Climate is stronger than you think: Exploring functional planting and TRIAD zoning for increased forest resilience to extreme disturbances

Abstract

In the face of global changes, forest management must now consider adapting forests to novel and uncertain conditions alongside objectives of conservation and production. In this perspective, we modified the TRIAD zoning approach to add a resilience component through functionally diverse plantations following harvesting in the extensive areas. We then assessed the capacity of this new "TRIAD+" zoning approach for improving the resilience of the mature forest biomass to climate change and three potential extreme pulse disturbances: a large fire, a severe drought, and an insect outbreak. We used the forest landscape simulation model LANDIS-II on a management unit in Mauricie (Quebec, Canada) to simulate and compare the TRIAD+ scenario with a classic TRIAD zoning scenario, and two business-as-usual harvesting scenarios with and without functional enrichment planting. We also simulated three different climate change scenarios (Baseline, RCP 4.5 and RCP 8.5) in which these management and extreme disturbance scenarios took place. We monitored the changes in three variables: the mature wood biomass across the landscape, the mature biomass of each functional group, and the functional diversity of stands in the landscape. Resilience was measured according to three indicators: resistance, net change and recovery time of mature biomass. TRIAD+ management resulted in a good compromise, harvesting the same amount of wood as other scenarios while increasing the surface of protected forests by around 240% compared to BAU scenarios, and improving the mean functional diversity of stands by around 15% compared to the classic TRIAD and BAU without plantations. Following the pulse disturbance events, TRIAD+ also increased the resilience of the mature biomass across the landscape. However, this increase was limited, depended on the resilience indicator and the event considered, and was negligible in terms of tree biomass recovered in the long term. It's uncertain whether these results stemmed from the relative lack of small-scale interactions in LANDIS-II through which the effect of functional diversity on stand resilience should occur, or if this effect is small to begin with. Overall, our study reveals that an adaptation component can be included in current or future management strategies, but that increasing functional diversity via plantations will likely be insufficient to significantly boost forest resilience. Future research should therefore explore other (combined) means of increasing forest resilience, and improve the representation of small-scale interactions in landscape-scale models.

Fig 1. Map showing the location and extent of our study area located in the Mauricie region of Quebec, Canada.

Satellite data from the Sentinel satellite, and edited by the Ministry of Forests and Natural Resources of Quebec under a CC-BY 4.0 license (https://www.donneesquebec.ca/recherche/dataset/mosaique-satellites).

Fig 2. Visual representation of the three categories of scenario simulated with LANDIS-II: forest management, climate, and catastrophic event.

Fig 3. Measures of resilience used in our study.

A, B and E refers to the value of the variable of interest before the disturbance (A), right after it (B), and at the end of the simulation (E), and are used in equations 2 and 3.

Fig 4. Temporal variation of the Total Mature Biomass in the landscape B_L for each combination of management, climate and catastrophe scenario.

Solid lines are mean values and envelops are standard deviation across 5 simulation replicates.

Fig 5. Temporal variation of the mean Functional Diversity across all stands of the landscape (F̅D̅_S) for each combination of management, climate and catastrophe scenario.

Solid lines are mean values and envelops are standard deviation across stands and 5 simulation replicates.

Fig 6. Evolution of the total biomass for the six functional groups of trees defined in our study for each combination of climate and management scenario, but without a catastrophic disturbance event at t = 100.

Values at each time step are mean values across 5 simulation replicates.

Fig 7. Bar plots showing the resilience values of the total biomass of the landscape (BL) for each resilience measure (row) and each catastrophic event (column).

The height of the bar are mean values and error lines are standard deviation across 5 simulation replicates.