Modelling seedling establishment and survival in Pinus pinea forests in Spain
In his PhD thesis entitled “An integral model of natural regeneration for Pinus pinea L. in the Northern Plateau (Spain)”, Rubén Manso uses multistage models. Such models consist of several sub-models corresponding to the transitional probabilities between different regeneration phases (dispersal, germination, predation, etc.). However, there was no good sub-model for predicting the survival probability of established seedlings. He recounts how EFIMED supported him to address this gap.
On being awarded an EFIMED Short Scientific Visit (SSV) grant, I had the opportunity to collaborate with an expert in survival analysis to address the missing sub-model for seedling establishment and survival in Pinus pinea forests. Thus, in late 2012 I worked on developing such a model at the Institute National de la Recherche Agronomique (INRA), based in Nancy (France), under the supervision of Dr. Mathieu Fortin. The title of the SSV was “Modelling seedling establishment and survival in Pinus pinea forests in Spain. Management implications in a context of global change”. From the model, we expected to understand the impact of seedling mortality in natural regeneration of Pinus pinea, as well as the climate- and management-related causes behind it.
The construction of the model was challenging, as several assumptions of basic models based on survival analysis were violated by the data. A non-trivial likelihood function was developed to counteract these shortcomings and a programming effort was undertaken with initial computations leading to promising preliminary results. The outputs of our model suggest that autumn frost is a strongly limiting factor for the survival of seedlings, accounting for the observed temporal variability of the process. At spatial level, seedlings located far from parent trees have less probability of survival than those occurring beneath the tree’s crown. Overall, seedling mortality was extraordinarily high for all autumn cohorts, particularly during the first months. Seedling population within a cohort seems to stabilise after two years, with a survival of 8% for individuals. Therefore, the impact of mortality on natural regeneration of the species is highly relevant, constituting a bottleneck in the regeneration process. We believe that felling schedules should be more flexible, permitting a high stand density until regeneration is achieved.
However, the most interesting results are yet to come. Data informing about mortality rate due to summer droughts, the main effect related to global change in the Mediterranean, is to be included in the model. This will provide essential information from which we will be able to infer future regeneration performance for this species, which has a significant economic and ecological importance.
I encourage everybody interested in improving and enriching their research to apply for EFIMED’s SSV scholarships. Collaborating with people from other laboratories, especially abroad, opens your mind, allows you to learn new techniques and gives you the opportunity of further collaboration in the future, as it is my case.
All photos, Rubén Manso
More information: Rubén Manso: rmgforestal (at) hotmail.com
Manso, R., Fortin, M., Calama, R., Pardos, M. 2013. Modelling seed germination in forest tree species through survival analysis. The Pinus pinea L. case study Forest Ecology and Management 289: 515–524.
Back to article listing