2.3.1.4 Discussion
We are not postulating that pathogens have necessarily created a biotic lock-in of deciduousness, but rather that the evolutionary inertia may have been decisive in some cases but not others. The evolutionary outcome could likely be influenced by various factors, including details of the pathogen life cycle. Possibly relevant details include: longevity of fungal fruiting bodies, level of host specificity, and absence or presence of alternate hosts. If alternate hosts exist, whether they are obligate, and what phases of the pathogen life occur in which host, are also likely to be relevant. A useful framework to examine these complex co-evolutionary dynamics that includes the context of abiotic factors is the ‘disease triangle’ (Stevens 1960), which illustrates the interplay between host, pathogen and environment, and their respective interaction and influence on disease and impact upon the host (Figure 3).
The story for tropical tree species does appear to be different. The production of foliage during the dry season in some species suggests strongly that, if pathogens are a selective force, they would have sometimes driven a switch to deciduousness.
The picture concerning the general incidence of deciduousness among woody perennial taxa is mixed. Conifers, which represent the ancient taxon, are very predominantly evergreen, suggesting profound evolutionary inertia against a shift to deciduousness. One exception is the entire genus Larix , within the Pinaceae, being deciduous. The remaining exceptions are all Taxodiaceae members of the Cupressaeae/Taxodiaceae complex, namely the genera Taxodium ,Glyptostrobus and Metasequoia , although T. mucronatum Ten. is semi-deciduous. They also tend to have few pathogens, along with some other taxonomically isolated species. Among angiosperms, which are evolutionarily more recent, deciduousness is far more common, in numbers of species if not in percent of total species. Even with genera, sympatric species can include both deciduous and evergreen members. This is so with Nothofagus in South America (as already mentioned), and Quercus in North America (where some species are semi-evergreen/deciduous) and the Mediterranean basin. Overall, the general incidence of deciduousness among taxa argues against any completely overriding influence of abiotic environment or taxonomic lineage per se .
2.3.2 Swiss needle cast and Douglas-fir
With native populations of Douglas-fir (Pseudotsuga menziesii(Mirb.) Franco in western north America, dates of spring bud burst show a coast-to-inland gradient from the Pacific coast. At a given elevation bud burst comes earlier the further from the coast (Campbell and Sugano 1979). This has been observed not only in situ but also in common-garden experiments, ruling out a simple effect of cooler coastal temperatures caused by the cold ocean current. One possible reason is that coastal populations have a longer humid season in which to complete vegetative growth and cone ripening, especially compared with more easterly populations where summer drought starts earlier (cf Campbell and Sorensen 1978). Another possible factor may be less insolation early in spring. Both factors might reduce the advantages of early bud burst in coastal populations. Yet another possibility, not mutually exclusive, is that later bud burst there escapes the worst of the seasonal hazard, created by humidity, of infection by foliage pathogens. Of such pathogens the most prominent is Phaeocryptopus gaeumannii(T.Rohde) Petrak, cause of Swiss needle cast (Boyce 1940; Mulvey et al. 2013), which is most aggressive in the humid coastal climates. Given the postulated selection pressure imposed by the pathogen, one might expect fog-belt Douglas-fir grown in very mild climates without the pathogen to show a genetic shift to earlier bud burst. However, any such study would both be long-term and face the difficulty of finding anywhere that will remain free of the pathogen. Chemical control of the pathogen would need to be long-term and is expensive, and could be complicated by the pathogen evolving resistance.
2.3.3 Two Pinusspecies
Two closely related coastal California pine species appear to be another case in point. They are Pinus radiat a D.Don (either Monterey pine or radiata pine) and P. muricata D.Don (bishop pine or muricata pine). They have overlapping (but almost entirely allopatric) geographic distributions. Pinus radiata ranges discontinuously from 28−37°N, and P. muricata discontinuously from 311/2−41°N. Pinus radiata has a very long, opportunistic growing season, being able to make growth all year if temperatures permit, which evidently accounts for a very fast growth potential (e.g. Burdon 2001; Burdon et al. 2017). By comparison.P. muricata makes minimal shoot elongation during winter, with apical buds remaining sealed, and is slower growing. Correspondingly, it has a later pollination season. These species are naturally challenged by strongly overlapping sets of pathogens, notably ones causing needle casts and shoot galls.