A systematic review by researchers from the University of Cape Town and the South African Environmental Observation Network finds major gaps and geographic biases in observational evidence of how plant communities in Mediterranean- type ecosystems (MTEs) are responding to climate change.

MTEs occur in five regions: the Mediterranean Basin, California, central Chile, southwestern Australia, and South Africa’s Cape (including the Cape Floristic Region).

These areas share a climate of hot, dry summers and cooler, wetter winters and are global biodiversity hotspots with high endemism. They face strong climate pressures, including extreme droughts reported by the IPCC across all five regions (notably severe in South Africa and California).

Much existing knowledge relies on models and experiments rather than direct field observations of natural plant communities.

Main findings

• Researchers screened databases and identified only 128 qualifying observational studies. Just 9% were from the Global South, and southern-hemisphere studies overall made up only about 17%.
• Geographic bias is pronounced: Mediterranean Basin (88 studies), California (18), southwestern Australia (10), central Chile (8), and South Africa’s Cape (only 4).
• Only 11 of the 128 studies appeared in IPCC assessment reports; in the latest IPCC land-ecosystems chapter, less than 3% of cited literature concerned these regions.
• Most frequently reported disturbances: climate variability, drought (dominant everywhere, 67-100% of reports by region), wildfire, frost, and pathogens/insect infestations. Wildfire was noted in most regions except Chile (despite frequent fires there); pathogen and insect studies were limited to the Northern Hemisphere; frost mainly in southwestern Australia.
• Studies mostly focused on vegetation condition and growth; far fewer examined plant diversity, phenology, or physiology. Trees and shrubs were overrepresented relative to grasses and wildflowers (which contribute heavily to diversity and can influence fire). Some response types were missing entirely in particular regions.
• Confidence was relatively high for links between climate- related disturbances and vegetation responses, but lower for attributing observed biological changes specifically to climate change (versus other drivers such as fire, land use, herbivory, invasive species, or natural variability). Few studies used modern causal-inference methods; only four included causal diagrams, and none scored highly on this.

Implications and recommendations

The geographic and topical imbalances make it hard to distinguish real ecological differences between regions from artifacts of research effort. Findings from well- studied areas (e.g., California or the Mediterranean Basin) may not apply to understudied ones like the Cape Floristic Region.

The authors call for coordinated research across all five MTEs that combines long- term monitoring, reanalysis of existing and historical datasets, remote sensing, experiments, and stronger causal- inference approaches. This would improve understanding of resilience mechanisms and better support conservation and adaptation in these biodiversity hotspots.

In short, while models and experiments provide projections, direct observational evidence of climate- driven changes in MTE plant communities remains sparse, unevenly distributed, and often insufficient to firmly attribute causes, creating significant blind spots for science and management.

The Mediterranean Basin is by far the best- studied of the five Mediterranean- type ecosystems (MTEs) in the review, accounting for the large majority of the 128 observational studies (approximately 88).

This gives it a relatively stronger evidence base than the other regions (especially the Global South ones), but the paper still identifies important implications, limitations, and opportunities specific to it.

Stronger observational evidence base, but incomplete

• Drought, climate variability, wildfire, and (to a lesser extent) pathogens and insect infestations are well- documented disturbances affecting plant communities.
• Most studies focus on vegetation condition and growth. Responses related to diversity, phenology, and physiology are less comprehensively covered than condition and growth (though better represented here than in some other MTEs, where certain categories are entirely absent).
• Confidence scores for studies from the Mediterranean Basin (and California) span a wide range but often include high values for clear prior expectations and hypotheses, data quality, and statistical analyses. Scores tend to be higher for understanding relationships between disturbances and responses than for rigorously detecting climate -change-driven biological change itself.

Key limitations and risks of over- generalisation

Even with the largest number of studies, the evidence is not comprehensive enough for full causal attribution in many cases. Few studies overall (and none achieving the top causal- inference score in the review’s modified confidence index) used formal causal diagrams (e.g., directed acyclic graphs) or modern causal- inference methods to isolate climate change from confounders such as land use, fire management, invasive species, herbivory, or natural variability.

This means that while many observed changes coincide with climate trends, stronger proof that climate change is the primary driver is often lacking. Findings from the Mediterranean Basin cannot automatically be assumed to apply to the other four MTEs (or even uniformly across the Basin itself), because ecological differences, fire regimes, species pools, and research priorities vary.

Implications for science, management, and policy in the Mediterranean Basin

• Scientific: The relatively rich observational record makes the Basin an important reference region. There is a clear opportunity to re- analyse existing long- term datasets with stronger causal methods, integrate them more systematically with remote sensing, experiments, and models, and expand underrepresented topics (especially diversity, phenology, physiology, and non- woody growth forms such as graminoids and forbs, which contribute heavily to diversity and can influence fire behaviour).
• Cross- regional learning: Because the Basin dominates the literature, its findings currently shape much of the global narrative on MTE climate impacts. A collaborative programme across all five MTEs (emphasising shared resilience mechanisms) would help test which responses are truly general versus region- specific and improve transferability of insights.
• Management and adaptation: Better causal understanding would support more targeted interventions for conservation, fire management, and climate adaptation in this biodiversity hotspot. The review stresses that identifying shared drivers of resilience and vulnerability across MTEs can strengthen management interventions, climate adaptation/mitigation efforts, and conservation action.
• IPCC and global assessments: Only a small fraction of the observational studies (11 of 128 overall) fed into IPCC reports. Strengthening the observational evidence base and causal attribution from the Mediterranean Basin (and ensuring it is better represented) would improve the robustness of future assessments for this region and for MTEs more broadly.

In short, the Mediterranean Basin benefits from the densest observational evidence among MTEs, providing a solid foundation for understanding drought, variability, and fire impacts on vegetation condition and growth.

However, gaps in causal attribution, certain response types, and cross- regional comparability remain.

Addressing these through improved methods, long- term monitoring coordination, and collaborative research would enhance both local management and the ability to generalise (or appropriately differentiate) findings to the other MTEs.

A systematic review of observational studies reporting climate change impacts on plant communities in Mediterranean-type ecosystems

Abstract summary

Mediterranean- type ecosystems (MTEs) are projected to be among the most vulnerable terrestrial ecosystems to climate change. While plant community responses have been extensively studied via experiments and models, observational studies remain relatively scarce.

This systematic review synthesises observed climate-change impacts on plant communities across the five global MTEs (Mediterranean Basin, California, central Chile, southwestern Australia, and the Cape Floristic Region of South Africa). It covers key disturbances and drivers of vegetation change, biological responses, geographic patterns, and confidence in the evidence.

Key findings:

• Only 128 publications met the strict inclusion criteria (observational studies of climate- related disturbances affecting plant communities; experiments, pure modelling and projections, paleo-studies, agricultural systems, etc., were excluded).
• Just 11 of these appeared in IPCC assessment reports.
• Strong geographic bias: only 9% from the Global South. Disturbances and response types were unevenly reported across regions.
• Most common disturbances: climate variability, drought, wildfire, frost, and pathogens and insect infestations, studied mainly in relation to vegetation condition and growth. Diversity, phenology, and physiology were less frequently examined (and sometimes absent in particular regions).
• A confidence index (adapted from O’Connor et al. 2015, with additions for reporting quality and causal inference) showed relatively high scores for understanding links between disturbances and observed responses, but lower scores for confidently detecting climate- change- induced biological changes.
• Few studies applied modern causal- inference methods. Differences among regions may partly reflect research effort and capacity rather than genuine ecological differences, limiting transferability of knowledge.
• Opportunity exists to re-analyse existing long- term datasets with stronger causal approaches.

Policy implications: A collaborative, integrative research programme focused on shared resilience mechanisms across MTEs would improve understanding of causal relationships, support better- targeted management, climate adaptation and mitigation, and conservation in these biodiversity hotspots.

Additional context from the paper and related coverage

The authors searched Web of Science and Scopus (up to September 2024), supplemented by literature cited in recent IPCC WGII reports, and applied rigorous PECO- based eligibility criteria and multi- stage screening. They extracted detailed data on study design, systems, disturbances, and responses, and scored confidence on multiple axes (expectations and alternative causes, data quality, statistics, reporting, and causal inference).

Press summaries of the work (University of Cape Town and SAEON) provide the regional breakdown of the 128 studies:

Mediterranean Basin ≈ 88, California ≈ 18, southwestern Australia ≈ 10, central Chile ≈ 8, Cape Floristic Region ≈ 4. Drought and climate variability dominated reports in every region.

The full paper is available at the DOI link above (open access). Supporting information, data, and code are typically deposited with the journal or associated repositories (e.g., OSF registration of the review protocol exists from earlier stages of the project).

Published: Ecological Solutions and Evidence – Wiley Online, (2026), first published 4 August 2026

DOI: 10.1002/2688-8319.70304

Provided: University of Cape Town

Authors: Hana Petersen, Ryan Blanchard, Jasper Slingsby

Abstract

1. Context. Mediterranean-type ecosystems (MTEs) are projected to be among the most vulnerable terrestrial ecosystems to global climate change. Plant community responses in MTEs to climate change have been well researched using experimental and modelling approaches, yet observational studies are less common.
2. Approach. We present a systematic review of observed climate change impacts on plant communities across the five global MTEs, highlighting key disturbances and drivers of vegetation change, biological responses to these disturbances, and prevailing geographic trends.
3. Literature trends. A relatively small number of publications met our criteria (N_pub = 128), of which only 11 were reported in assessment reports by the Intergovernmental Panel on Climate Change. The publications exhibited a large geographic bias, with only 9% from the Global South. Reported disturbances and responses were unevenly represented across MTEs.
4. Drivers and responses. Climate variability, drought, wildfire, frost, and pathogens/insect infestations were the most reported disturbances, largely studied in relation to vegetation condition and growth. Studies on diversity, phenology and physiology were less common, and absent in some MTEs.
5. Confidence assessment. An index assessing confidence in studies’ methods and findings revealed high scores for understanding relationships between climate change impacts and observed responses, and lower scores for detection of climate change-induced biological responses. The extent to which reported differences among regions reflect different effects of climate change and biological responses versus research focus and capacity remains unclear. While the included literature has improved our understanding of observed climate change impacts on plant communities in each MTE, the transferability of this knowledge across regions is limited by unequal reporting between regions. Since few studies applied causal inference methods for observational data, there is an opportunity to reanalyse existing datasets to improve our understanding of causal relationships.
6. Policy implications. Developing a collaborative, integrative research programme that considers common resilience mechanisms among MTEs will aid our understanding of causal relationships among climate change-induced disturbances and vegetation responses. By identifying shared drivers of resilience and vulnerability across MTEs, such research can support more targeted and effective management interventions, strengthen climate adaptation and mitigation efforts, and inform conservation action in these global biodiversity hotspots.