Higher temperatures and low precipitation strongly decrease conifer recruitment in the Rocky Mountains even in heat-adapted species and populations
Data files
Jun 22, 2026 version files 1.85 MB
-
Full_Data.csv
1.82 MB
-
Metadata.xlsx
12.37 KB
-
R_code.md
10.66 KB
-
README.md
1.58 KB
Abstract
Climate change is triggering recruitment failures in Rocky Mountain forests, yet the specific factors limiting regeneration remain unclear. Studies of established seedlings suggest the very earliest stages of recruitment are a key bottleneck in regeneration resilience, which we hypothesize is caused by the acute sensitivity of conifer germinants to climate.
In this study, we asked how climate impacts rates of seed germination, initial (first growing-season) survivorship, and seedling establishment (2- or 3-year survivorship after the first growing-season) in Colorado Front Range forests, USA. We sowed in situ seeds of common tree species – Picea engelmannii, Pinus contorta, Pinus flexilis, Pinus ponderosa, and Pseudotsuga menziesii – over a large climate gradient. Using seeds sourced across the wider region, 3-5 populations of each species were sown to explore how local adaptation in recruitment varies with climate. We expected strong inter- and intra-specific differences, predicting seeds from warmer and drier distributions experience higher recruitment.
The study years encompassed a 5°C range in growing-season temperatures and a 175% range in growing-season precipitation. Species-specific rates of mean germination (r = 0.935), initial survivorship (r = 0.786), and seedling establishment (r = 0.709) were strongly correlated with the mean seed mass of the species. For all species, temperature × precipitation interactions were important, while higher temperatures alone increased germination but markedly decreased initial survivorship. Initial survivorship overall was significantly higher in wetter conditions; for P. flexilis, P. ponderosa and P. menziesii, it remained robust even in the hottest temperatures as long as precipitation was high. Intraspecific differences were evident, as seedlings of all species had optimal initial survivorship in precipitation ~20-60% higher than origin, and negligible or steeply declining initial survivorship in conditions drier than origin. Seedling establishment rates overall were less than half initial survivorship, as strong climatic filtering continued for older seedlings, and species distribution models indicate climate change will cause recruitment failures across a substantial portion of the species’ current ranges, particularly in lower elevations, even under a low emissions scenario.
Synthesis. Climate change is increasing the likelihood of recruitment failure in the Rocky Mountains, especially in the warmest and driest portions of forest distributions, although regeneration may prove resilient to hotter conditions in areas with sufficient precipitation. Given low recruitment in hot and dry conditions for all populations, assisted migration efforts require better-adapted populations than used in this study. Overall, these results highlight the threat that warming and especially drying pose to the regeneration resilience of forests in the Rocky Mountains.
Dataset DOI: https://datadryad.org/dataset/doi:10.5061/dryad.rbnzs7htk
Rocky Mountains tree species germination and seedling recruitment data
This repository contains data and code associated with an experimental study of conifer seed germination, initial first growing-season survivorship, and seedling establishment in four experimental common gardens in the Colorado Front Range, USA.
The dataset is organized at the individual seed and seedling level. Each row represents one sown seed and includes site, quadrat, cohort year, species, population, seed-source, climate, light, germination, initial first growing-season, and longer-term seedling establishment information. Climate variables include site-year climate, seed-source climate, and climate-transfer variables.
Description of the Data and file structure
Data is structured in a 'wide' format such that data variables for each subject is in a separate column. Some analyses may require reshaping or filtering the data; for example, survivorship analyses use only seeds that germinated, and seedling establishment analyses use only seedlings that survived through the first growing season.
See the Metadata.xlsx file for the relationship between data files, data codes, and other abbreviations used. The Full_Data.csv contains the raw data for the entire study, and the R_code.md contains the R code used to model the data and generate the figures.
