TY - JOUR
T1 - Interactions between landscape and local factors inform spatial action planning in post-fire forest environments
AU - Peeler, Jamie L.
AU - Smithwick, Erica A.H.
N1 - Funding Information:
The National Science Foundation (BCS-1901630), Joint Fire Science Program (1910144), National Geographic, National Aeronautics and Space Administration Pennsylvania Space Grant Consortium, and Penn State Center for Landscape Dynamics funded the study.
Funding Information:
The National Science Foundation (BCS-1901630), Joint Fire Science Program (1910144), National Geographic, National Aeronautics and Space Administration Pennsylvania Space Grant Consortium, and Penn State Center for Landscape Dynamics funded the study. We thank Hunter Mitchell and Shaina Walker for their hard work in the field, as well as Shelby Duncan for their assistance with the object-based image analysis. We also thank Diane Abendroth and Bridger-Teton National Forest for their logistical support and permitting the study. Laura Leites, Doug Miller, and Alan Taylor served on JLP’s dissertation committee and provided helpful feedback on the manuscript and study.
Publisher Copyright:
© 2021, The Author(s), under exclusive licence to Springer Nature B.V.
PY - 2021/12
Y1 - 2021/12
N2 - Context: Landscape and local factors govern tree regeneration across heterogeneous post-fire forest environments. But their relative influence is unclear—limiting the degree that managers can consider landscape context when delegating resources to help stand-replacing patches restock successfully. Objectives: We investigated how landscape and local factors shape tree regeneration across heterogeneous post-fire forest environments. Our research questions were: What is the relative influence of landscape and local factors on tree species presence (RQ1) and stocking density (RQ2) at stand-replacing patches? Do thresholds occur when landscape factors are influential (RQ3)? Methods: We sampled landscape and local variables at 71 plots near Jackson, Wyoming, United States. We used Random Forests to investigate how local and landscape variables affect post-fire tree recovery. Relative influence was determined using mean decrease in accuracy. Partial dependence plots were used to visualize whether thresholds occurred for variables with mean decrease in accuracy > 15%. Results: Landscape factors like seed source area were associated with subalpine fir presence and stocking density. But different thresholds occurred. Specifically, subalpine fir presence required 10% seed source area, while stocking density required 40%. Northeast aspects surrounded by > 10% seed source area were most likely to support subalpine fir presence. Conversely, local factors like soil nutrients were associated with lodgepole pine presence, highlighting effects of different regeneration strategies. Conclusions: Landscape factors bolster spatial resilience and help stand-replacing patches restock naturally. But landscape factors do not support tree regeneration equally across heterogeneous post-fire forest environments. Consequently, considering stand-replacing patches in their landscape context will be critical for future spatial action planning.
AB - Context: Landscape and local factors govern tree regeneration across heterogeneous post-fire forest environments. But their relative influence is unclear—limiting the degree that managers can consider landscape context when delegating resources to help stand-replacing patches restock successfully. Objectives: We investigated how landscape and local factors shape tree regeneration across heterogeneous post-fire forest environments. Our research questions were: What is the relative influence of landscape and local factors on tree species presence (RQ1) and stocking density (RQ2) at stand-replacing patches? Do thresholds occur when landscape factors are influential (RQ3)? Methods: We sampled landscape and local variables at 71 plots near Jackson, Wyoming, United States. We used Random Forests to investigate how local and landscape variables affect post-fire tree recovery. Relative influence was determined using mean decrease in accuracy. Partial dependence plots were used to visualize whether thresholds occurred for variables with mean decrease in accuracy > 15%. Results: Landscape factors like seed source area were associated with subalpine fir presence and stocking density. But different thresholds occurred. Specifically, subalpine fir presence required 10% seed source area, while stocking density required 40%. Northeast aspects surrounded by > 10% seed source area were most likely to support subalpine fir presence. Conversely, local factors like soil nutrients were associated with lodgepole pine presence, highlighting effects of different regeneration strategies. Conclusions: Landscape factors bolster spatial resilience and help stand-replacing patches restock naturally. But landscape factors do not support tree regeneration equally across heterogeneous post-fire forest environments. Consequently, considering stand-replacing patches in their landscape context will be critical for future spatial action planning.
UR - http://www.scopus.com/inward/record.url?scp=85113268428&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85113268428&partnerID=8YFLogxK
U2 - 10.1007/s10980-021-01325-4
DO - 10.1007/s10980-021-01325-4
M3 - Article
AN - SCOPUS:85113268428
SN - 0921-2973
VL - 36
SP - 3523
EP - 3537
JO - Landscape Ecology
JF - Landscape Ecology
IS - 12
ER -