A continual improvement of our understanding about factors that influence tree regeneration is needed to inform landscape-scale restoration efforts across the western USA. This is increasingly important because changing climate and disturbance regimes will influence where restoration efforts should be implemented across landscapes. In this study, we used Forest Inventory and Analysis (FIA) data to model ponderosa pine (Pinus ponderosa C. Lawson var. scopulorum Engelm.) seedling presence across FIA plots in Arizona and New Mexico, USA. Additionally, we analyzed archived soil samples from FIA plots for percent clay, silt, and sand; samples were collected from the upper 20 cm of the mineral soil. Soil particle size is correlated with soil parent material, which has been shown to influence ponderosa pine regeneration. We used structural equation modeling to determine the relative importance of direct and indirect factors affecting ponderosa pine seedling presence. For the 0–10 cm depth increment of the mineral soil, soils with high inorganic C (carbonate) concentrations, a low percentage of coarse fragments (>2 mm) in the mineral soil, a low percentage of sand in the fine fractions (<2 mm) of the mineral soil, and low average May minimum temperature had the greatest probability of ponderosa pine seedling presence. Average annual precipitation also influenced pine regeneration through its negative correlation with inorganic C concentration. A partial explanation for the latter correlation could be that carbonates in the upper soil horizons leach into lower layers over time in areas with greater precipitation. Inorganic C concentration was also associated with soil parent material and was correlated with the soil cation exchange capacity and several macronutrient concentrations. Overall, our results indicate that soils and climate are key drivers of tree regeneration in the southwestern USA. Other factors such as the forest overstory, including the presence of seed trees, and the herbaceous plant community can also influence ponderosa pine regeneration. However, these factors were not statistically significant in the current study, likely because a small subset of FIA plots were used in the analyses. As soils are collected on additional FIA plots, new models of tree regeneration based on the current structural equation model could be evaluated and refined. Our results also indicate that information about the soil series associated with FIA plots would be useful for modeling tree regeneration, especially because soil series are commonly used by land managers in decision-making processes.
Read full abstract