Tree growth and survival over 61 years at the Second International Larch Provenance Test in southeastern Michigan, USA

Abstract Provenance trials remain an important tool for understanding genetic variation in forest trees. Maintaining decades-old experiments is critical for revealing maladaptation of translocated populations to local sites and meaningful divergence in survival and growth. We revisited the Second International Larch Provenance Test in southeastern Michigan, which featuring 24 provenances of European and Japanese larch. We compiled data collected when the plantation was 25, 41, and 61 years from seed to supplement data published at 19 years from seed. Divergence in survival rates first became pronounced at 41 years from seed. Plantation sources of unknown origin and continental Sudeten provenances had the highest rates of survival. High-elevation sources from the southern Alps performed the poorest. Survival differences were likely driven by ice storm damage and severe growing season droughts, both of which occurred at least 1-2 times per decade in southeastern Michigan over the last 60 years. Provenance performance for growth mirrored that of survival, although sources varied more in their survival than in their growth. Growth rate among sources varied over the 43-year sampling period, with intervals of fast growth occurring in several of the overall poor-performing provenances. Diameter was moderately correlated to both elevation and latitude, with inconsistent correlations to climatic variables. Broad geographical variation in provenance performance was also important in explaining variation in growth. Our data corroborate that monitoring of long-term provenance trials could be an important source of information needed to predict and understand how forest species may respond to imminent climate change and may be critical for developing strategies for its mitigation.