*2.2. Content and Composition of Volatile Organic Compounds (VOCs)*

Applying GC-MS analysis, 12 monoterpenes and one sesquiterpene were identified in the leaf extracts of *Thymus vulgaris* L., representing ≥ 94% of all detected volatile constituents. Major identified volatile organic compounds (VOCs) in the leaf extracts of *Thymus vulgaris* L. under the supplemental lighting systems were thymol, *γ*-terpinene, and *p*-cymene, respectively, which is consistent with the results of former research [48,49]. Thereby, the chemical makeup remained unaffected by the different lighting systems. The total content of VOCs per g of LDM is highly enhanced by LED (2.7%) and HPS (2.3%) as compared to by FL (1.1%). The difference in quantity of VOCs per g of LDM between thyme plants cultivated under LED and HPS is not significant (*p* = 0.088). The LED considerably increased the amounts of all 13 evaluated terpenoids in the leaves of *Thymus vulgaris* L. in contrast to the FL system. The HPS system also enabled considerable increases in comparison to the FL system, even though the differences between the amounts of *γ*-terpinene and borneol are less profound, with *p* = 0.099 and *p* = 0.075, respectively. Differences between LED and HPS treatments were only detected for *α*-pinene, while myrcene (*p* = 0.077) as well as limonene (*p* = 0.057) differed only in tendency. All results are summarized in Table 2.


**Table 2.** Effect of three different supplemental lighting systems on the chemical composition of 13 main volatile organic compounds (VOCs) of *Thymus vulgaris* L. cultivated in the greenhouse during fall and winter of Berlin, Germany.

\* Retention indices (RI) relative to C6-C<sup>24</sup> n-alkanes on a HP-5MS column for compound identification. Indices are presented as means ± *SD* with *n* = 192. \*\* Percentages were calculated from GC-FID TIC data after weight correction and presented as means ± *SD* with *n* = 64. \*\*\* Amounts of major compounds were calculated based on density corrected calibration functions obtained from reference standards analyzed under the same GC-FID conditions as the samples. Presented are mean amounts of volatile compounds (µg 100 mg−<sup>1</sup> LDM (=leaf dry matter)) of four independent spatial replications per light treatment (*n* = 4) ± *SD* of 16 collected dried leaf samples per spatial replication and light treatment (*N* = 192, *n* = 64 dry leaf samples per supplemental light treatment, *n* = 16 dry leaf samples per spatial replication). Significant differences (*p* ≤ 0.05) were determined according to Dunnett's T3 multiple comparisons test after Brown-Forsythe and Welch ANOVA test (*p* ≤ 0.02). Different letters within a row indicate significant differences at *p* ≤ 0.05, and bold amounts indicate significant differences at *p* ≤ 0.1. \*\*\*\* Percentage of total VOCs (volatile organic compounds) was calculated based on the results of the internal standard (6-methyl-5-penten-2-one), which was co-analyzed in each sample. Presented are mean percentages per g LDM (% g−<sup>1</sup> LDM (=leaf dry matter)) of four independent spatial replications per light treatment (*n* = 4) ± *SD* of 16 collected dried leaf samples per spatial replication and light treatment (*N* = 192, *n* = 64 dry leaf samples per supplemental light treatment, *n* = 16 dry leaf samples per spatial replication). Significant differences (*p* ≤ 0.05) were determined according to Dunnett's T3 multiple comparisons test after Brown-Forsythe and Welch ANOVA test (*p* ≤ 0.001). Different letters within a row indicate significant differences at *p* ≤ 0.05 and bold amounts indicate significant differences at *p* ≤ 0.1.

> Gouinguene and Turlings (2002) showed in their study that young corn plants (*Zea mays* L.) significantly increased their emissions of volatiles as light intensity increased up to 10,000 lm [50]. However, beyond 10,000 lm volatile emissions in *Zea mays* L. did not enhance any further, suggesting a kind of saturation or limitation was reached. The authors of [51,52] also detected this proposed light quantity-dependency. Multiple studies also suggest that terpene synthesis involves phytochromes [51–54], red and far-red light-sensing photoreceptors (reviewed by [55]), making the production of terpenes also dependent on light quality, specifically on the *R/FR* ratio [56]. For example, in thyme seedlings, red light strongly promoted the production of mono- and sesquiterpenes (thymol, *γ*-terpinene, *p*-cymene and carvacrol, *β*-caryophyllene) and the number of essential oil-containing trichomes per cotyledone, two stimulatory effects that proved to be completely reversible by a subsequent exposure to far-red irradiation [53,54]. Later, a partial reduction of volatile emissions was detected in *Arabidopsis thaliana* (L.) Heynh exposed to a low *R/FR* ratio of 0.2 as compared to plants exposed to a high *R/FR* ratio of 2.2 when controlling for light intensity [56]. These findings could explain the comparatively low VOC contents detected

in thyme leaves grown under FL, as both their light intensity (*PFD* 60 µmol m−<sup>2</sup> s −1 , *PFD*-R 7.6 µmol m−<sup>2</sup> s −1 ) as well as their *R/FR* ratio (0.1) were significantly reduced as compared to LED and HPS under our experimental conditions. It would also suggest that the similar contents of VOCs per gram of thyme leaves found under LED and HPS are the result of their similar high *R/FR* ratios of 2.8 and 2.4, respectively. However, *R/FR* ratios dramatically decline under vegetational canopies. As described by Franklin (2008), a single leaf reduces a given *R/FR* ratio of 1.2 to 0.2 for the leaves growing underneath, and the ratio reduces further to 0.1 underneath a second leaf [55]. As leaf and shoot yields of *Thymus vulgaris* L. significantly increased under the LED system, it is reasonable to believe that the actual *R/FR* ratio underneath the more densely stands of thyme plants grown under the LED system was much lower than under the less dense canopy of thyme plants grown under HPS. This idea coincides with the result from Kegge et al. (2013), who detected a reduction of volatile emissions in plants grown in high density stands [56]. Another explanation for the similar contents of VOCs per gram of thyme leaves found under LED and HPS may be found in the high B light proportion found under the broad LED light spectrum, as it was recently shown that essential oil contents of *Thymus vulgaris* L. decrease with increasing proportions of blue light [57]. The associated suppressions of terpene synthesis under low *R/FR* ratios as well as under low *R/B* ratios may have been partially compensated by the LEDs' elevated light intensity (*PFD* 232 µmol m−<sup>2</sup> s −1 , *PFD*-R 55.3 µmol m−<sup>2</sup> s −1 ) compared to the intensity of the HPS system (*PFD* 143 µmol m−<sup>2</sup> s −1 , *PFD*-R 24.3 µmol m−<sup>2</sup> s −1 ) in our study. Additionally, though air temperatures under the given experimental conditions did not differ between the supplemental lighting systems, it is known that leaf temperature increases under HPS lights as comparted to other lighting systems [58]. As elevated temperatures evidently increase the emission of volatiles [50], a greater leaf temperature under HPS may have been present and contributed to the terpene synthesis in HPS-grown thyme plants. Further, as we did not adjust fertilization, though the LED system yielded much greater biomasses than FL and HPS, it is plausible that a reduced nutrient availability for LED-grown thyme plants limited their production of VOCs, as demonstrated by Gouinguene, and Turlings (2002), who showed that fertilization rate positively effects volatile emissions [50].

Nevertheless, as the LED lights were able to increase the production of volatiles in thyme leaves significantly compared to the HPS lights, the LEDs' volatile productivity per square meter doubled in absolute terms (2.5 vs. 1.3 g m−<sup>2</sup> ) with *p* < 0.06 (Table 3).
