The tolerance of tree growth to drought in diverse mixed forests subjected to seasonal water shortage is understudied despite their ecological and economic relevance. By combining intra- and inter-annual analyses of radial growth responses to climate and drought at daily and monthly scales, different strategies to acclimate to hydroclimate variability of coexisting conifers and broadleaves were unveiled. We analyzed the growth patterns and responses to hydroclimate variability in two conifers (Pinus engelmannii, Juniperus deppeana) and two broadleaves (Quercus grisea, Arbutus arizonica) co-occurring in a Madrean pine-oak forest located in NW México. The strongest positive response to daily precipitation was found in the two conifers, but this response peaked earlier in J. deppeana than in P. engelmannii, which presented a more delayed formation of radially-enlarging tracheids. The latest negative response to temperature was found in Q. grisea, which agrees with its more delayed xylogenesis than A. arizonica. P. engelmannii presented the highest responsiveness to water shortage, driven by lower precipitation and high maximum temperatures, responding to longer droughts ending in autumn (r ​= ​0.72), whilst A. arizonica showed the lowest responsiveness to short spring droughts (r ​= ​0.39). Growth of P. engelmannii was linked to climate-atmospheric circulation patterns over the near Pacific Ocean. Overall, P. engelmannii and A. arizonica showed high growth rates and earlier growth onset, whilst J. deppeana and Q. grisea showed slower growth rates and later growth onset. The Vaganov-Shashkin growth model evidenced that winter-spring soil moisture was the key driver of growth. Under more arid conditions and more frequent and hotter droughts, pine stands could rapidly shift towards mixed pine-oak forests.

Hydroclimate affects the radial growth responses of trees, but the drivers of their spatial and population variability are not sufficiently understood. We addressed this issue by sampling several conifer populations located at the same latitude, but at different longitude and elevation in western (W) and eastern (E) Mexican regions. We used dendroecology to disentangle how earlywood width (EW), latewood width (LW) and adjusted LW (LWadj), i.e. the residuals after removing EW influences on LW, responded to climate variables (temperature and precipitation), climate indices (Southern Oscillation Index, SOI, Niño 3.4, Pacific Decadal Oscillation, PDO) and a drought index (Standardised Precipitation-Evapotranspiration Index, SPEI). The W species (Pinus herrerae Martínez, Pinus durangensis Martínez, Abies durangensis Martínez and Cupressus lusitanica Mill.) showed lower growth rates than the E species (Pinus hartwegii Lindl., Picea mexicana Martínez, Pseudotsuga menziesii (Mirb.) Franco and Abies vejari Martínez). Growth in W benefits mostly from high precipitation in the prior winter and current spring and it is limited by high temperatures in spring, whereas growth in the E showed similar but weaker responses. Furthermore, positive (negative) correlations were found in radial growth with the Niño 3.4 (SOI) and the PDO from the prior to current autumns, which were again stronger in absolute terms in the W than in the E regions, excepting SOI in summer. In the W, P. durangensis and C. lusitanica were the least and most responsive species to spring drought, respectively; whilst P. menziesii and A. vejari were very responsive to spring drought compared to P. hartwegii in the E. Our results suggest greater responsiveness to hydroclimate and atmospheric patterns in the W than in the E region. These findings allow better interpretations of future changes in growth and composition in Mexican conifer forests, considering that climate models forecast warmer spring conditions and increased water shortage.