Four plants, two hemispheres, same baseline?

Changes in plant phenology due to anthropogenic climate change are now well documented – although, the phenological record base for the northern hemisphere is much richer. Currently, the overall picture between the two hemispheres is similar (e.g. earlier occurrence of spring events), however, the rate of change in many cases appears to be faster in the southern hemisphere. So is this in part attributable to climate or species sensitivity or a combination of both? A step to this is to examine the early phenology of species prior the significant influence of climate.

Hence we have started to find and compare early/legacy records (pre-1900) from the southern and northern hemisphere. This, naturally, necessitates comparison of only native species from one hemisphere and/or agricultural species. As indicated southern hemisphere records are sparse. However, pre-1900 datasets are available for a few locations, of these the records from the Tasmanian Royal Society (Hobart, Australia) covering the period 1864 to 1886 contain primarily northern hemisphere species. The PEP725 (http://www.pep725.eu/) database was examined for matching species, phases and coinciding time periods. Four species (Aesculus hippocastanum – first flowering; Fraxinus excelsior – leafing, Robinia pseudoacacia – leafing and Sambucus niger – leafing) from seven locations overlapped. Each of these is from the Dutch Royal Meteorological Organisation (KNMI) records (1868 to 1898). In all there were seven locations with one location, Oostkapelle, shared between species. We have assumed that the descriptive terms used in Hobart (e.g. commencing to flower is equivalent to 60 in the BBCH code used in PEP725 data).

As a first step we compared the seasons the phases occurred in as well as correlations between and among species, and locations. For A. hippocastanum and F. excelsior the respective phases occurred in the same season but in early Spring in Hobart and mid to late Spring in the Netherlands. For S. niger and R. pseudoacacia the phases did not occur in the same season. In Hobart leafing in these species always occurred earlier: S. niger was always coming into leaf by late winter (August) compared to late Spring and early Summer (May/June) in the Netherlands, R. pseudoacacia was in leaf by early Spring (September) in Hobart compared to Summer (June/July). These differences in timing could in part be explained by Hobart being between 2.4 and 3.5 °C warmer on average.

The strongest correlation (R = 0.60) within species was in F. excelsior between Oostkapelle and Hobart. Although this was not significant (P = 0.11) and over a very limited period; 1869 and 1876. The largest number of coinciding years, from 1868 to 1883, was in S. niger at both Slijk-Ewijk and Zaandam. These correlations were weak (R = 0.13 and -0.16, respectively) and also not significant (P > 0.2). Weak agreement also occurred within species in the Netherlands: for example in S. niger between Aardenburg and Varsseveld (R = 0.04) and A. hippocastanum between Aardenburg and Bovenkarspel (R = -0.19).

Interestingly, equally as strong or stronger correlations were found between the species. The strongest of these relationships was between F. excelsior in Hobart and R. pseudoacacia in Oostkapelle (R = 0.92, P = 0.009), however, it was over the same limited time period previously mentioned (1869 and 1876). This does, however, rise the possibly of using surrogate species for current monitoring.

This limited examination of phenological phases across the two hemispheres has not provided a clear answer on whether the current differences are attributable to climate or species sensitivity or a combination of both. The reasons for this could be differences in interpretation of phases, the limited number of coinciding years, phenological plasticity or real regional differences. Hence further exploration such as the examination of commonality of growing degree days within the longer periods of these data as well as locating legacy datasets with greater overlapping periods are needed to provide a clear answer.