Assessing Seed Production and Genetic Resources Adaptability
| by Delphine Grivet (INIA)
The main objective of WP3 is to provide dynamic descriptors of Genetic Conservation Units (GCUs) integrating genotypic, phenotypic and environmental characteristics, with the ultimate goal of evaluating and forecasting population adaptive capacity. One of the most immediate goals during the first year of the project was to measure proxies of adaptive traits in two GCUs with contrasting ecological characteristics for two target species (black poplar, Populus nigra and maritime pine, Pinus pinaster). The heritability or these traits and their relationship with fitness will be determined in the next steps. Among these traits, microdensity was selected as it provides estimates of wood density while near-infrared spectroscopy (NIRS) provides information on wood properties.
The use of high-throughput phenotyping, reduced the need of an intensive sampling (500 trees per GCU) and the labour and cost of phenotyping. We adopted a strategy that combines: (i) phenotyping a minimized number of trees (10 per GCU) to analyse in parallel long cores (from bark to pith) by X-ray microdensitometry, and short cores or wood drilldust (external 5 cm) by NIRS. This will allow us to develop models that predict wood properties from NIRS data. (ii) Phenotyping a large number of trees (500 per GCU) to analyse short cores or wood drilldust (external 5 cm) to access wood properties via NIRS.
This work was carried out thanks to coordination between WP3 local teams (INIA-CSIC & CITA in Spain, BFW in Austria, CNR in Italy and INRAE Pierroton in France) and the WP2 mobile team (INRAE Avignon in France and CREAF in Spain). Samples are being analysed in the partners’ laboratories (CETEMAS in Spain and INRAE Orléans in France). Workflow for XR analysis includes the following steps (following CETEMAS protocol):
Phase 1: in the field. This phase includes sampling using an auger to get a core from bark to pith. It is a delicate operation as the cores may be fragile. As soon as they are extracted, they are preserved in plastic straws to keep them moist until sawing and to avoid bending.
Phase 2: in the laboratory In this phase, the samples are prepared to be sawn to obtain a 1.5mm width slice for X-ray analysis. They are first glued and then sawed, trying not to break or bend them. It is very important that the cores stay moist during this phase, otherwise, the cores will bend or break easily and it will not possible to obtain suitable samples.
Phase 3: in the lab. Once the X-ray samples are ready, they are first extracted with ethanol for 24h to remove resin. This is necessary to avoid errors in density determination. Samples are dried after extraction and before density determination, handling them with care while they are soaking so there are no breakages. Dried samples are then analysed using an Itrax multiscanner, a device that combines optical measurement (RGB image) and density measurements using soft X-rays of high intensity for microdensitometry. This equipment was developed exclusively for the non-destructive analysis of some types of samples, including wood. It allows us to obtain high-precision data on growth and density at intra- and inter-annual scales from the growth rings. We get the results by analysing the optical and radiographic images with specific software that, among other things, compares the radiographic images with standards of known density.
Phase 4: in the computer. Using Windendro software we can define a path and determine the number of tree rings, average density or ring density and obtain different parameters, such as early density or late density for example. However, the tree species and the growth rate can impact the resolution and affect the analysis.
This workflow will allow us to estimate wood density values and to use them as references to developing NIRS calibration models, matching them with NIR spectra collected from the drilldust extracted from the same trees. We will work on the models using spectra vs. reference value in 10 trees/12 GCUs. These models will then be validated for samples corresponding to 50 tree cores from 2 GCUs, and used to predict wood density in samples from 500 trees in 2 GCUs.
Relevant terminology
- Near-Infrared spectroscopy (NIRS) : a spectroscopic method that uses the near-infrared region of the electromagnetic spectrum (from 780 nm to 2500 nm).
- Wood microdensity : the proportion of cell wall in a given woody tissue.
- Adaptive trait : an aspect of the developmental pattern which facilitates the survival and/or reproduction of its carrier in a certain succession of environments.