Robustness of high-throughput prediction of leaf ecophysiological traits using near infra-red spectroscopy and poro-fluorometry
Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background
Water resource is a major limiting factor impacted by climate change that threatens crop production and quality. Understanding the ecophysiological mechanisms involved in the response to water deficit is crucial to select new varieties more drought tolerant. A major bottleneck hampering such advances is the lack of methods for measuring ecophysiological traits on large populations of individuals. We investigated the relevance of spectroscopy and poro-fluorometry with related high-throughput measurement devices to predict leaf morphological and ecophysiological traits using partial least square regression (PLSR) models. This work relies on a grapevine diversity panel grown in pots under contrasted conditions, outdoors under well watered conditions and in a greenhouse with three different soil water treatments. We took advantage of these experimental designs to specifically assess the robustness of predictive models.
Results
Some complementarity between measuring devices were found, with spectrometers being able to predict leaf mass per area, water content and water quantity (R²>0.58), while the poro-fluorometer could predict net CO 2 assimilation (R²>0.72), regardless of the water treatment. The prediction of leaf mass per area appeared to be quite robust between outdoors and greenhouse experiments. The prediction of water use efficiency was highly dependent on the water treatment, with much better predictions under moderate (R²>0.73) than severe water deficit. We then used the high-throughput devices alone to measure the whole grapevine panel, by applying calibration models to predict ecophysiological traits and estimated their broad sense heritability. Leaf mass per area was also directly determined on the entire diversity panel and its heritability was similar whether calculated on observed or predicted values. The highest computed heritabilities for these traits reached values close to 0.5.
Conclusions
This study showed that spectrometers and poro-fluorometer could be reliable nondestructive tools for the high-throughput phenotyping of ecophysiological traits on thousands of plants, paving the way for studying the genetic determinism of such traits.
