Abstract/Summary

Widespread forest dieback is a phenomenon of global concern that requires an improved understanding of the relationship between tree growth and climate to support conservation efforts. One priority for conservation is the Atlas cedar (Cedrus atlantica), an endangered species exhibiting dieback throughout its North African range. In this study, we evaluate the long-term context for recent dieback and develop a projection of future C. atlantica growth by exploring the periodic variability of its growth through time. First, we present a new C. atlantica tree- ring chronology (1150–2013 CE) from the Middle Atlas mountains, Morocco. We then compare the new chronology to existing C. atlantica chronologies in Morocco and use principal components analysis (PCA) to isolate the common periodic signal from the seven longest available records (PCA7, 1271–1984 CE) in the Middle and High Atlas portions of the C. atlantica range. PCA7 captures 55.7% of the variance and contains significant multidecadal ( ̃95yr, ̃57yr, ̃21yr) periodic components, revealed through spectral and wavelet analyses. Parallel analyses of historical climate data (1901–2016 CE) suggests that the multidecadal growth signal ori- ginates primarily in growing season (spring and summer) precipitation variability, compounded by slow- changing components of summer and winter temperatures. Finally, we model the long-term growth patterns between 1271–1984 CE using a small number (three to four) of harmonic components, illustrating that sup- pressed growth since the 1970s – a factor implicated in the dieback of this species – is consistent with recurrent climatically-driven growth declines. Forward projection of this model suggests two climatically-favourable periods for growth in the 21st century that may enhance current conservation actions for the long-term survival of the C. atlantica in the Middle and High Atlas mountains.