Physical disturbance expands the climatic envelope of annual plant dominance
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Genomics Radio. Generated commentary on the latest computational biology and genomics papers.
Ines: We're looking at the paper: "Physical disturbance expands the climatic envelope of annual plant dominance".
Marcus: Here's the gist of the paper. Disturbance nearly doubled the proportion of annual species and more than doubled the relative cover of annuals, whereas fertilisation had little influence and did not interact with disturbance Disturbance nearly doubled the proportion of annual species and more than doubled the relative cover of annuals, whereas fertilisation had little influence and did not interact with disturbance Life History Theory Context Classical lifehistory theory predicts that disturbance should shift communities from long-lived perennials toward short-lived annuals, and that nutrient enrichment may amplify this shift Annual species constitute ∼15% of the world’s herbaceous flora, with substantially higher representation in hot, dry climates Annuals often allocate more to rapid aboveground growth and seed production, while perennials tend to invest more in storage organs and belowground structures Disturbance can create these conditions by reducing survival of established individuals, particularly long-lived perennials, while opening recruitment windows for species that regenerate rapidly from seeds Experimental Findings on Drivers The study tested how physical disturbance (vegetation removal and shallow soil tillage) and fertilisation reshape annual–perennial balance, and whether disturbance relaxes the climatic limits of annual dominance Disturbance nearly doubled the proportion of annual species and more than doubled the relative cover of annuals Fertilisaton had little influence and did not interact with disturbance The disturbance-driven shift arose through contrasting pathways: in graminoids and legumes, it reflected the loss of perennial cover, while in forbs, the expansion of annual cover Climatic Envelope Expansion In the absence of disturbance, annual dominance was restricted to systems with extremely hot and dry summers, but disturbance nearly tripled the extent of climate space in which annuals dominated Disturbance expands the climatic limits of annual dominance by relaxing constraints that otherwise restrict annual dominance to sites with high temperatures and low precipitation during the warmest season The proportion of climatic space dominated by annuals increased from 11% to 29% when measured by species share, and from 12% to 34% when measured by relative cover Functional Group Variation Disturbance shifted annual–perennial balance through different pathways across functional groups In graminoids and legumes, the shift was driven mainly by reductions in perennial cover, although annual graminoid cover also increased modestly In forbs, perennial cover showed no clear response to disturbance, while annual forb cover roughly doubled Some of this variation is expected because annual prevalence is bounded between 0 and 100% leaving less scope for increase where annuals are already common Ecosystem Implications Annual plants can buffer grasslands against repeated disturbance and increasing droughts by rapidly reassembling from seed and sustaining vegetation cover and biomass production Perennial dominance and extensive belowground biomass are associated with many ecosystem services provided by grasslands, including long-term carbon storage, erosion control, and regulation of water runoff As disturbance regimes intensify and climates become more favorable to annuals, the key risk is erosion of perennial dominance, with consequences for long-term ecosystem functioning The findings advance understanding of the broad range of conditions under which disturbance promotes annual dominance, helping to predict and manage grassland responses to global change The paper concludes that disturbance can shift annual–perennial balance consistently across grasslands worldwide and expand the climatic conditions under which annual strategies dominate The analysis utilized the coordinated Disturbance and Recovery Across Global Grasslands Network (DRAGNet 27), which applies common experimental treatments and vegetation sampling protocols across 37 grassland sites The experiment applied common experimental treatments and vegetation sampling protocols across 37 grassland sites At each site, DRAGNet follows a randomized block design with 3–5 blocks Each block contains one 5 × 5, m plot assigned to each of four treatments used here: an unmanipulated Control, Disturbance, Nutrient Addition (NPK), and their combination (Disturbance + NPK) The Disturbance treatment is applied once per year for three consecutive years and is designed to create conditions for community reassembly by removing established vegetation and disrupting shallow belowground structures Within each 5 × 5 m plot, vegetation is sampled once per year during the growing season in a permanently marked 1 m2 quadrat located at least 1 m inside the plot boundaries We analysed the first three years after treatment initiation at each site, corresponding to the period during which the annual disturbance treatment was actively applied Statistical Modeling Network-wide treatment effects were estimated using generalized linear mixed models (GLMMs) to plot-level data These GLMMs allowed for combining non-Gaussian distributions (Beta, negative binomial, lognormal), zero-inflation components, and hierarchical random effects suitable for the multi-site, repeated-measures design Site-level variation in disturbance effects was quantified using the log response ratio and additive difference in percentage points Climate–disturbance models were fitted using zero–oneinflated Beta regression (BEINF) to model site-level proportions as functions of climate, identifying mean temperature of the warmest quarter and precipitation in the warmest quarter as the best predictors The analysis focused on these two variables because they emerged as the best predictors for both responses Models were fitted separately for each treatment to avoid imposing a restrictive parametric form on what are likely to be complex and treatment-specific interactions between climate and disturbance or fertilisation The expected mixture mean, E
Y: = (1 − ν − τ) µ + τ, was evaluated on a regular grid covering the convex hull of the observed climate space to visualize the climatic “dominance zone” of annuals Data and Site Details The dataset comprised 37 sites and 1,425–1,444 plot–year observations The study sites span a broad climatic gradient, from mean annual temperatures of approximately −5 to 25◦C and annual precipitation from roughly 200 to 1700 mm Data availability is provided through the Environmental Data Initiative repository at https://doi.org/10.6073/pasta/65799ea6ae60abdf0590b35134a9ab43 The authors acknowledge funding to E.
Ines: Borer and E.
Marcus: Seabloom from the National Science Foundation Research Coordination Network (NSF-DEB-2311608) and the Long Term Ecological Research program (DEB-1831944) to the Cedar Creek LTER The work was generated using data from the Disturbance and Recovery Across Grasslands experiment (DRAGNet), funded at the site-scale by individual researchers Key Statistical Results The Continuous Model for relative annual cover showed that Disturbance had a significant effect, with an estimated coefficient of 2.71 The Zero-Inflated Model for relative annual cover indicated that Disturbance had a significant effect, with an estimated coefficient of 0.11 The Continuous Model for annual species share showed that Disturbance had a significant effect, with an estimated coefficient of 2.21 The Zero-Inflated Model for annual species share indicated that Disturbance had a significant effect, with an estimated coefficient of 0.11 The continuous model for absolute perennial cover showed that Disturbance had a significant effect, with an estimated coefficient of 0.59 The zero-inflated model for absolute perennial cover indicated that Disturbance had a significant effect, with an estimated coefficient of 3.61 The continuous model for absolute annual graminoid cover showed that Disturbance had a significant effect, with an estimated coefficient of 1.20 The zero-inflated model for absolute annual graminoid cover indicated that Disturbance had a significant effect, with an estimated coefficient of 0.55 The continuous model for absolute annual forb cover showed that Disturbance had a significant effect, with an estimated coefficient of 1.75 The zero-inflated model for absolute annual forb cover indicated that Disturbance had a significant effect, with an estimated coefficient of 0.12 The continuous model for absolute perennial forb cover showed that Disturbance had a non-significant effect, with an estimated coefficient of 0.94 The zero-inflated model for absolute perennial forb cover indicated that Disturbance had a significant effect, with an estimated coefficient of 1.64 The continuous model for absolute annual legume cover showed that Disturbance had a non-significant effect, with an estimated coefficient of 0.98 The zero-inflated model for absolute annual legume cover indicated that Disturbance had a significant effect, with an estimated coefficient of 1.14 The continuous model for absolute perennial legume cover showed that Disturbance had a non-significant effect, with an estimated coefficient of 0.70 The zero-inflated model for absolute perennial legume cover indicated that Disturbance had a significant effect, with an estimated coefficient of 0.
Paper summary: Ines: So to wrap up what we've discussed about "Physical disturbance expands the climatic envelope of annual plant dominance," this paper is testing classical life history predictions against real-world data from a global experiment on thirty-seven grassland sites <ref:2607.08251#pg1>.
Marcus: The thesis they are pushing is that physical disturbance, like vegetation removal and shallow soil tillage, doesn't just cause some random change; it actively reshapes the annual versus perennial balance and expands the climatic conditions where annuals can succeed.
Yuki: They are specifically looking at how disturbance interacts with nutrient inputs too, because classical theory suggests nutrient enrichment might amplify the shift toward annuals, but this study found that fertilisation actually had little influence and didn't interact with the disturbance.
Ines: The main results they highlight show that physical disturbance nearly doubled the proportion of annual species and more than doubled their relative cover across different plant functional groups.
Marcus: They also quantified how this shift happened differently in graminoids versus forbs, noting that for graminoids and legumes, the change was driven mostly by a reduction in perennial cover.
Yuki: Meanwhile, for forbs specifically, the perennial cover showed no clear response to the disturbance treatment at all, while the annual forb cover roughly doubled.
Ines: And then they move on to climate space; they found that disturbance nearly tripled the extent of the climate space where annuals dominate.
Marcus: They quantified this by showing that when measuring species share, that proportion went from eleven percent to twenty-nine percent, and when measuring relative cover, it went from twelve percent up to thirty-four percent.
Yuki: This means the constraints on annual dominance—the need for extremely hot and dry summers—are relaxed by the disturbance itself.
Ines: So what this means for us is that we used to think annual dominance was highly sensitive to climate, but physical disturbance acts as a modulator that broadens the conditions under which those annual strategies can operate.
Marcus: It’s a significant finding because it suggests that managing grassland health isn't just about keeping conditions stable; it's about understanding how disturbance regimes fundamentally alter the ecological possibilities for plant communities.
Yuki: The paper sets up this idea by showing that annual species constitute around fifteen percent of the world’s herbaceous flora, and they already show a substantial representation in hot, dry climates.
Conclusion: Ines: So looking at the title and the overall findings of "Physical disturbance expands the climatic envelope of annual plant dominance," what does this actually change for how we think about these ecosystems?
Marcus: It changes our view on stability in grasslands because it shows that annuals aren't just a response to harsh climate; they can expand their realized niche through physical interactions with the environment.
Yuki: For population genetics, it suggests that the historical spread and persistence of annual species might have been more flexible than we assumed, as their dominance isn't strictly locked into narrow climatic windows.
Ines: I think the most important thing is that this work provides a mechanism—physical disturbance—that allows us to predict how grassland structure might shift under increasing climate change scenarios.
Marcus: Exactly, and it's important because they specifically showed that this effect is driven by physical processes rather than just nutrient availability or slow climatic drifts alone.
Yuki: So we see that the interplay between disturbance and climate isn't just about temperature; it’s about how those physical stresses reorganize the life history trade-offs of the plants themselves.
Ines: It moves us away from a static picture where annual dominance is only possible in very specific, extreme arid zones, toward one where disturbance opens up more possibilities globally.
Niv DeMalach, Roberto Salguero-Gomez, Oded Hollander, Jane Lucas, Alexander T. Strauss, Tilak Chaudhary, Y. Anny Chung, James McCarthy, Bruno Moreira
Institute of Plant Sciences and Genetics in Agricultures at The Hebrew University of Jerusalem · Department of Biology at University of Oxford, Oxford, United Kingdom · Cary Institute of Ecosystem Studies
q-bio.OT
Submitted: 2026-07-09
Updated: 2026-11-03
License: http://creativecommons.org/licenses/by/4.0/
Importance score: 97/100
The gist: Disturbance nearly doubled the proportion of annual species and more than doubled the relative cover of annuals…
Key concepts
- Disturbance
- Physical disturbance involves removing established vegetation and disrupting shallow belowground structures in grassland plots. This process creates opportunities for rapid regeneration of species that grow quickly from seeds, shifting the community balance away from long-lived perennial plants.
- Annual Dominance
- This refers to a situation where annual plants become the most abundant species in an ecosystem. In undisturbed systems, this is limited to areas with extremely hot and dry summers. Disturbance relaxes these climatic constraints, allowing annuals to dominate in a wider range of temperature and precipitation conditions.
- Functional Group Variation
- The effect of disturbance on the annual-perennial balance differs depending on the plant's functional group (e.g., graminoids vs. forbs). In some groups, like legumes, perennial cover is unaffected by disturbance, but in others, like forbs, annual cover roughly doubles.
- Climatic Envelope Expansion
- This describes how disturbance widens the range of climate conditions where annual plants can thrive. By relaxing constraints on temperature and precipitation that usually restrict annuals to hot/dry summers, disturbance expands the total climatic space in which annual dominance is possible.
Terminology
Summary
Disturbance nearly doubled the proportion of annual species and more than doubled the relative cover of annuals, whereas fertilisation had little influence and did not interact with disturbance
Summary
Disturbance nearly doubled the proportion of annual species and more than doubled the relative cover of annuals, whereas fertilisation had little influence and did not interact with disturbance
Life History Theory Context
Classical lifehistory theory predicts that disturbance should shift communities from long-lived perennials toward short-lived annuals, and that nutrient enrichment may amplify this shift Annual species constitute ∼15% of the world’s herbaceous flora, with substantially higher representation in hot, dry climates Annuals often allocate more to rapid aboveground growth and seed production, while perennials tend to invest more in storage organs and belowground structures Disturbance can create these conditions by reducing survival of established individuals, particularly long-lived perennials, while opening recruitment windows for species that regenerate rapidly from seeds
Experimental Findings on Drivers
The study tested how physical disturbance (vegetation removal and shallow soil tillage) and fertilisation reshape annual–perennial balance, and whether disturbance relaxes the climatic limits of annual dominance Disturbance nearly doubled the proportion of annual species and more than doubled the relative cover of annuals Fertilisaton had little influence and did not interact with disturbance The disturbance-driven shift arose through contrasting pathways: in graminoids and legumes, it reflected the loss of perennial cover, while in forbs, the expansion of annual cover
Climatic Envelope Expansion
In the absence of disturbance, annual dominance was restricted to systems with extremely hot and dry summers, but disturbance nearly tripled the extent of climate space in which annuals dominated Disturbance expands the climatic limits of annual dominance by relaxing constraints that otherwise restrict annual dominance to sites with high temperatures and low precipitation during the warmest season The proportion of climatic space dominated by annuals increased from 11% to 29% when measured by species share, and from 12% to 34% when measured by relative cover
Functional Group Variation
Disturbance shifted annual–perennial balance through different pathways across functional groups In graminoids and legumes, the shift was driven mainly by reductions in perennial cover, although annual graminoid cover also increased modestly In forbs, perennial cover showed no clear response to disturbance, while annual forb cover roughly doubled Some of this variation is expected because annual prevalence is bounded between 0 and 100% leaving less scope for increase where annuals are already common
Ecosystem Implications
Annual plants can buffer grasslands against repeated disturbance and increasing droughts by rapidly reassembling from seed and sustaining vegetation cover and biomass production Perennial dominance and extensive belowground biomass are associated with many ecosystem services provided by grasslands, including long-term carbon storage, erosion control, and regulation of water runoff As disturbance regimes intensify and climates become more favorable to annuals, the key risk is erosion of perennial dominance, with consequences for long-term ecosystem functioning The findings advance understanding of the broad range of conditions under which disturbance promotes annual dominance, helping to predict and manage grassland responses to global change The paper concludes that disturbance can shift annual–perennial balance consistently across grasslands worldwide and expand the climatic conditions under which annual strategies dominate
Methods Overview
The analysis utilized the coordinated Disturbance and Recovery Across Global Grasslands Network (DRAGNet 27), which applies common experimental treatments and vegetation sampling protocols across 37 grassland sites The experiment applied common experimental treatments and vegetation sampling protocols across 37 grassland sites At each site, DRAGNet follows a randomized block design with 3–5 blocks Each block contains one 5 × 5, m plot assigned to each of four treatments used here: an unmanipulated Control, Disturbance, Nutrient Addition (NPK), and their combination (Disturbance + NPK) The Disturbance treatment is applied once per year for three consecutive years and is designed to create conditions for community reassembly by removing established vegetation and disrupting shallow belowground structures Within each 5 × 5 m plot, vegetation is sampled once per year during the growing season in a permanently marked 1 m2 quadrat located at least 1 m inside the plot boundaries We analysed the first three years after treatment initiation at each site, corresponding to the period during which the annual disturbance treatment was actively applied
Statistical Modeling
Network-wide treatment effects were estimated using generalized linear mixed models (GLMMs) to plot-level data These GLMMs allowed for combining non-Gaussian distributions (Beta, negative binomial, lognormal), zero-inflation components, and hierarchical random effects suitable for the multi-site, repeated-measures design Site-level variation in disturbance effects was quantified using the log response ratio and additive difference in percentage points Climate–disturbance models were fitted using zero–oneinflated Beta regression (BEINF) to model site-level proportions as functions of climate, identifying mean temperature of the warmest quarter and precipitation in the warmest quarter as the best predictors The analysis focused on these two variables because they emerged as the best predictors for both responses Models were fitted separately for each treatment to avoid imposing a restrictive parametric form on what are likely to be complex and treatment-specific interactions between climate and disturbance or fertilisation The expected mixture mean, E[Y] = (1 − ν − τ) µ + τ, was evaluated on a regular grid covering the convex hull of the observed climate space to visualize the climatic “dominance zone” of annuals
Data and Site Details
The dataset comprised 37 sites and 1,425–1,444 plot–year observations The study sites span a broad climatic gradient, from mean annual temperatures of approximately −5 to 25◦C and annual precipitation from roughly 200 to 1700 mm Data availability is provided through the Environmental Data Initiative repository at https://doi.org/10.6073/pasta/65799ea6ae60abdf0590b35134a9ab43 The authors acknowledge funding to E. Borer and E. Seabloom from the National Science Foundation Research Coordination Network (NSF-DEB-2311608) and the Long Term Ecological Research program (DEB-1831944) to the Cedar Creek LTER The work was generated using data from the Disturbance and Recovery Across Grasslands experiment (DRAGNet), funded at the site-scale by individual researchers
Key Statistical Results
The Continuous Model for relative annual cover showed that Disturbance had a significant effect, with an estimated coefficient of 2.71 The Zero-Inflated Model for relative annual cover indicated that Disturbance had a significant effect, with an estimated coefficient of 0. p. 6 a d: *
f: ***
d × f: **
0
20
40
60
Absolute cover (%)
Annual graminoids
b d: ***
f: *
d × f: *
Perennial graminoids
c d: ***
f: ns
d × f: ns
0
10
20
30
40
Absolute cover (%)
Annual forbs
d d: ns
f: ***
d × f: *
Perennial forbs
e d: ns
f: ns
d × f: ns
0
5
10
15
20
Control
Disturbance
NPK
Disturbance+NPK
Absolute cover (%)
Annual legumes
f d: ***
f: ns
d × f: ns
Control
Disturbance
NPK
Disturbance+NPK
Perennial legumes
Figure 3: Disturbance shifts annual–perennial balance through different
functional-group pathways. Bars show model-based marginal means of absolute
cover with 95% confiIdence intervals, estimated from the continuous component of
zero-inflated lognormal mixed-effects models; values therefore represent the percentage of plot area covered by each functional-group when present, i. a d: *
f: ***
d × f: **
0
20
40
60
Absolute cover (%)
Annual graminoids
b d: ***
f: *
d × f: *
Perennial graminoids
c d: ***
f: ns
d × f: ns
0
10
20
30
40
Absolute cover (%)
Annual forbs
d d: ns
f: ***
d × f: *
Perennial forbs
e d: ns
f: ns
d × f: ns
0
5
10
15
20
Control
Disturbance
NPK
Disturbance+NPK
Absolute cover (%)
Annual legumes
f d: ***
f: ns
d × f: ns
Control
Disturbance
NPK
Disturbance+NPK
Perennial legumes
Figure 3: Disturbance shifts annual–perennial balance through different
functional-group pathways. Bars show model-based marginal means of absolute
cover with 95% confiIdence intervals, estimated from the continuous component of
zero-inflated lognormal mixed-effects models; values therefore represent the percentage of plot area covered by each functional-group when present, i. a d: *
f: ***
d × f: **
0
20
40
60
Absolute cover (%)
Annual graminoids
b d: ***
f: *
d × f: *
Perennial graminoids
c d: ***
f: ns
d × f: ns
0
10
20
30
40
Absolute cover (%)
Annual forbs
d d: ns
f: ***
d × f: *
Perennial forbs
e d: ns
f: ns
d × f: ns
0
5
10
15
20
Control
Disturbance
NPK
Disturbance+NPK
Absolute cover (%)
Annual legumes
f d: ***
f: ns
d × f: ns
Control
Disturbance
NPK
Disturbance+NPK
Perennial legumes
Figure 3: Disturbance shifts annual–perennial balance through different
functional-group pathways. Bars show model-based marginal means of absolute
cover with 95% confiIdence intervals, estimated from the continuous component of
zero-inflated lognormal mixed-effects models; values therefore represent the percentage of plot area covered by each functional-group when present, i. a d: *
f: ***
d × f: **
0
20
40
60
Absolute cover (%)
Annual graminoids
b d: ***
f: *
d × f: *
Perennial graminoids
c d: ***
f: ns
d × f: ns
0
10
20
30
40
Absolute cover (%)
Annual forbs
d d: ns
f: ***
d × f: *
Perennial forbs
e d: ns
f: ns
d × f: ns
0
5
10
15
20
Control
Disturbance
NPK
Disturbance+NPK
Absolute cover (%)
Annual legumes
f d: ***
f: ns
d × f: ns
Control
Disturbance
NPK
Disturbance+NPK
Perennial legumes
Figure 3: Disturbance shifts annual–perennial balance through different
functional-group pathways. Bars show model-based marginal means of absolute
cover with 95% confiIdence intervals, estimated from the continuous component of
zero-inflated lognormal mixed-effects models; values therefore represent the percentage of plot area covered by each functional-group when present, i. a d: *
f: ***
d × f: **
0
20
40
60
Absolute cover (%)
Annual graminoids
b d: ***
f: *
d × f: *
Perennial graminoids
c d: ***
f: ns
d × f: ns
0
10
20
30
40
Absolute cover (%)
Annual forbs
d d: ns
f: ***
d × f: *
Perennial forbs
e d: ns
f: ns
d × f: ns
0
5
10
15
20
Control
Disturbance
NPK
Disturbance+NPK
Absolute cover (%)
Annual legumes
f d: ***
f: ns
d × f: ns
Control
Disturbance
NPK
Disturbance+NPK
Perennial legumes
Figure 3: Disturbance shifts annual–perennial balance through different
functional-group pathways. Bars show model-based marginal means of absolute
cover with 95% confiIdence intervals, estimated from the continuous component of
zero-inflated lognormal mixed-effects models; values therefore represent the percentage of plot area covered by each functional-group when present, i. a d: *
f: ***
d × f: **
0
20
40
60
Absolute cover (%)
Annual graminoids
b d: ***
f: *
d × f: *
Perennial graminoids
c d: ***
f: ns
d × f: ns
0
10
20
30
40
Absolute cover (%)
Annual forbs
d d: ns
f: ***
d × f: *
Perennial forbs
e d: ns
f: ns
d × f: ns
0
5
10
15
20
Control
Disturbance
NPK
Disturbance+NPK
Absolute cover (%)
Annual legumes
f d: ***
f: ns
d × f: ns
Control
Disturbance
NPK
Disturbance+NPK
Perennial legumes
Figure 3: Disturbance shifts annual–perennial balance through different
functional-group pathways. Bars show model-based marginal means of absolute
cover with 95% confiIdence intervals, estimated from the continuous component of
zero-inflated lognormal mixed-effects models; values therefore represent the percentage of plot area covered by each functional-group when present, i. a d: *
f: ***
d × f: **
0
20
40
60
Absolute cover (%)
Annual graminoids
b d: ***
f: *
d × f: *
Perennial graminoids
c d: ***
f: ns
d × f: ns
0
10
20
30
40
Absolute cover (%)
Annual forbs
d d: ns
f: ***
d × f: *
Perennial forbs
e d: ns
f: ns
d × f: ns
0
5
10
15
20
Control
Disturbance
NPK
Disturbance+NPK
Absolute cover (%)
Annual legumes
f d: ***
f: ns
d × f: ns
Control
Disturbance
NPK
Disturbance+NPK
Perennial legumes
Figure 3: Disturbance shifts annual–perennial balance through different
functional-group pathways. Bars show model-based marginal means of absolute
cover with 95% confiIdence intervals, estimated from the continuous component of
zero-inflated lognormal mixed-effects models; values therefore represent the percentage of plot area covered by each functional-group when present, i. a d: *
f: ***
d × f: **
0
20
40
60
Absolute cover (%)
Annual graminoids
b d: ***
f: *
d × f: *
Perennial graminoids
c d: ***
f: ns
d × f: ns
0
10
20
30
40
Absolute cover (%)
Annual forbs
d d: ns
f: ***
d × f: *
Perennial forbs
e d: ns
f: ns
d × f: ns
0
5
10
15
20
Control
Disturbance
NPK
Disturbance+NPK
Absolute cover (%)
Annual legumes
f d: ***
f: ns
d × f: ns
Control
Disturbance
NPK
Disturbance+NPK
Perennial legumes
Figure 3: Disturbance shifts annual–perennial balance through different
functional-group pathways. Bars show model-based marginal means of absolute
cover with 95% confiIdence intervals, estimated from the continuous component of
zero-inflated lognormal mixed-effects models; values therefore represent the percentage of plot area covered by each functional-group when present, i. a d: *
f: ***
d × f: **
0
20
40
60
Absolute cover (%)
Annual graminoids
b d: ***
f: *
d × f: *
Perennial graminoids
c d: ***
f: ns
d × f: ns
0
10
20
30
40
Absolute cover (%)
Annual forbs
d d: ns
f: ***
d × f: *
Perennial forbs
e d: ns
f: ns
d × f: ns
0
5
10
15
20
Control
Disturbance
NPK
Disturbance+NPK
Absolute cover (%)
Annual legumes
f d: ***
f: ns
d × f: ns
Control
Disturbance
NPK
Disturbance+NPK
Perennial legumes
Figure 3: Disturbance shifts annual–perennial balance through different
functional-group pathways. Bars show model-based marginal means of absolute
cover with 95% confiIdence intervals, estimated from the continuous component of
zero-inflated lognormal mixed-effects models; values therefore represent the percentage of plot area covered by each functional-group when present, i. Bars show model-based marginal means of absolute
cover with 95% confiIdence intervals, estimated from the continuous component of
zero-inflated lognormal mixed-effects models; values therefore represent the percentage of plot area covered by each functional-group when present, i. conditional on
non-zero cover. a d: *
f: ***
d × f: **
0
20
40
60
Absolute cover (%)
Annual graminoids
b d: ***
f: *
d × f: *
Perennial graminoids
c d: ***
f: ns
d × f: ns
0
10
20
30
40
Absolute cover (%)
Annual forbs
d d: ns
f: ***
d × f: *
Perennial forbs
e d: ns
f: ns
d × f: ns
0
5
10
15
20
Control
Disturbance
NPK
Disturbance+NPK
Absolute cover (%)
Annual legumes
f d: ***
f: ns
d × f: ns
Control
Disturbance
NPK
Disturbance+NPK
Perennial legumes
Figure 3: Disturbance shifts annual–perennial balance through different
functional-group pathways. Bars show model-based marginal means of absolute
cover with 95% confiIdence intervals, estimated from the continuous component of
zero-inflated lognormal mixed-effects models; values therefore represent the percentage of plot area covered by each functional-group when present, i. Bars show model-based marginal means of absolute
cover with 95% confiIdence intervals, estimated from the continuous component of
zero-inflated lognormal mixed-effects models; values therefore represent the percentage of plot area covered by each functional-group when present, i. conditional on
non-zero cover. a d: *
f: ***
d × f: **
0
20
40
60
Absolute cover (%)
Annual graminoids
b d: ***
f: *
d × f: *
Perennial graminoids
c d: ***
f: ns
d × f: ns
0
10
20
30
40
Absolute cover (%)
Annual forbs
d d: ns
f: ***
d × f: *
Perennial forbs
e d: ns
f: ns
d × f: ns
0
5
10
15
20
Control
Disturbance
NPK
Disturbance+NPK
Absolute cover (%)
Annual legumes
f d: ***
f: ns
d × f: ns
Control
Disturbance
NPK
Disturbance+NPK
Perennial legumes
Figure 3: Disturbance shifts annual–perennial balance through different
functional-group pathways. Bars show model-based marginal means of absolute
cover with 95% confiIdence intervals, estimated from the continuous component of
zero-inflated lognormal mixed-effects models; values therefore represent the percentage of plot area covered by each functional-group when present, i. Bars show model-based marginal means of absolute
cover with 95% confiIdence intervals, estimated from the continuous component of
zero-inflated lognormal mixed-effects models; values therefore represent the percentage of plot area covered by each functional-group when present, i. conditional on
non-zero cover. a d: *
f: ***
d × f: **
0
20
40
60
Absolute cover (%)
Annual graminoids
b d: ***
f: *
d × f: *
Perennial graminoids
c d: ***
f: ns
d × f: ns
0
10
20
30
40
Absolute cover (%)
Annual forbs
d d: ns
f: ***
d × f: *
Perennial forbs
e d: ns
f: ns
d × f: ns
0
5
10
15
20
Control
Disturbance
NPK
Disturbance+NPK
Absolute cover (%)
Annual legumes
f d: ***
f: ns
d × f: ns
Control
Disturbance
NPK
Disturbance+NPK
Perennial legumes
Figure 3: Disturbance shifts annual–perennial balance through different
functional-group pathways. Bars show model-based marginal means of absolute
cover with 95% confiIdence intervals, estimated from the continuous component of
zero-inflated lognormal mixed-effects models; values therefore represent the percentage of plot area covered by each functional-group when present, i.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Climate niche prediction for annual dominance: The improved AI system can predict whether a given climate location will be a climatic
dominance zone
of annuals by calculating the expected proportion of annuals using BEINF models based on mean temperature and precipitation, as shown in Table S22. -
Disturbance response modeling: The system can quantify how physical disturbance affects annual prevalence across different functional groups by analyzing the
Disturbance × NPK
interaction terms in the zero-inflated beta GLMMs (Table S4). -
Fertilizer effect assessment: The improved AI can determine if nutrient enrichment systematically favors annuals by examining the treatment coefficients for NPK in the species share and relative cover models, noting that
fertilisation had little influence and did not interact with disturbance
in some contexts. -
Functional group specific dynamics: The system can predict how disturbance reshapes the annual–perennial balance by distinguishing between graminoids, legumes, and forbs based on the observed pathways described in Figure 3.
-
Climatic constraint relaxation: The AI can model how physical disturbance
markedly expanded this zone
in which annual dominance occurs, demonstrating thatdisturbance expands the climatic limits of annual dominance.
Abstract
Disturbance regimes and nutrient inputs are changing worldwide, with consequences for the structure and functioning of plant communities. Life-history theory predicts that disturbance should shift communities from long-lived perennials toward short-lived annuals, and that nutrient enrichment may amplify this shift. However, these predictions have not been tested experimentally across broad environmental gradients. Here, using a global grassland experiment spanning 37 sites, we tested how severe physical disturbance (vegetation removal and shallow soil tillage) and fertilisation alter the annual-perennial balance, and whether disturbance expands the climatic envelope of annual dominance. Disturbance increased annual species share from 15% to 29% and relative annual cover from 9% to 29%, whereas fertilisation had little influence and did not interact with disturbance. The disturbance-driven shift arose through contrasting pathways: in graminoids and legumes, perennial cover declined, while in forbs, annual cover expanded. In the absence of disturbance, annual dominance was restricted to systems with extremely hot and dry summers, yet disturbance consistently increased annual prevalence across climatic gradients and doubled the extent of climate space in which annuals dominated. During repeated disturbance, annuals may help maintain vegetation cover, but their expansion also signals loss of perennial cover and associated ecosystem functions.
Related papers
- Current validation practice undermines surgical AI development
- QUAREP-LiMi: A community-driven initiative to establish guidelines for quality assessment and reproducibility for instruments and images in light microscopy
- FISH Omics Format: a standard enabling FAIR chromatin tracing data sharing across repositories
- Cancer genomics, targeted therapy to precision oncology: the treatment paradigm is a holistic approach to cancer therapeutics
- Building Barricades: Red-Teaming Agentic Biosynthesis Planning
- The Kinetic Basis of the Genetic Code