Recovery of European temperate forests after large and severe disturbances
Data files
Jan 08, 2024 version files 250.96 KB
-
Post_disturbance_recovery_dataset.csv
246.46 KB
-
README.md
4.50 KB
Abstract
As climate change progresses, there is increasing concern that large and severe disturbances may diminish the resilience of forest ecosystems and alter their recovery dynamics. We investigated recovery of temperate forests in Europe following large and severe disturbance events (more than 70 % canopy cover reduction in patches larger than 1 ha) that span a time since disturbance range of one to five decades. Across a ground-based plot network at 143 sites, featuring various forest types and management practices, subjected to 28 disturbance events, including windthrow, fire, and bark beetle, we studied post-disturbance tree density and composition, which are key indicators of forest resilience. We used this dataset to qualitatively assess forest recovery in structure and composition by comparing plot-level post-disturbance height-weighted densities with site-specific pre-disturbance densities. Additionally, we analyzed ecological drivers of post-windthrow tree density, including forest management, topography, and bioclimate, using a series of generalized additive models. The present dataset includes the plot-level data necessary to carry out the aforementioned analyses.
# Recovery of European temperate forests after large and severe disturbances
The csv file “Post_disturbance_recovery_dataset” include all the data necessary to carry out the analyses performed in the manuscript that we are going to submit to the journal “Global Change Biology”. Analyses include the statistical modeling of tree densities (i.e.response variables) after wind disturbances and their climatic, topographic and management drivers (i.e.explanatory variables); and the assessment of forest recovery in structure and composition, by comparing forest density and composition before and after the disturbance.
Description of the data and file structure
The csv file “Post_disturbance_recovery_dataset” include plot-level data
to carry out the aforementioned analyses:
- “site” is the site identifier
- “plot” is the plot alphanumeric identifier
- “Pre-disturbance management” indicates the type of forest management
before the disturbance happened, it’s a categorical variable with 3
levels: “Y” if managed at the time of the disturbance, “L” if the
forest was managed until less than 50 years before the disturbance,
“N” if the forest was unmanaged - “Post-disturbance management” indicates the type of forest
management after the disturbance happened, it’s a categorical
variable with 3 levels: “None” if there was no post-disturbance
management, “Logging only” if only salvage logging (i.e.removal of
biomass damaged by a disturbance) was carried out, “Intensive” if
salvage logging was carried out together with tree planting - “Years since disturbance” is a discrete variable indicating the
number of years passed between when the disturbance happened and
when the plot sampling was carried out - “Disturbance agent” is a categorical variable with 3 levels:
“BarkBeetle”, “Fire”, “Wind” - “Elevation” is a continuous variable indicating the elevation of the
plot center, unit measure is m a.s.l. - “Aridity Change” is a continuous variable. Yearly aridity index (AI)
was calculated as the average of monthly June, July, August aridity
indices, which were computed as 1- P/pev, with P for Total
precipitation, and pev for Potential evaporation. Aridity change was
calculated as the difference between the mean AI of the 7 years
post-disturbance and the aridity baseline, calculated as average of
the index for years between 1950 and 1980. It’s a proxy for
post-disturbance summer drought anomalies, which could have a
negative impact on seedlings and small trees - “HLI” is a continuous variable standing for Heat Load Index,
calculated after McCune & Keon (2002). It’s a proxy for
topographically driven potential heat load - Eight columns featuring the response variables, related to the
density per hectare, used in statistical modeling: combinations of
raw densities (“RawDensity”, i.e.number of individual trees) and
weighted densities (“WDensity”, i.e.individual trees were weighted
according to their respective heights, then summed up, following a
scheme similar to Vickers et al.(2019), with the difference that we
gave the weight of 1 to any trees larger than 7 cm in DBH or taller
than 7.54 m), with tree species groups (“Total” and according to
their role in forest succession, i.e.”Pioneer”,
“Early-successional”, and “Late-successional”) - “Pre_Density” indicates the tree density per hectare of the forest
before the disturbance happened - “Pre_Sp1” and “Pre_Sp2” are categorical variables that indicate the
dominant and co-dominant tree species present before the disturbance
happened, respectively. In case that no co-dominant species was
present Pre_Sp2 equals “N” - “WDensity_Sp1” and “WDensity_Sp2” indicate the post-disturbance\
weighted densities per hectare of the pre-disturbance dominant \
and co-dominant tree species, respectively.
References
McCune, B., & Keon, D. (2002). Equations for potential annual
direct incident radiation and heat load. Journal of Vegetation Science,
13, 603–606.
Vickers, L. A., McWilliams, W. H., Knapp, B. O., D’Amato,
A. W., Dey, D. C., Dickinson, Y. L., Kabrick, J. M., Kenefic, L. S.,
Kern, C. C., Larsen, D. R., Royo, A. A., Saunders, M. R., Shifley, S.
R., & Westfall, J. A. (2019). Are Current Seedling Demographics Poised
to Regenerate Northern US Forests? Journal of Forestry, 117(6),
592–612. https://doi.org/10.1093/jofore/fvz046