Higher sample sizes and observer inter‐calibration are needed for reliable scoring of leaf phenology in trees

Abstract

Reliable phenological observations are needed to quantify the impact of climate change on tree phenology. Ground observations remain a prime source of phenological data, but their accuracy and precision have not been systematically quantified. The high subjectivity of ground phenological observations affects their accuracy, and the high within‐population variability of tree phenology affects their precision. The magnitude of those effects is unknown to date. We first explored the inter‐observer variability in the timing of bud development and leaf senescence in trees using a unique dataset of seven observer inter‐calibration sessions. Then, using tree phenological data collected in three European forests (n = 2,346 observations for budburst, n = 539 for leaf senescence), we quantified how the ‘observer uncertainty’ (accuracy of the observations) and the ‘population sampling uncertainty’ (precision of the observations) combine to affect the estimates of the budburst and the leaf senescence dates. The median observer uncertainty was 8 days for budburst (BBCH = 7) and 15 days for leaf senescence (BBCH = 95). As expected, the population sampling uncertainty decreased with increasing sample size, and was about 6 days for budburst and 10 days for leaf senescence for a sample of 10 individuals monitored per population (corresponding to the median sample size in the phenological literature). As a whole, the overall uncertainty of phenological observations could reach up to 2 weeks for budburst and 1 month for leaf senescence. Synthesis. This paper quantifies for the first time the accuracy and precision of ground phenological observations in forest trees and as such offers tables to estimate the uncertainty of phenological data. We show that reliable estimates of budburst and leaf senescence require three times (n = 30) to two times (n = 20) larger sample sizes as compared to sample sizes usually considered in phenological studies. We further call for an increased effort of observer inter‐calibration, required to increase the accuracy of phenological observations. These recommendations reduce the uncertainty of phenological data, thereby improving the estimation of phenological trends over time, the response of phenology to temperature or the inference of phenological model parameters.

Spring and Autumn Phenology in Sessile Oak (Quercus petraea) Near the Eastern Limit of Its Distribution Range

“…For example, Tomescu [67] reported a DOY of 289 for sessile oak in the Dobrina forest, and Tomescu [84] reported a DOY of 273 for sessile oak in the Poieni forest located 360 m from our RAD population, suggesting a delay in leaf senescence of 1.7-4.6 days per decade. As mentioned previously, these differences should be treated with caution, since the phenological observation protocol is not the same, and the influence of inter-observer variability is very high [85]. Similarly, we observed a small delay in the average date of leaf senescence in the reference HEL population (DOY 282) and the FAN comparative trial (DOY 293) compared with nearby sites such as the Bradatel forest (DOY 273) and Casa cu Nuci forest (DOY 289; [84]).…”

“…These changes in the occurrence of phenological phases are also reflected in the length of the vegetation season [87,[96][97][98], as can be seen in comparisons between our results and those reported previously for bioactive season lengths. Based on observations between 1956-1965, Tomescu [85] found that the vegetation period in sessile oak lasted, on average, 180 days regardless of geographical gradients (200 days at 100 m altitude and 160 at 650 m). Therefore, we can observe a lengthening of the growing season by ca.…”

Process-oriented models of autumn leaf phenology: ways to sound calibration and implications of uncertain projections

“…the timing of spring phenology with respect to autumn phenology; Keenan and Richardson, 2015). Models of spring phenology regularly outcompete models of autumn phenology by several days when assessed by the root mean square error between observed and modelled dates (4-9 vs. 6-13 days, respectively; Basler, 2016;Liu et al, 2020). 60…”

“…While the uncertainty associated with climate projections has been extensively researched (e.g. Palmer et al, 2005;Foley, 2010;Braconnot et al, 2012), so far the uncertainty associated with process-oriented phenology models has only been described in 75 a few notable studies: Basler (2016) compared spring phenology models calibrated per species and per site as well as calibrated per species with pooled sites, Liu et al (2020) compared autumn phenology models with a focus on inter-site and inter-annual variability, and Liu et al (2021) focused on sample size and observer bias in observations of spring and autumn phenology.…”

Phenological Stages of Development of Cornus L. S. Str. Species (Cornaceae) According to BBCH Scale

“…The expanded BBCH scale is growing in popularity. Now it has become international and has already been adapted for many different in origin, ecology, and economic use of plants of common and little-known species and agricultural crops, including fruit crops (Danner et al, 2019;Vaidya, 2019;Liu et al, 2021).…”