Effets directs et différés du climat sur l'architecture et la phénologie des pins

[Departement_IRSTEA]Territoires [TR1_IRSTEA]SEDYVINInternational audienceGoals In this study, we tried to quantify the influence of climate and climatic accidents (scorching heat, drought, snow) on the primary production and architectural development of five pine species growing in Mediterranean regions: Pinus halepensis, P. pinea, P. pinaster, P. nigra and P. silvestris. Measured parameters include branch length growth, number of annual flushes for polycyclic species, branch ramification rate, size and number of needles per year, male and female flowers production and successful fructification. The study also aimed at assessing the direct and delayed effects of climatic accidents on the leaf area, to be related to observed crown transparency. Material and methods The study area includes 8 sites distributed between 80 and 1300 m of elevation. 2100 branches from 150 trees were sampled from 2004 to 2010 and architectural development was retrospectively measured from morphological markers over a period of 15 to 25 years. The sampling design considered separately for each tree various parts of the crown (top, middle and base), orientation (north and south), and also branch hierarchy (primary or secondary axis) and vigor. A weekly follow-up of phenology and architectural development was performed from 2008 to 2011 on one of the sites to understand their relations with climate. The observed variations of the architectural parameters were integrated into a specifically designed model of pine branch architectural development and in a model of full tree development to simulate the impact of climate on pine leaf area. Results Models of architectural parameters vs climate relationships were calibrated with the monthly or seasonal data of rain and temperature and their extremes. For a given year, most of the architectural parameters were influenced by the climate of the previous year, and generally less by the current year. Their medium-term evolution was clearly influenced by recent climate changes, particularly repeated droughts and increasing temperatures. For branch length growth, we were able to disentangle the respective weight of climatic accidents and of the normal evolution with age. With the exception of Pinus silvestris at high elevation in the hinterland, 2003 scorching heat and repeated droughts from 2003 to 2007 considerably reduced the primary production for all species and on all sites: weak branch length growth, abnormally short needles and low number of needle per shoot, low polycyclism, ramification and fruiting rates. This reduction began in 1998 for most of these parameters in the low and sometimes middle parts of the crowns, and on weak branches. For the most Mediterranean species (Pinus halepensis and P. pinea), these parameters recovered gradually from 2008 to 2010 in the top crown, but they remained low in the middle and base of the crown and, for the other species, in the whole crown. The weakness of trees in the 2003-07 period was highlighted by the impossibility for twigs to develop normally when they carried cones. At low or intermediate elevation, climate warming resulted in a time shift in phenological phases and disturbance in the architectural development: male flowers appearing in the middle of winter, continuous shoot growth over 2 successive years. This led to high rates of twig abortion caused by frost and fungus. Simulations with architectural development models allowed calculating the resilience of leaf area according to the intensity and duration of climate accidents. Following the 2003-2007 drought, the deficit of active twigs (carrying needles) and of leaf area amounted respectively to 60 % and 75 % for Pinus halepensis compared to the values simulated with an average climate. Because of the 2 to 3-year persistence of needles and of the time necessary for restoring a normal ramification rate, this deficit was only slowly absorbed: it remained close to 30% and 40 % respectively two years after the end of the drought and may need four years to disappear if no new accident occurs. With a warmer and drier climate, phenological disorders may lead to a regular mismatch between the life cycle and climate conditions, increasing tree vulnerability to climate accidents.The long lasting reduction of potential tree LAI after a climatic accident or repeated stress, due to a simplified architecture and small needle size, may contribute to carbon shortage and starvation, and therefore to delayed forest die-back