4. Individual and population fitness and some ramifications
Relationships between individual and population fitness in this connection can be problematic. Population fitness depends not just on the metric values for the fitness traits in its habitat which determine individual fitness, but also on the amount of heritable variation for those traits. In the poplar case, key adaptive traits in the wild would include pathogen resistance, timing of seasonal growth, and the sexual reproductive potential that allows genetic selection to operate. Thus, the impact of some ill-adapted segregants can be outweighed by a concurrent incidence of segregants that excel their parents’ fitness, since the latter contribute most to the parentage of the next generation. However, with pathogens involved, population fitness can be complicated by the nature of infection processes. With a ‘polycylic’ pathogen, there can be both reinfection from within the host individual and cross-infection between individuals. For instance, where pathogens impair fitness, an individual becoming deciduous gains little or no protection if the population remains evergreen and can thereby provide a continuous supply of inoculum. In this way, becoming deciduous is unlikely to be advantageous, unless either this results in asynchrony between the window of host susceptibility and presence of inoculum or individuals are too sparse to favour cross-infection. An individual becoming evergreen, however, might impair the fitness of the rest of the population through producing a continuous supply of inoculum, unless it suffers so badly from internal reinfection as to incur selective elimination. To meet that condition, internal reinfection would presumably need to predominate strongly over cross-infection between individuals, which may be common with polycyclic infection.
Becoming evergreen could require evolution of enhanced pathogen resistance and/or some more direct fitness advantage of the habit. Natural selection for resistance may be limited by ‘selective opportunity cost’ whereby the fitness gains from additional resistance are in trade-off with the fitness gains from responses to other selective pressures. Some of those gains, it may be noted, will be needed to make good the decay of non-additive genetic components of fitness that occurs upon sexual reproduction. Given the dynamic nature of biotic interactions, along with pathogen mutations, natural selection for ecological fitness can be expected to entail a “red queen” model (Van Valen 1973) requiring endless genetic shifts to maintain fitness. With long-lived perennials, however, such shifts may occur most readily within generational cohorts, during lifespans. With the complex and highly dynamic poplar rust pathosystem, poplar plantations have proved notoriously vulnerable to pathotype mutations (e.g. Persoons et al. 2014). Such plantations are typically monoclonal, which makes them convenient to produce and manage. Present understanding of pathosystems suggests that natural populations of poplars, reproducing by seed, might be expected to be more resilient towards the rusts. However, fully functioning natural populations, which would provide a good basis for comparison, are becoming increasingly scarce.
It must be noted that the biotic vulnerability of poplar plantations, based on species that are Melampsora hosts, contrasts with the biotic resilience of natural stands of American (or trembling) aspen (Popiulus tremuloides Michaux) which are typically monoclonal (Latutrie et al. 2015). The aspen, with different pathogens, is evidently involved in very different pathosystems.
5. Pre-adaptive relationships ?
As indicated above, seasonal timing of phenological events may serve as defence against pathogens through avoidance of infection seasons while the plant is vulnerable. Alternatively, the timing may be a source of vulnerability. This seems evident in some cases of freshly introduced pathogens, rather than in historically co-evolved cases. With Dutch elm disease, caused by ascomycete fungi Ophiostoma spp, early bud burst and flushing, while possibly incurring climatic vulnerability, can evidently avoid the height of the infection season (Ghelardini & Santini 2009). With sudden oak death, caused by Phytophthora ramorum Werres et al., host vulnerability has been linked to timing of bud burst and onset of cambial activity in the hosts (Dodd et al. 2008). Similarly, the pathogen Austropuccinia psidii (G. Winter) Beenken, a rust affecting a wide range of hosts in the Myrtaceae that has invaded parts of Asia and the Pacific, can only infect actively flushing host tissues (Beresford et al. 2020). In each case, invasive pathogens appear likely to place strong selective pressures on the phenology of native host populations, and understanding the potential for adaptive responses appears to be worth further study.
6. Concluding
We propose that pathogens can be strong evolutionary drivers of seasonal phenology, especially in respect of deciduousness. This appears to be largely novel. Some evidence is given for our proposal for deciduousness in a temperate zone. In the tropics, the proposal is made for want of any other satisfactory explanation. Other examples are given of how pathogens may influence seasonality of shoot flushing, where simple climatic explanations seem inadequate. We suggest that evolutionary barriers can operate against a shift from a deciduous to an evergreen habit in some temperate-zone cases.
For herbivory we have proposed that predator satiation may operate as an evolutionary driver in addition to purely climatic adaptation to the occurrence of winters. Herbivory is also proposed as an evolutionary constraint on growing or flowering seasons. For the tropics, where climates are largely characterised by wet and dry seasons, predator satiation has already been implicitly identified as a quite common evolutionary driver of seasonal phenology.
Ontogenetic phenology appears strongly influenced evolutionarily by biotic factors, including both herbivory and pathogens.
How a particular situation plays out will depend on many factors, notably the general climatic context and the constraints it imposes (along with other environmental factors), and the nature of biotic pressures, viz pathogens and/or herbivory. More specifically, the factors for pathogens involve the reproductive biology of pathogens including seasonality; the host plant ecology, in the scale from highly gregarious to essentially solitary; and the genetic systems of both host and pathogens. For herbivory, they include the life cycles and seasonal abundance of insects, along with the nature and abundance of vertebrate browsers. Not to be ignored are interactions with other plants in the ecosystem and among herbivores.
Influences of climate are both direct and indirect, the latter category involving how much climate favours pathogens and herbivores. With pathogens, there is the classic ‘triangle’ of interactions involving host, pathogen and the environment, so the direct and indirect evolutionary influences are in some degree interdependent. With the poplar/Melampsora pathosystem, a direct influence of a climate that is milder than the native one may favour an evergreen habit. But the indirect influence, through the effect of the same climate on the pathogen, is likely to militate against that shift. Often, however, the evolutionary pressures from climatic and biotic factors would be mutually reinforcing, as in the Medicago case.
In calling for a broad evaluation of biotic factors as evolutionary drivers (or co-drivers) of plant phenology, the coverage is preliminary and selective, representing an alert to the topic. Even a more systematic literature review would still require interpretation of individual cases. Challenges certainly exist in such evaluation. To identify likely cases of biotic drivers, we consider it appropriate to look largely to those who are closely familiar with individual plant species and their ecology. Then detailed review and analysis of available evidence is needed, but interpretation in individual cases may still be problematic. From there, the general postulate of biotic influences will need to be carried forward into hypotheses that are observationally or experimentally testable. While woody perennials have provided tantalising case histories, they can also pose major logistical challenges in testing hypotheses.