TY - JOUR
T1 - A comparison of methods to estimate seasonal phenological development from BBCH scale recording
AU - Cornelius, Christine
AU - Petermeier, Hannes
AU - Estrella, Nicole
AU - Menzel, Annette
PY - 2011/11
Y1 - 2011/11
N2 - The BBCH scale is a two-digit key of growth stages in plants that is based on standardised definitions of plant development stages. The extended BBCH scale, used in this paper, enables the coding of the entire development cycle of all mono- and dicotyledonous plants. Using this key, the frequency distribution of phenological stages was recorded which required a less intense sampling frequency. The onset dates of single events were later estimated from the frequency distribution of BBCH codes. The purpose of this study was to present four different methods from which those onset dates can be estimated. Furthermore, the effects of (1) a less detailed observation key and (2) changes in the sampling frequency on estimates of onset dates were assessed. For all analyses, phenological data from the entire development cycle of four grass species were used. Estimates of onset dates determined by Weighted Plant Development (WPD), Pooled pre-/post-Stage Development (PSD), Cumulative Stage Development (CSD) and Ordinal Logistic Regression (OLR) methods can all be used to determine the phenological progression of plants. Moreover, results show that a less detailed observation key still resulted in similar onset dates, unless more than two consecutive stages were omitted. Further results reveal that the simulation of a less intense sampling frequency had only small impacts on estimates of onset dates. Thus, especially in remote areas where an observation interval of a week is not feasible, estimates derived from the frequency distribution of BBCH codes appear to be appropriate.
AB - The BBCH scale is a two-digit key of growth stages in plants that is based on standardised definitions of plant development stages. The extended BBCH scale, used in this paper, enables the coding of the entire development cycle of all mono- and dicotyledonous plants. Using this key, the frequency distribution of phenological stages was recorded which required a less intense sampling frequency. The onset dates of single events were later estimated from the frequency distribution of BBCH codes. The purpose of this study was to present four different methods from which those onset dates can be estimated. Furthermore, the effects of (1) a less detailed observation key and (2) changes in the sampling frequency on estimates of onset dates were assessed. For all analyses, phenological data from the entire development cycle of four grass species were used. Estimates of onset dates determined by Weighted Plant Development (WPD), Pooled pre-/post-Stage Development (PSD), Cumulative Stage Development (CSD) and Ordinal Logistic Regression (OLR) methods can all be used to determine the phenological progression of plants. Moreover, results show that a less detailed observation key still resulted in similar onset dates, unless more than two consecutive stages were omitted. Further results reveal that the simulation of a less intense sampling frequency had only small impacts on estimates of onset dates. Thus, especially in remote areas where an observation interval of a week is not feasible, estimates derived from the frequency distribution of BBCH codes appear to be appropriate.
KW - Flowering
KW - Grassland
KW - Observation key
KW - Onset dates
KW - Sampling frequency
UR - http://www.scopus.com/inward/record.url?scp=80955137609&partnerID=8YFLogxK
U2 - 10.1007/s00484-011-0421-x
DO - 10.1007/s00484-011-0421-x
M3 - Article
AN - SCOPUS:80955137609
SN - 0020-7128
VL - 55
SP - 867
EP - 877
JO - International Journal of Biometeorology
JF - International Journal of Biometeorology
IS - 6
ER -