Introduction

A new manuscript from SPECIAL group PhD student Yicheng Shen has recently been published in Ecological Indicators, discussing the factors affecting post-fire recovery. You can read on for a short summary of the paper below or read the exciting new research here!

What is post-fire recovery?

Post-fire recovery can simply be described as the time taken for vegetation to recover after a fire. This is an important vegetation property because it impacts the ability of a plant to re-grow after a disturbance, photosynthesise, and sequester carbon; having implications for the global carbon cycle. Due to the scale at which recovery can impact land-atmosphere interactions it is important to understand the properties influencing recovery.

Post-fire recovery has already been to show to vary considerably globally, however, most studies on this topic are completed in regional and local scale. The literature suggests fire intensity, fire size and heterogeneity,  topographic factors, and climate, are important factors impacting recovery. Additionally, it is clear that vegetation properties such as serotiny and resprouting will influence post-fire recovery however, little has been done to quantify this.

A Novel Approach to Estimating Post-Fire Recovery

Typically, remote-sensing data is used to study post-fire recovery as it can produce a measure of vegetation before and after fire. However, accurate measurements can be challenging to obtain in areas of dense vegetation. Solar-induced chlorophyll fluorescence (SIF) provides a direct measure of photosynthetic activity and is a valuable tool for monitoring vegetation changes at large scales. Studies have shown that SIF is more sensitive than traditional vegetation indices (e.g., NDVI, EVI) in capturing post-fire recovery, making it a reliable method for assessing fire and drought impacts. In this study, SIF is used to estimate post-fire recovery time for over 10,000 fires globally, as photosynthetic recovery is a key step toward full ecosystem regeneration.

Individual fires were obtained from The Fire Atlas along for study with their fire characteristics e.g. fire size, fire intensity. Additional factors included in the analysis include climate factors, vegetation factors such as gross primary productivity (GPP), human activity and topographic factors. A relaxed lasso approach was first taken to identify variables important for post-fire recovery and then a linear model was fit to further assess variable importance and their relationships with recovery time.

Key Findings

Of those variables included in the analysis, 19 were determined to be importance for post-fire recovery. However, some variables had a strong collinearity with others and were removed during the modelling process. This resulted in a linear model fit to 11 total variables. A summary of the final model, fitted globally, can be found in Table 2 of the paper.

“Post-fire recovery of photosynthetic activity is relatively fast: in 75 % of the cases recovery occurred in < 4 years.”

Core variables in the model:

  • GPP was identified to be the most important of these variables which intuitively makes sense as higher production may enable faster recovery.
  • Fire intensity and duration are key, more severe fires = longer recovery times.
  • Inclusion of biome improves the model fit showing a clear patter of hot and dry climates having a short recovery time, whilst cold climates demonstrate slower recovery times.
  • Resprouting is a vegetation trait associated with fire-adapted species. When assessing the percentage of resprouters in biomes, the same pattern can be found indicating that there may be a relationship between fire-adapted traits and recovery times.
  • Contrasting past studies, topographic factors were not found to be important and human activity was only found to have an impact in non-fire adapted biomes.

Figure 1: Post-fire photosynthetic recovery times across biomes after controlling for other factors influencing recovery time. The number of fire candidates in each biome is shown in brackets. Significant differences between biomes, as measured by the t-test, are indicated by the red letters. (Taken from, Shen et al. 2025, Figure 4, pg. 7)

Overall, this study provides an exciting step towards developing the next generation of fire-vegetation models. Specially, within the SPECIAL group this study will be used as we continue to develop a fire-vegetation interactions model as part of the LEMONTREE and Leverhulme Fire-Vegetation Interactions Working Group. Fire damage and subsequent recovery will become an important limitation on biomass production and carbon cycling in future modelling efforts.

Keep an eye out on the LEMONTREE website for fire-vegetation working group updates as well as the SPECIAL site for more publications from the group!