Phenolic concentration under (simulated) climate change .41

In the greenhouse experiment, elevated CO2 concentration increased the concentrations of salicylates and phenolic acids in stem bark, and as a result the plants exposed to elevated CO2 concentration had a higher concentration of total low-molecular-weight phenolics compared with plants grown in ambient conditions (I). Many earlier studies with deciduous tree species have also reported that elevated CO2 concentration increased the synthesis of secondary metabolites, although there are reports that an elevated CO2 concentration decreased secondary metabolites or had no

42

effect (Peltonen et al. 2005; Lindroth 2012; Lavola et al. 2013; Nissinen et al.

2016). Peñuelas and Estiarte (1998) suggested that elevated CO2

concentration increases the carbon-to-nitrogen ratio in plants, and consequently the utilisation of carbon for growth is diminished, therefore the surplus carbon is allocated to the synthesis of carbon-based secondary metabolites. In this study, the CO2-treated plants might have nitrogen deficiency and as a result the surplus carbon might have been allocated to synthesize carbon-based secondary metabolites.

On the other hand, elevated temperature decreased most of the individual compounds of salicylates, flavonoids, and phenolic acids, with a consequent decrease in the concentration of total low-molecular-weight phenolics in stem bark of P. tremula (I). In response to elevated temperature, the plants had greater growth. During a period of higher growth, L-phenylalanine, a precursor of proteins and phenolic compounds (Matsuki 1996), is apparently diverted away from phenolic synthesis, and thus downregulates synthesis of secondary metabolites relative to protein synthesis (Tuomi et al. 1991; McDonald et al. 1999; Fritz et al. 2006). In consistence with the greenhouse experiment, elevated temperature also decreased the concentration of total low-molecular-weight phenolics in P.

tremula stems in the field experiment, however it did not show the temporal variation during the three-year growth period, which contrasts with our expectation (II). Yet, the concentration of condensed tannins varied over time in response to elevated temperature, and as a result the concentration of condensed tannins increased in stems in the second growing season when compared to the first season (II). These results are in accordance with the growth differentiation balance hypothesis, which states that when growth increment in plants slows down, it makes the carbon substrates available for phenolic synthesis (Herms and Mattson 1992). The decline in proportional increment in the diameter growth rate in temperature-treated plants through time might have facilitated resources that could be allocated to condensed tannins.

There was no effect of elevated UVB radiation on salicylates, flavonoids, phenolic acids, and condensed tannins in stems of P. tremula in any

43 particular year or over the entire three-year growth period. Although salicylates do not show very strong responses to elevated UVB radiation, flavonoids, such as quercetins and phenolic acids, are often reported to increase in plants exposed to elevated levels of UVB radiation (Julkunen-Tiitto et al. 2005; Nybakken et al. 2012; Nissinen et al. 2017). The induction of flavonoids and phenolic acids in plants in response to elevated UVB radiation, however, can differ depending on the plant part. Tissue thickness and anatomical surface features of different plant organs affect the attenuation of UVB radiation in plants (Day et al. 1992; Schreiner et al. 2009).

The relatively thick outer cell layer and exposed surface area of P. tremula stems may reduce the penetration of UVB radiation and thus minimize the synthesis of UVB-radiation-induced phenolics.

3.2.2 Effect of bud removal on phenolic concentration under elevated temperature and UV radiation

Alongside growth, I investigated the effect of mild (~5%) bud removal on phenolic concentration in the leaves of P. tremula which were grown under elevated temperature and UV radiation (III). The concentrations of salicylates and phenolic acids did not change due to the removal of buds as both intact and bud-remove individuals had a similar level of concentration in the leaves (III). However, bud-removed individuals had greater concentrations of flavonoids and condensed tannins as compared to the intact plants.

Although salicylates are very abundant in Populus species, they are highly genetically determined and are less responsive to environmental stresses such as lack of resources and tissue damage than other phenolic compounds (Roth et al. 1998; Hemming et al. 1999; Osier et al. 2000). Roth et al. (1998) found that salicylates are minimally affected, although levels of condensed tannins showed strong responses due to tissue damage in P.

tremuloides and Acer saccharum (Marshall.), even when the damage was severe.

In this study, greater accumulation of flavonoids and condensed tannins in leaves due to bud removal was not expected, because bud-removed individuals also had greater biomass growth. More growth in plants should

44

cost them via a reduction in the synthesis of defensive chemical constituents (Bryant et al. 1983; Herms and Mattson 1992). Removal of buds might have activated induced defense in the plants, and as a result the concentration of flavonoids and condensed tannins increased to mitigate future damage. It has been suggested that induced defense is activated in plants when tissue is damaged or removed or in response to abiotic stresses (Kessler and Baldwin 2002; Young et al. 2010; Jiang et al. 2018). Although induced defense incurs costs, it may be relatively cheap because response to damage enhances a plant’s competitiveness, and induced secondary chemical constituents are only synthesized when needed (Miranda et al. 2007; Young et al. 2010; Hood and Sala 2015). Moreover, higher concentrations of flavonoids and condensed tannins in association with higher biomass growth in bud-removed individuals can be partly a result of the fact that P.

tremula is an inherently growing species. Earlier studies found that fast-growing trees are generally better capable of compensating for the tissue damage than slow-growing ones owing to their greater soil nutrient acquisition capability, which increases their utilization of carbon for growth (Lavigne et al. 2001; Hikosaka et al. 2005; Endara and Coley 2011; O’Reilly- Wapstra et al. 2014). Accordingly, these bud-removed plants, under favorable conditions, might have prioritized both growth and defense instead of allocating resources to only one of these processes.