common in grassland systems. Shrubs can reduce drought severity (Cruz-Alonso et al. 2020) and abate the negative effects of livestock grazing (Bailey 1970; Eldridge et al. 2015; Perea et al. 2017) as well as acting as nurse plants after fire events (Raffaele and Veblen 1998). The reductions in disturbance severity under shrub canopies, coupled with light reductions, might interact to create an assembly filter selecting for shade-tolerant species, resulting in a compositional shift away from a natural grassland and towards either a shrub-dominated state (Lett and Knapp 2003) or composition more typical of a forested system (Dalotto et al. 2018). However, the degree to which shrubs act as assembly filters through resource modification should be a direct consequence of the architecture and spatial organization of the establishing shrub.

Savannah ecosystems represent a unique context for quantifying differences between shrubbed and shrubless areas. Defined by the presence of scattered woody vegetation in a grassland matrix, savannahs are held in (dis)equilibrium by resource limitation and disturbance regimes (Accatino et al. 2010). The permanent presence of woody species in savannahs contrasts with grassland systems which typically have much lower woody cover (in the absence of encroachment) (Van Auken 2000). The Aspen Parkland region of North America, characterized by patches of aspen forest (Populus tremuloides Michx.) interspersed within a grassland matrix, provides a good system to examine shrub-grassland associations because it is a transitionary zone between the great plains to the south and boreal forest to the north (Bird 1930). This system naturally experiences transitions between woody and non-woody community types (Pinno and Wilson 2011), with the rate of transition being influenced by the effects of grazing (Bork et al. 2013), drought (Hogg et al. 2008) and fire (Bailey et al. 1990). Changes in these disturbance regimes due to climate change (Schneider et al. 2009) or human activity (Michaelian et al. 2011) will likely lead to changes in the relative abundance of woody plants within the savannah matrix.

Of the woody parkland species, the shrub wolf-willow (Elaeagnus commutata Bernh. ex Rydb.) has a set of traits that will likely make its impacts on understory plant community assembly unique. Wolf-willow is an actinorhizal nitrogen-fixing shrub (Visser et al. 1991), reproducing clonally via its root system to form continuous stands rather than occurring as a single or a few isolated stems. Increased fertility via nitrogen fixation could interact with shrub canopy light interception to modify community assembly filters within the local grassland (Stevens et al. 2017). Wolf-willow is also an ‘increaser’ in overgrazed rangelands (USDA 2020) with density typically controlled by fire and moderate grazing/mowing (Arnold and Higgins 1986). Combined anthropogenic fire suppression and changes to grazing regimes have the potential to drive changes in the abundance of this shrub. At least one study has reported increases in abundance of wolf-willow in fescue grasslands (Bailey 1970); however, the potential impacts on assembly filters and plant understory community composition remain unclear. Information on how wolf-willow communities differ from adjacent shrubless grasslands is sparse and more data is needed to understand the pattern and processes associated with wolf-willow presence in grasslands. In this study, we aim to test the impacts of wolf-willow on plant community composition, understory biomass, and nutrient cycling within a native grassland. We build on previous work that has documented grassland community shifts in response to wolf-willow (Whysong and Bailey 1975) by incorporating the architecture of wolf-willow stands and examining its effects on both biotic and abiotic properties of the grassland. Here we answer three main questions:

  • 1) Do wolf-willow and grassland plots differ in core aspects of biomass, fertility and biodiversity?

  • 2) Are aspects of wolf-willow architecture associated with observed variation in understory biomass, fertility and biodiversity?

  • 3) Does wolf-willow show evidence of acting as an assembly filter?

Methods

Study site

The study occurred in a 50-hectare field at the Roy Berg Kinsella Research Ranch in Alberta, Canada (53.0848°, -111.5600°). The site is in the Aspen Parkland ecoregion and the dominant grassland species include Festuca hallii, Hesperostipa curtesita and Poa pratensis (Lamb 2008). The site has historically been lightly grazed by cattle, with the last grazing event occurring 3 years prior to this study. Within the site, wolf-willow stands occur both at forest margins as well as in core grassland communities.

Study design

We established 10 replicate blocks that were each comprised of two 3 × 3 m plots. These plots sizes were chosen to optimize our ability to detect local community differences; however, we recognize they may reduce our ability to measure landscape-level patterns. These paired plots were separated by <10 m, with one located in a wolf-willow stand and the second in a nearby native grassland with <1 % wolf-willow cover. We avoided selecting paired plot locations that bordered aspen stands to avoid potential confounding effects from aspen stands. Blocks were separated by at least 30 m, with plots within blocks placed randomly within the suitable grassland and wolf-willow regions. Within each block, we had three scales of measurement: the two plots, quadrants within plots and points within plots. Quadrants were created by subdividing each plot into four subplots, each 1.5 × 1.5 m. Point measurements were taken at a series of 25 10 cm diameter points laid out identically within each plot (Fig. 1). This pattern is subset of the sampling design used in Chagnon et al. (2018). We generated a series of random subsets of their sampling point pattern, and selected a pattern based on the pair-wise distance distribution and the number of points per quadrant. We selected for the pattern that had both a relatively flat distribution (increasing replication at nearest and furthest pair-wise distances) and relative parity in the number of points per quadrant. Table 1 displays the final structure of our study, outlining which data were measured at each scale.

Figure 1.

Figure 1.

Open in a new tab

Sampling design schematic for examining patterns of wolf-willow presence in the Aspen Parkland. On the left, a conceptual diagram of how plots were arranged within blocks. Community type designation is written in each plot. On the right, an expanded view of the sampling design within one plot. Solid circles are the points at which images were taken and biomass and soil moisture measured. Hollow squares designate the location of soil cores for nutrient analysis.

Table 1.

Scales of measurement in this study, and the corresponding measurements taken at each scale.

Scale Data
Plot Wolf-willow map
Wolf-willow architecture
• Height
• Basal area
• Canopy area
Quadrant Soil nitrogen
Soil moisture
Soil pH
Point Above-ground biomass
Below-ground biomass
Litter mass
Soil moisture
Canopy pictures (Wolf-willow only)
Community composition (Presence/Absence)
Open in a new tab

Wolf-willow structure and spatial characteristics

To characterize the architecture of wolf-willow stands, all wolf-willow stems in plots were mapped to the nearest centimetre. Additionally, we recorded basal area, maximum stem height and canopy area for each stem. Canopy area was calculated as the oval area of the widest diameter and the corresponding perpendicular diameter. To summarize these architectural traits and get an overall measure of wolf-willow patch characteristics, we first summed or averaged the measures at the plot-level, then used a principal component analysis (PCA) to ordinate the wolf-willow measurements: total canopy area, stem density, total basal area, average height and average canopy cover (seeSupporting Information—Fig. S2). Due to asexual reproduction by wolf-willow, stem density can be high both in very young and in mature stands. By using architectural measures, we hope to have a more sensitive analysis than simple shrub presence or stem density, examining how the density and size of wolf-willow within a stand influences the understory community. We termed the axis scores on the primary axis as ‘wolf-willow score.’ The score was positively associated with all traits, so a higher score generally corresponds to more wolf-willow. It should be noted that these traits likely have correlations to stand age (total basal area, average height) and stand developmental stage (total canopy cover, stem density). We did not quantify these relationships though, and as such will avoid making inferences linking score to stand age or developmental stage.

Understory community composition and standing biomass

To investigate how wolf-willow and grassland plots differ in biodiversity and standing biomass, we sampled at the point scale recording species presence. All vegetation and litter sampling occurred at peak biomass (13–16 August). We collected peak above-ground herbaceous biomass via clipping, collected below-ground biomass using root cores (20 cm deep × 5 cm diameter) and collected litter via raking. Due to ongoing experiments at the site, we did not harvest wolf-willow above-ground biomass. Above-ground biomass and litter were dried at 60 °C for at least 48 h, then weighed. Below-ground biomass was field sieved and washed over a 2 mm sieve, dried and weighed. We were unable to reliably differentiate wolf-willow roots from other species, and for consistency weighed all roots occurring within a core, regardless of identity. As we were interested primarily in the root mass, any bulbs found were separated and weighed separately. During this process, we observed nodulation of wolf-willow roots, however, this was not quantified.

Understory resource availability

To determine the effect of wolf-willow on soil resource availability, we measured soil moisture at each of the 25 points in each plot (13 August) using a Theta Probe (Delta-T Devices, Cambridge, UK). Though single measurements are unreliable, this serves as a coarse look at differences in soil moisture, rather than gathering detailed hydrological data. Due to cost considerations, we used a different sampling design to measure soil chemistry parameters, measuring these parameters at the quadrant scale. We took four soil cores (15 cm deep × 2 cm diameter) per quadrant (Fig. 1). Cores were then pooled by quadrant and sieved with a 2 mm sieve to discard any roots, creating four samples per plot. We used these samples to measure nitrogen and pH. Mineral nitrogen was extracted using 2.0 M KCl (Saha et al. 2018) and then measured by flow injection analysis (Lachat QuickChem QC8500 FIA Automated Ion Analyzer, Loveland, CO, USA) and pH was measured using a pH probe (ThermoFisher Scientific, Waltham, MA, USA). Like soil moisture, one-time sampling of nitrogen does not capture the long-term dynamics of soil chemistry. However, a one-time sampling can still provide insights into potential differences in inorganic N availability under similar environmental conditions.

To determine the effects of wolf-willow on light availability, we used hemispherical photography to measure the canopy cover at the 25 points within each wolf-willow plot (Fig. 1). We used a smartphone camera equipped with an Aukey 198° fisheye lens mounted on a selfie stick (Bianchi et al. 2017). Images were taken from ground level, post-vegetation clipping. Images were processed in ImageJ (Schneider et al. 2012) and we extracted the gap fraction using CIMES-FISHEYE (Gonsamo et al. 2011). To make the gap fraction measurements more intuitive we converted it to canopy closure (1 – Gap Fraction). We assumed a fixed height for canopies, based on the average height of shrub canopies within our study (1.08 m). We limited the search angle of the image analysis to 2.5 degrees (Θ = tan-1[point radius ÷ canopy height]), creating a search window the same size as our understory sampling points (5 cm radius).

Statistical analysis

To quantify the spatial distribution of wolf-willow stems, we used a pooling procedure with Ripley’s L-function (Besag 1977; Baddely et al. 2015). We compared the observed stem distribution to a null model of random distribution, allowing us to identify scales at which stems were either clustered or overdispersed more than expected by chance.

To quantify the differences between wolf-willow and grassland plots in aspects of fertility and biomass stocks we used mixed-effects models comparing wolf-willow and grassland plots with block as a random effect. In eight separate models, we fit either above-ground biomass, below-ground biomass, litter, soil moisture, nitrogen (NO3, NH4, total mineral nitrogen) or pH as response variables. To assess the response of vascular plant diversity to wolf-willow presence, we used mixed-effects models with either richness or evenness (Pielou’s evenness index) as response variables, wolf-willow presence as a fixed effect and block as a random effect. The effects of wolf-willow presence on plant community composition were tested using a permutational multivariate analysis of variance (PERMANOVA) with 999 permutations. Wolf-willow presence was the predictor variable, and we incorporated the block design by constraining randomization to within a block. Before the ordination we used a Hellinger transformation on our species matrix to down-weight the importance of rare species (Legendre and Gallagher 2001).

To investigate how wolf-willow architecture is associated with our fertility, biomass and biodiversity parameters, we used 10 linear regressions, regressing each of above-ground biomass, below-ground biomass, litter, soil moisture, NO3, NH4, total mineral nitrogen, pH, average alpha diversity (richness per point within plot) and beta diversity (between points within plot) against wolf-willow score. Beta diversity was calculated using the betadisper function in R package vegan. This method is based on Anderson et al. (2006) and uses Euclidean distance to quantify the multivariate dispersion of plot plant community composition compared with the spatial median of all wolf-willow plots. Blocks were not specified in the model, as each block contains one wolf-willow plot.

To investigate the potential of wolf-willow to serve as an assembly filter we first used a constrained ordination, testing if plant composition was better explained by an environmental factor than by our community types. We used stepwise model selection to identify which of soil moisture, NO3, NH4, total mineral nitrogen, pH and community type best explained the data. Soil moisture and community type emerged as the top variables. We then used redundancy analysis and PERMANOVA with soil moisture as the predictor variable to determine how well community composition was explained by soil moisture.

To get at finer scale in changes in the understory community, we investigated single species’ responses to wolf-willow presence. We used generalized linear mixed-effects models testing the response of species abundance to a community type fixed effect. For this analysis, we defined species abundance as fraction of total sampling points with observed species’ presence. Block was inputted as a random effect. We tested all species, as our interest was primarily in seeing if wolf-willow and grassland communities exhibit differences in assembly, rather than making conclusions about specific species’ responses to either of the community types.

All analyses were done in R 3.5.3 (R Core Team 2019), using packages spatstat (Baddely et al. 2015), lme4 (Bates et al. 2015) and vegan (Oksanen et al. 2019).

Results

Wolf-willow structural and spatial characteristics

Within our shrub plots, wolf-willow densities ranged from 1.78 to 5.22 stems/m2 with an average density of 2.91 ± 0.40 stems/m2 (mean, SE). Average canopy closure was 13.3 ± 1.19 %, with a range of 4.3–24.7 %. Stems were significantly clustered spatially at small scales (4–30 cm) but were randomly distributed at larger scales (31–80 cm) (seeSupporting Information—Fig. S1). No wolf-willow individuals were detected in the grassland plots.

Patterns of wolf-willow presence

Wolf-willow had few associations with core community metrics. Despite an average canopy cover of 13 %, we found no difference between wolf-willow dominated and grassland community types for species richness, biomass, soil nutrients or soil moisture (Table 2; Fig. 3). In contrast, wolf-willow stands were associated with greater litter biomass and lower soil pH (Table 2; Fig. 3). Neither species richness nor evenness differed between wolf-willow and grassland community types (Table 2; Fig. 2); however, the NMDS and PERMANOVA showed a significant divergence in composition (Fig. 2).

Table 2.

Results of linear mixed effects models on the effect of wolf-willow presence on factors of the understory community. All tests had 1,8 degrees of freedom, and all values are rounded to three decimal places.

Measurement F P
Soil pH 5.457 0.044
Soil ammonium (NH+4) 3.557 0.093
Soil total mineral nitrogen 3.265 0.104
Soil nitrate (NO−3) 0.978 0.348
Soil moisture 0.476 0.508
Litter mass 5.541 0.043
Above-ground biomass 2.969 0.119
Below-ground biomass 0.068 0.801
Richness 0.425 0.531
Evenness 2.924 0.122
Open in a new tab

Figure 3.

Figure 3.

Open in a new tab

Differences between grassland and wolf-willow (Elaeagnus commutata) dominated communities in fertility and biomass stocks. Results of mixed effects models on these data are presented in Table 2. Error bars are 1 standard error.

Figure 2.

Figure 2.

Open in a new tab

The effects wolf-willow (Elaeagnus commutata) presence on community composition in the Aspen Parkland. On the left, an ordination of the compositional data, with the results of a PERMANOVA in the top right of the plot. On the right, two bar plots showing the differences in richness (top), and evenness (bottom) between the two community types, neither richness nor evenness were significantly different (df = 9 for both, t = -0.653, P = 0.531 and t = -1.7098, P = 0.122, respectively). Error bars are 1 standard error.

Effects of wolf-willow architecture

The lack of overall effects from the binary presence/absence of wolf-willow masks variation observed within wolf-willow stands. The PCA generated a primary axis that explained 73.8 % of variation and was positively associated with all inputted measures (seeSupporting Information—Fig. S2). As our wolf-willow score increased, there was significant positive association with NO3 (F = 5, df = 1,8, P = 0.001) and total mineral nitrogen (F = 7.513, df= 1,8, P = 0.025) and significant negative associations with root biomass (F = 2.924, df = 1,8, P = 0.0102) and soil moisture (F = 5.813, df = 1,8, P = 0.042) (Fig. 4). The total mineral nitrogen association is likely driven by the NO3 association, as NH4 had weak evidence of association (F = 2.039, df = 1,8, P = 0.191). Wolf-willow had no effect on average alpha or beta diversity (F = 2.786, F = 0.664, df = 1,8 both, P = 0.134, 0.439, respectively) or plot-level species richness (F = 0.893, df = 1,8, P = 0.893), though there was a trend towards reduced species evenness as the wolf-willow score increased (F = 3.980, df = 1,8, P = 0.081).

Figure 4.

Figure 4.

Open in a new tab

Linear models of how wolf-willow (Elaeagnus commutata) score relates to biomass stocks and fertility components of understory communities. Wolf-willow score is derived from Axis 1 on a PCA of wolf-willow structural traits. A line of best fit indicates a significant relationship between the two variables.

In the constrained ordination, we did not find evidence of an environmental assembly filter driving our observed plant composition. The primary axis of soil moisture explained 8.8 % of the variation in composition. The PERMANOVA showed that soil moisture was not as strong a predictor of composition as community type (P = 0.093, F = 0.187, df = 1,18, compared with P = 0.034 for community type).

Beneath the breadth of mostly null results at the community levels, we did find evidence that wolf-willow influences certain species within the grassland. Among species, there is a response gradient to wolf-willow presence, with species’ responses ranging from significantly negative to significant positively (Table 3; Fig. 5).

Table 3.

Species’ count responses to wolf-willow presence. Using a GLM with a Poisson distribution, differences in species counts between grassland and wolf-willow communities were analysed. A negative estimate indicates bias towards grassland, while a positive estimate indicates bias towards wolf-willow. Species with singular fitting models or model errors were excluded.

Species Name Estimate SE Z P
Distichlis spicata -1.7918 0.7638 -2.346 0.019
Danthonia intermedia -1.4663 0.6404 -2.29 0.022
Astragalus agrestis -0.9555 0.3721 -2.568 0.0102
Hesperostipa curtesita -0.6408 0.1745 -3.671 0.000241
Carex sp. -0.3477 0.138 -2.52 0.0117
Poa pratensis 0.24 116 0.10 068 2.395 0.0166
Galium boreale 0.6931 0.2094 3.309 0.000935
Erigeron caepsitosus -31.28 6 196 000 0 1
Viola canadensis -0.77 317 0.49 354 -1.567 0.117
Potentilla arguta -0.6931 1.2247 -0.566 0.5714
Symphyotrichum laeve -0.5108 0.5164 -0.989 0.323
Thermopsis rhombifolia -0.5108 0.7303 -0.699 0.484
Viola adunca -0.3365 0.414 -0.813 0.416
Elymus glaucus -0.3054 0.3522 -0.867 0.386
Pascopyrum smithii -0.2877 0.7638 -0.377 0.706
Artemisia ludoviciana -0.1206 0.2163 -0.558 0.577
Artemisia frigida -9.822E-07 1 0 1
Viola nephrophylla 4.476E-05 1 0 1
Symphyotrichum falcatum 0.04082 0.28 544 0.143 0.886
Elymus trachycaulus 0.24 116 0.40 291 0.599 0.549
Achillea millefolium 0.2809 0.2172 1.294 0.196
Vicia americana 0.5108 0.7301 0.7 0.4841
Rosa arkansana 0.5232 0.3155 1.659 0.0972
Helianthus pauciflorus 0.6286 0.4378 1.436 0.151
Campanula rotundifolia 0.7621 0.4577 1.665 0.0959
Lactuca tatarica 0.8473 0.4879 1.737 0.0825
Bromus inermis 31.13 883.37 0.035 0.972
Viola pedatifida 39.78 5792.62 0.007 0.995
Open in a new tab

Figure 5.

Figure 5.

Open in a new tab

Response of species counts to wolf-willow presence. Error bars represent 1 standard error. Point fill corresponds to significance. White is P ≥ 0.1, grey is 0.1 > P ≥ 0.05 and black is P < 0.05. Erigeron caespitosus, Bromus inermis and Viola pedatifida all had extreme values, and are not shown in the plot area. Exact values for all species can be found in Table 3.

Discussion

Though wolf-willow is a dominant plant in the Aspen Parkland, our results show that its presence alone is associated with relatively minor changes in community composition, understory biomass and soil resource availability. These results run counter to many shrub encroachment studies, where the establishment of shrubs can have strong negative effects on the species richness and productivity of the herbaceous species in a system (Briggs et al. 2002, Price and Morgan 2008, Castro and Freitas 2009, López-Díaz et al. 2015). These negative effects can arise due to increased litter deposition (Hilger and Lamb 2017; Zhang et al. 2019). While we saw increased litter in our shrub understories, we did not detect an effect on richness or standing biomass. Though we observed equal above-ground biomass between the two community types, our biomass measures for wolf-willow stands excluded wolf-willow biomass. Had we measured this we likely would have found shrub stands to have higher above-ground biomass than grasslands. Our effect sizes are likely driven by the architectural traits of wolf-willow stands, where the low stem density, sparse canopy and clumped arrangement constrain the strength of effect that wolf-willow has on the communities where it occurs (Bork and Burkinshaw 2009; Sabo et al. 2009).

Based on the number of significant relationships, our ‘wolf-willow score’ analyses seem to be more sensitive than our presence/absence comparisons. The associations between wolf-willow and soil nitrogen and pH are likely driven by interactions between nitrogen fixation and litter deposition. The observed nitrogen enrichment may have arisen directly through fixation (Paul et al. 1971), or through deposition and decomposition of leaf litter (Siemann and Rogers 2003). Acidification in wolf-willow stands could be driven indirectly by fixation-derived ammonium ions. Plant use of ammonium can acidify the rhizosphere (von Wirén et al. 2000) while nitrification of ammonium releases H+ ions into the soil (Cole 1995). Both acidification and nitrogen enrichment can have impacts on the understory plant community, altering the bioavailability of NO−3/NO−2 (Kemmitt et al. 2005), as well as the nodulation of legume roots (Coventry et al. 1985). Allelopathy is one plausible explanation of the observed changes in soil chemistry. Other studies on allelopathy have observed changes in soil pH resulting from allelochemical inputs (Morris et al. 2012; Uddin et al. 2020). A congener has shown some evidence of allelopathic effects (Orr et al. 2005), as well as altering rates of nitrification in soil (Lunares et al. 1993), but more detailed work is needed on wolf-willow to investigate this possibility. Our soil chemistry measurements lack a temporal dimension and adding this data may help to solidify our understanding of wolf-willow presence on soil, as well as how these conditions develop.

There are two possible explanations to the observed negative association between wolf-willow score and root biomass. Wolf-willow could be reducing overall root biomass in the system, and thus changing root:shoot ratios of understory plants. However, because we did not sample wolf-willow shoot biomass, it is difficult to say conclusively how root:shoot ratios may differ between grasslands and shrub stands. Other studies have found understories of woody plants to have lower root:shoot ratios than grasslands (Wilson 1993; Grouzis and Akpo 1997; Pinno and Wilson 2011). Examples of studies finding declines in overall root mass associated with increasing shrub presence are difficult to find (but see Yusuf et al. 2015), so at this time we lack a clear mechanism to support this hypothesis. Allelopathy (mentioned previously) may be responsible for declines in root mass (Orr et al. 2005), but without clear evidence of wolf-willow allelopathy this is merely speculation. The second possibility explaining observed root mass declines is that wolf-willow roots deeper than many grassland species, causing a deeper allocation of root biomass (thus out of range of our cores). This could be evidence of soil niche partitioning (Kulmatiski and Beard 2013), though other studies on niche partitioning of grass and shrub roots typically see grass rooting become shallower rather than deeper in response to shrub presence (Peláez et al. 1994, Nippert and Holdo 2015). In the absence of root:shoot allocational shifts, this explanation would also logically predict lower herbaceous above-ground biomass, as deeper allocation explaining lower biomass assumes a zero-sum scenario of root biomass. The fact that we do not see reduced shoot biomass makes it unlikely for this to be the sole explanation, though the true process might involve components of both explanations presented here. Grasslands are important carbon sinks (Bengtsson et al. 2019), and the potential for wolf-willow presence to reduce root mass in grasslands may be detrimental to the provisioning of this service, particularly at high shrub densities.

Plot-level richness seems to be too coarse a measure to detect wolf-willow effects on understory community composition. Using more sensitive analyses, we can see evidence that wolf-willow is affecting the community composition of the system. Our ordination plot shows minor divergence, supported by a marginally significant PERMANOVA. Evidence of wolf-willow fine-tuning the understory composition can further be seen by the evenness response. The negative association between wolf-willow score and evenness suggests that dominance increases in wolf-willow understories. A prime beneficiary of this effect is Poa pratensis, a dominant grass in the Aspen Parkland (Lamb 2008), which showed a significant positive association to wolf-willow presence. However, not all dominant species increase under wolf-willow. Hesperostipa curtesita is another dominant grass in this system (Lamb 2008), yet showed a significant negative association with wolf-willow. While many studies report richness effects associated with shrub presence (Anthelme et al. 2007; Price and Morgan 2008; Ratajczak et al. 2012; Madsen et al. 2020), our study shows that important community changes can occur even in the absence of a richness effect. Future studies should consider incorporating architectural or abundance measures (like our wolf-willow score) as well as looking at other components of biodiversity.

Though wolf-willow presence dramatically alters the visual landscape of grasslands in the Aspen Parkland, the realized changes in the grassland community we observed are more subtle, supporting a nuanced view of community dominance (Grieg-Smith 1986; Frieswyk et al. 2007). With evidence of changes in soil chemistry, soil moisture and below-ground biomass, wolf-willow seems to be associated with important below-ground processes; however, to understand both the mechanisms and outcomes, more detailed studies are needed. The results of our study provide baseline information on the functionality of wolf-willow dominated areas in the Aspen Parkland. The results of our architectural approach suggest that architecture may be a currently understudied driver of shrubland dynamics. By incorporating architectural traits and spatial arrangement, we might more accurately compare shrubland dynamics across species and ecosystems, aiding in the development of a more generalized understanding of shrubland patterns and processes.

Supporting Information

The following supporting information is available in the online version of this article—

Figure S1. Summarizes wolf-willow physical and spatial characteristics, including the plot of the L-Function.

Figure S2. The results of the PCA on wolf-willow architectural measures and how they relate to wolf-willow score (the primary PCA axis).

plab011_suppl_Supplementary_Figures
Click here for additional data file. (590.7KB, pdf)

Acknowledgements

We would like to thank R. Vinebrooke, P.L. Chagnon and M. Dale for their assistance in experimental design and data analysis. J. Grenke, H. Mahal, T. Barber-Cross and J. Wild assisted in the field and laboratory.

Populations & Communities. Chief Editor: Jean Burns

Sources of Funding

This study was funded by an Natural Sciences and Engineering Research Council (NSERC) Undergraduate Student Research Award (IPH), an NSERC Discovery Grant (JFC) and an NSERC Discovery Accelerator Supplement Grant (JFC).

Conflict of Interest

None declared.

Contributions by the Authors

All authors but KO designed the study. IPH, MAD and KO collected and processed field samples. IPH analysed the data and wrote the manuscript. CB, JFC and MAD significantly contributed to revisions.

Data Availability

All code and data used in this study are available at Peetoom Heida, Isaac; Brown, Charlotte; Dettlaff, Margarete A.; Oppon, Kenneth J.; and Cahill, James F. (2021): Presence of a dominant native shrub is associated with minor shifts in the function and composition of grassland communities in a northern savannah. Figshare. Dataset. https://doi.org/10.6084/m9.figshare.13697590

Literature Cited

  1. Accatino F, De Michele C, Vezzoli R, Donzelli D, Scholes RJ. 2010. Tree-grass co-existence in savanna: Interactions of rain and fire. Journal of Theoretical Biology 267:235–242. [DOI] [PubMed] [Google Scholar]
  2. Aerts R, November E, der Borght IV, Behailu M, Hermy M, Muys B. 2006. Effects of pioneer shrubs on the recruitment of the fleshy-fruited tree Olea europaea ssp. cuspidata in Afromontane savanna. Applied Vegetation Science 9:117–126. [Google Scholar]
  3. Anderson MJ, Ellingsen KE, McArdle BH. 2006. Multivariate dispersion as a measure of beta diversity. Ecology Letters 9:683–693. [DOI] [PubMed] [Google Scholar]
  4. Anthelme F, Villaret J-C, Brun J-J. 2007. Shrub encroachment in the Alps gives rise to the convergence of sub-alpine communities on a regional scale. Journal of Vegetation Science 18:355–362. [Google Scholar]
  5. Archer S, Scifres C, Bassham CR, Maggio R. 1988. Autogenic succession in a subtropical savanna: conversion of grassland to thorn woodland. Ecological Monographs 58:111–127. [Google Scholar]
  6. Arnold TW, Higgins KF. 1986. Effects of shrub coverages on birds of North Dakota mixed-grass prairies. Canadian Field-Naturalist 100:20–14. [Google Scholar]
  7. Baddely A, Rubak E, Turner R. 2015. Spatial point patterns: methodology and applications with R. Boca Raton, FL: Chapman and Hall/CRC Press. [Google Scholar]
  8. Baer SG, Church JM, Williard KWJ, Groninger JW. 2006. Changes in intrasystem N cycling from N2-fixing shrub encroachment in grassland: multiple positive feedbacks. Agriculture, Ecosystems & Environment 115:174–182. [Google Scholar]
  9. Báez S, Collins SL. 2008. Shrub invasion decreases diversity and alters community stability in northern Chihuahuan desert plant communities. Plos ONE 3:e2332. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Bailey AW. 1970. Barrier effect of the shrub Elaeagnus commutata on grazing cattle and forage production in central Alberta. Journal of Range Management 23:248–251. [Google Scholar]
  11. Bailey AW, Irving BD, Fitzgerald RD. 1990. Regeneration of woody species following burning and grazing in Aspen Parkland. Journal of Range Management 43:212–215. [Google Scholar]
  12. Bates D, Mächler M, Bolker B, Walker S. 2015. Fitting linear mixed-effects models using lme4. Journal of Statistical Software 67:1–48. [Google Scholar]
  13. Bengtsson J, Bullock JM, Egoh B, Everson C, Everson T, O’Connor T, O’Farrell PJ, Smith HG, Lindborg R. 2019. Grasslands—more important for ecosystem services than you might think. Ecosphere 10:e02582. [Google Scholar]
  14. Berkowitz AR, Canham CD, Kelly VR. 1995. Competition vs. facilitation of tree seedling growth and survival in early successional communities. Ecology 76:1156–1168. [Google Scholar]
  15. Besag JE. 1977. Contribution to the discussion of the paper by Ripley (1977). Journal of the Royal Statistical Society, Series B 39:193–195. [Google Scholar]
  16. Bianchi S, Cahalan C, Hale S, Gibbons JM. 2017. Rapid assessment of forest canopy and light regime using smartphone hemispherical photography. Ecology and Evolution 7:10556–10566. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Bird RD. 1930. Biotic communities of the Aspen Parkland of central Canada. Ecology 11:356–442. [Google Scholar]
  18. Bork EW, Burkinshaw AM. 2009. Cool-season floodplain meadow responses to shrub encroachment in Alberta. Rangeland Ecology & Management 62:44–52. [Google Scholar]
  19. Bork EW, Carlyle CN, Cahill JF, Haddow RE, Hudson RJ. 2013. Disentangling herbivore impacts on Populus tremuloides: a comparison of native ungulates and cattle in Canada’s Aspen Parkland. Oecologia 173:895–904. [DOI] [PubMed] [Google Scholar]
  20. Brantley ST, Young DR. 2007. Leaf-area index and light attenuation in rapidly expanding shrub thickets. Ecology 88:524–530. [DOI] [PubMed] [Google Scholar]
  21. Briggs JM, Hoch GA, Johnson LC. 2002. Assessing the rate, mechanisms, and consequences of the conversion of tallgrass prairie to Juniperus virginiana forest. Ecosystems 5:578–586. [Google Scholar]
  22. Camill P, Umbanhowar CE, Teed R, Geiss CE, Aldinger J, Dvorak L, Kenning J, Limmer J, Walkup K. 2003. Late-glacial and holocene climatic effects on fire and vegetation dynamics at the prairie–forest ecotone in south-central Minnesota. Journal of Ecology 91:822–836. [Google Scholar]
  23. Castro H, Freitas H. 2009. Above-ground biomass and productivity in the Montado: From herbaceous to shrub dominated communities. Journal of Arid Environments 73:506–511. [Google Scholar]
  24. Chagnon P-L, Brown C, Stotz GC, Cahill JF. 2018. Soil biotic quality lacks spatial structure and is positively associated with fertility in a northern grassland. Journal of Ecology 106:195–206. [Google Scholar]
  25. Cole DW. 1995. Soil nutrient supply in natural and managed forests. Plant and Soil 168:43–53. [Google Scholar]
  26. Coventry DR, Hirth JR, Reeves TG, Jones HR. 1985. Development of populations of Rhizobium trifolii and nodulation of subterranean clover following the cropping phase in crop-pasture rotations in southeastern Australia. Soil Biology and Biochemistry 17:17–22. [Google Scholar]
  27. Cruz-Alonso V, Villar-Salvador P, Ruiz-Benito P, Ibáñez I, Rey-Benayas JM. 2020. Long-term dynamics of shrub facilitation shape the mixing of evergreen and deciduous oaks in Mediterranean abandoned fields. Journal of Ecology 108:1125–1137. [Google Scholar]
  28. Dalotto CES, Sühs RB, Dechoum MS, Pugnaire FI, Peroni N, Castellani TT. 2018. Facilitation influences patterns of perennial species abundance and richness in a subtropical dune system. Aob PLANTS 10:ply017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Eldridge DJ, Beecham G, Grace JB. 2015. Do shrubs reduce the adverse effects of grazing on soil properties? Ecohydrology 8:1503–1513. [Google Scholar]
  30. Eldridge DJ, Bowker MA, Maestre FT, Roger E, Reynolds JF, Whitford WG. 2011. Impacts of shrub encroachment on ecosystem structure and functioning: towards a global synthesis. Ecology Letters 14:709–722. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Eldridge DJ, Soliveres S. 2015. Are shrubs really a sign of declining ecosystem function? Disentangling the myths and truths of woody encroachment in Australia. Australian Journal of Botany 62:594–608. [Google Scholar]
  32. Frieswyk CB, Johnston CA, Zedler JB. 2007. Identifying and characterizing dominant plants as an indicator of community condition. Journal of Great Lakes Research 33:125–135. [Google Scholar]
  33. Gonsamo A, Walter J-MN, Pellikka P. 2011. CIMES: a package of programs for determining canopy geometry and solar radiation regimes through hemispherical photographs. Computers and Electronics in Agriculture 79:207–215. [Google Scholar]
  34. Grieg-Smith P. 1986. Chaos and order—organization. In: Kikkawa J, Anderson DJ, eds. Community ecology: pattern and process. Melbourne, Australia:Blackwell Scientific Publications, 19–29. [Google Scholar]
  35. Grouzis M, Akpo L-E. 1997. Influence of tree cover on herbaceous above- and below-ground phytomass in the Sahelian zone of Senegal. Journal of Arid Environments 35:285–296. [Google Scholar]
  36. Hilger H, Lamb EG. 2017. Quantifying optimal rates of litter retention to maximize annual net primary productivity on mixed-grass prairie. Rangeland Ecology & Management 70:219–224. [Google Scholar]
  37. Hogg EH (Ted), Brandt JP, Michaelian M. 2008. Impacts of a regional drought on the productivity, dieback, and biomass of western Canadian aspen forests. Canadian Journal of Forest Research 38:1373–1384. [Google Scholar]
  38. Jackson RB, Banner JL, Jobbágy EG, Pockman WT, Wall DH. 2002. Ecosystem carbon loss with woody plant invasion of grasslands. Nature 418:623–626. [DOI] [PubMed] [Google Scholar]
  39. Kemmitt SJ, Wright D, Jones DL. 2005. Soil acidification used as a management strategy to reduce nitrate losses from agricultural land. Soil Biology and Biochemistry 37:867–875. [Google Scholar]
  40. Knapp AK, Briggs JM, Collins SL, Archer SR, Bret-Harte MS, Ewers BE, Peters DP, Young DR, Shaver GR, Pendall E, Cleary MB. 2008. Shrub encroachment in North American grasslands: shifts in growth form dominance rapidly alters control of ecosystem carbon inputs. Global Change Biology 14:615–623. [Google Scholar]
  41. Kulmatiski A, Beard KH. 2013. Root niche partitioning among grasses, saplings, and trees measured using a tracer technique. Oecologia 171:25–37. [DOI] [PubMed] [Google Scholar]
  42. Lamb EG. 2008. Direct and indirect control of grassland community structure by litter, resources, and biomass. Ecology 89:216–225. [DOI] [PubMed] [Google Scholar]
  43. Legendre P, Gallagher ED. 2001. Ecologically meaningful transformations for ordination of species data. Oecologia 129:271–280. [DOI] [PubMed] [Google Scholar]
  44. Lett MS, Knapp AK. 2003. Consequences of shrub expansion in mesic grassland: Resource alterations and graminoid responses. Journal of Vegetation Science 14:487–496. [Google Scholar]
  45. López-Díaz ML, Rolo V, Benítez R, Moreno G. 2015. Shrub encroachment of Iberian dehesas: implications on total forage productivity. Agroforestry Systems 89:587–598. [Google Scholar]
  46. Lunares F, Muñoz-Mingarro D, Pozuelo JM, Ramos B, Castro FBD. 1993. Microbial inhibition and nitrification potential in soils incubated with Elaeagnus angustifolia L. leaf litter. Geomicrobiology Journal 11:149–156. [Google Scholar]
  47. Madsen B, Treier UA, Zlinszky A, Lucieer A, Normand S. 2020. Detecting shrub encroachment in seminatural grasslands using UAS LiDAR. Ecology and Evolution 10:4876–4902. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Michaelian M, Hogg EH, Hall RJ, Arsenault E. 2011. Massive mortality of aspen following severe drought along the southern edge of the Canadian boreal forest. Global Change Biology 17:2084–2094. [Google Scholar]
  49. Morris SJ, Herrmann DL, McClain J, Anderson J, McConnaughay KD. 2012. The impact of garlic mustard on sandy forest soils. Applied Soil Ecology 60:23–28. [Google Scholar]
  50. Nippert JB, Holdo RM. 2015. Challenging the maximum rooting depth paradigm in grasslands and savannas. Functional Ecology 29:739–745. [Google Scholar]
  51. Oksanen J, Blanchet FG, Kindt R, Legendre P, Minchin PR, O’Hara RB, Simpson GL, Solymos P, Stevens MHH, Wagner H. 2014. R package version 2.2-0. Vegan: Community Ecology Package. [Google Scholar]
  52. Orr SP, Rudgers JA, Clay K. 2005. Invasive plants can inhibit native tree seedlings: Testing potential allelopathic mechanisms. Plant Ecology 181:153–165. [Google Scholar]
  53. Parizek B, Rostagno CM, Sottini R. 2002. Soil erosion as affected by shrub encroachment in northeastern Patagonia. Rangeland Ecology & Management / Journal of Range Management Archives 55:43–48. [Google Scholar]
  54. Paul EA, Myers RJK, Rice WA. 1971. Nitrogen fixation in grassland and associated cultivated ecosystems. Plant and Soil 35:495–507. [Google Scholar]
  55. Peláez DV, Distel RA, Bóo RM, Elia OR, Mayor MD. 1994. Water relations between shrubs and grasses in semi-arid Argentina. Journal of Arid Environments 27:71–78. [Google Scholar]
  56. Perea R, López-Sánchez A, Dirzo R. 2017. Differential tree recruitment in California oak savannas: are evergreen oaks replacing deciduous oaks? Forest Ecology and Management 399:1–8. [Google Scholar]
  57. Pinno BD, Wilson SD. 2011. Ecosystem carbon changes with woody encroachment of grassland in the northern Great Plains. Écoscience 18:157–163. [Google Scholar]
  58. Price JN, Morgan JW. 2008. Woody plant encroachment reduces species richness of herb-rich woodlands in southern Australia. Austral Ecology 33:278–289. [Google Scholar]
  59. R Core Team . 2019. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. [Google Scholar]
  60. Raffaele E, Veblen TT. 1998. Facilitation by nurse shrubs of resprouting behavior in a post-fire shrubland in northern Patagonia, Argentina. Journal of Vegetation Science 9:693–698. [Google Scholar]
  61. Ratajczak Z, Nippert JB, Collins SL. 2012. Woody encroachment decreases diversity across North American grasslands and savannas. Ecology 93:697–703. [DOI] [PubMed] [Google Scholar]
  62. Sabo KE, Sieg CH, Hart SC, Bailey JD. 2009. The role of disturbance severity and canopy closure on standing crop of understory plant species in ponderosa pine stands in northern Arizona, USA. Forest Ecology and Management 257:1656–1662. [Google Scholar]
  63. Saha UK, Sonon L, Biswas BK. 2018. A comparison of diffusion-conductimetric and distillation-titration methods in analyzing ammonium- and nitrate-nitrogen in the KCl-extracts of Georgia soils. Communications in Soil Science and Plant Analysis 49:63–75. [Google Scholar]
  64. Schneider RR, Hamann A, Farr D, Wang X, Boutin S. 2009. Potential effects of climate change on ecosystem distribution in Alberta. Canadian Journal of Forest Research 39:1001–1010. [Google Scholar]
  65. Schneider CA, Rasband WS, Eliceiri KW. 2012. NIH Image to ImageJ: 25 years of image analysis. Nature Methods 9:671–675. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Scott RL, Huxman TE, Williams DG, Goodrich DC. 2006. Ecohydrological impacts of woody-plant encroachment: seasonal patterns of water and carbon dioxide exchange within a semiarid riparian environment. Global Change Biology 12:311–324. [Google Scholar]
  67. Siemann E, Rogers WE. 2003. Changes in light and nitrogen availability under pioneer trees may indirectly facilitate tree invasions of grasslands. Journal of Ecology 91:923–931. [Google Scholar]
  68. Smit C, den Ouden J, Díaz M. 2008. Facilitation of Quercus ilex recruitment by shrubs in Mediterranean open woodlands. Journal of Vegetation Science 19:193–200. [Google Scholar]
  69. Stevens N, Lehmann CE, Murphy BP, Durigan G. 2017. Savanna woody encroachment is widespread across three continents. Global Change Biology 23:235–244. [DOI] [PubMed] [Google Scholar]
  70. Uddin MN, Asaeda T, Shampa SH, Robinson RW. 2020. Allelopathy and its coevolutionary implications between native and non-native neighbors of invasive Cynara cardunculus L. Ecology and Evolution 10:7463–7475. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. USDA, N . 2020. The PLANTS Database. Database.http://plants.usda.gov/.
  72. Van Auken OW 2000. Shrub invasions of North American semiarid grasslands. Annual Review of Ecology and Systematics 31:197–215. [Google Scholar]
  73. Van Auken OW. 2009. Causes and consequences of woody plant encroachment into western North American grasslands. Journal of Environmental Management 90:2931–2942. [DOI] [PubMed] [Google Scholar]
  74. Visser S, Danielson RM, Parkinson D. 1991. Field performance of Elaeagnus commutata and Shepherdia canadensis (Elaeagnaceae) inoculated with soil containing Frankia and vesicular–arbuscular mycorrhizal fungi. Canadian Journal of Botany 69:1321–1328. [Google Scholar]
  75. von Wirén N, Gazzarrini S, Gojon A, Frommer WB. 2000. The molecular physiology of ammonium uptake and retrieval. Current Opinion in Plant Biology 3:254–261. [PubMed] [Google Scholar]
  76. Whysong GL, Bailey AW. 1975. Production and nitrogen content of herbage in a silverberry (Elaeagnus commutata) community compared to adjacent grassland and forest communities. Canadian Journal of Plant Science 55:801–808. [Google Scholar]
  77. Wilson SD. 1993. Belowground competition in forest and prairie. Oikos 68:146–150. [Google Scholar]
  78. Yang L, Ren H, Liu N, Wang J. 2010. The shrub Rhodomyrtus tomentosa acts as a nurse plant for seedlings differing in shade tolerance in degraded land of South China. Journal of Vegetation Science 21:262–272. [Google Scholar]
  79. Yusuf HM, Treydte AC, Sauerborn J. 2015. Managing semi-arid rangelands for carbon storage: grazing and woody encroachment effects on soil carbon and nitrogen. Plos ONE 10:e0109063. [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. Zhang A, Wang D, Wan S. 2019. Litter addition decreases plant diversity by suppressing seeding in a semiarid grassland, Northern China. Ecology and Evolution 9:9907–9915. [DOI] [PMC free article] [PubMed] [Google Scholar]
  81. Zhou L, Shen H, Chen L, Li H, Zhang P, Zhao X, Liu T, Liu S, Xing A, Hu H, Fang J. 2019. Ecological consequences of shrub encroachment in the grasslands of northern China. Landscape Ecology 34:119–130. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

plab011_suppl_Supplementary_Figures
Click here for additional data file. (590.7KB, pdf)

Data Availability Statement

All code and data used in this study are available at Peetoom Heida, Isaac; Brown, Charlotte; Dettlaff, Margarete A.; Oppon, Kenneth J.; and Cahill, James F. (2021): Presence of a dominant native shrub is associated with minor shifts in the function and composition of grassland communities in a northern savannah. Figshare. Dataset. https://doi.org/10.6084/m9.figshare.13697590

Articles from AoB Plants are provided here courtesy of Oxford University Press

RESOURCES