*3.4. Can Evolutionary Adaptation Explain the Phenotype of an Accession?*

Adaptation to a specific ecological niche shapes an organism's genome and its response to changes in its surroundings [33]. Correlation analysis revealed a greater reduction in ΦPSII under FL the further north on the globe the accession had originated from (Figure 8). This finding could potentially be explained by the following factors, considering that the main phase of vegetative growth of most Arabidopsis accessions is in early spring: (i) lower temperatures with increasing latitude, thus, accessions from further north may have optimized their photosynthesis to lower temperatures than those in our experiment (16 ◦C night, 21 ◦C day); (ii) differences in day length, because changes in gene expression that drive acclimation in response to FL have been shown to intercept with the circadian clock [3]; and/ or (iii) light intensity and availability.

A negative correlation was also found between Fv/Fm and number of leaves until flowering (Figure 7C). The number of leaves until flowering is an indicator for whether Arabidopsis accessions germinate in spring or whether they germinate in autumn and require a long cold period in winter (vernalization) before flowering. Thus, it is tempting to speculate that species which form many leaves before flowering germinate in autumn and are thus better adapted to photosynthesis under cold temperatures, as they overwinter. A decrease in Fv/Fm is associated with sustained photoinhibitory quenching (qI) that accompanies PSII damage. It has been speculated that slowly reversible qI may have a photoprotective function under some conditions [34]. Thus, qI may be prematurely switched on in winter-grown accessions in order to protect against cold-stress.

#### **4. Materials and Methods**

## *4.1. Plant Material, Growth Conditions and Treatments*

Thirty-five accessions of *Arabidopsis thaliana* were randomly selected from a collection of 330 accessions, and Col-0 was additionally selected as a reference genotype (36 accessions were used in total) due to its use as a wildtype in many reverse genetics studies. Information on country of origin, longitude, and the number of leaves until flowering (Table 1) was accessed on the 1001 genomes website (https://1001genomes.org/; [35]). In some cases, the Google Maps Arabidopsis viewer was additionally used to locate an accession. Latitudes for Col-0, Ler-1, and Ws-0 were omitted from Table 1 and correlation analyses (see below) as these genotypes have been cultivated in laboratories for decades and may not anymore be representative of the original accessions. All accessions are from the Northern Hemisphere, 28 (78%) were initially collected in Europe and another eight in Asia and North America (Table 1).

Seeds were sown on substrate prepared for Arabidopsis ('Arabidopsis substrate'; 70% white peat, 20% vermiculite, 10% sand; Stender, Schermbeck, Germany) which was enriched with1gL−<sup>1</sup> each of two standard fertilizers: Osmocote Start® (ICL Specialty Fertilizers, Tel Aviv, Israel; composition: 11% N, 11% P2O5, 17% K2O, 2% MgO, 0.38% Fe, 0.05% Mn, 0.01% B, 0.09% Cu, 0.009% Mo, 0.014% Zn) and Triabon® (Combo Expert, Münster, Germany; composition: 16% N, 8% P2O5, 12% K2O, 4% MgO, 9% S, 0.02% B, 0.04% Cu, 0.1% Fe, 0.1% Mn, 0.02% Mo, 0.01% Zn; [36]). For the first 14 days, plants that were later used in the uniform light treatment were grown under a 16 h photoperiod, at ~150 μmol m−<sup>2</sup> s−<sup>1</sup> photosynthetically active radiation (PAR; 400–700 nm), while plants later used for the fluctuating light treatment were grown in a 12 h photoperiod, at 250 μmol m−<sup>2</sup> s−<sup>1</sup> PAR, for the first 14 days. This means that FL grown plants were initially exposed to a higher daily light integral than those in U (10.8 mol photons d−<sup>1</sup> and 8.6 mol photons d<sup>−</sup>1, respectively). Then, single plants were placed in a 0.11 L pot containing Arabidopsis substrate, and exposed to the light treatments until they were 28 days old. Day/night temperatures and relative humidity were 20/16 ◦C and 60/75% in all cases (for light and temperature recordings in the growth chambers, see Figure S3).

The treatments were as follows: uniform light (U) of 250 μmol m−<sup>2</sup> s−<sup>1</sup> PAR, and fluctuating light (FL) of alternating cycles of 900 μmol m−<sup>2</sup> s−<sup>1</sup> and 90 μmol m−<sup>2</sup> s−<sup>1</sup> PAR for one and four minutes, respectively (average light intensity: 252 μmol m−<sup>2</sup> s−<sup>1</sup> PAR). In both treatments, the photoperiod was 12 h, totaling 144 high/low light cycles in the FL treatment. Changes between the two light intensities in the FL experiment were very rapid and accurate (Figure S3). Three types of LED lamps (Roschwege, Greifenstein, Germany) were used in both treatments: white (LED-Star 2700 K 10 W), red (LED-Star DR 660 nm 5 W) and blue (LED-Star DB 460 nm 5 W). The output setpoints of all LED lamps were kept identical for any given light intensity, ensuring that there were no changes in light spectrum as light intensity changed. Accessions under the FL treatment were grown in two separate experiments: 20 accessions were grown in experiment 1, 15 different accessions were grown in experiment 2, and Col-0 was grown in both experiments (Table 1). Utmost care was taken that conditions in both FL experiments were identical. In the U treatment, all accessions were grown in one experiment. Plants were watered 2–3 times per week depending on substrate wetness to the touch (as per usual practice in

the institute). To account for position effects in the climate chamber, plants were randomized during the treatment period.
