Plant material

For this study, in vitro cultured seedlings -approximately 1.0 cm tall and three month oldwere used. Seedlings were obtained from the germination of seeds of O. tigrinum and L. autumnalis plants that belong to Mexico’s National Phyto Genetic Resources System (SINAREFI), held by President Juárez Agrobiology School of the Universidad Michoacana de San Nicolás de Hidalgo, in Uruapan, Michoacán.

Culture medium

A Murashige and Skoog (MS) medium with no plant hormones, plus thiamine (0.4 mg L^-1), inositol (100 mg L^-1), saccharose (30 g L^-1), agar (6 g L^-1), and with pH adjusted at 5.7 was used to grow seedlings. This medium was sterilized in an autoclave, during 15 minutes, at 121 °C.

Establishment and experimental design

For this study, 100 mL glass jars with 20 ml of MS medium were used and 30 seedlings were placed on each one. The procedure was performed in a laminar flow hood. The labeled jars were incubated in 0.10125 m³ plywood boxes painted in white on the inside. On the top of each box an aluminum base with four SiLed^® high-efficiency 1 W LED lights was placed. LEDs were red, blue or white, coupled in series to an external 12 V source that was connected to the 110 V AC.

Experimental design was completely random and treatments were: red and blue LEDs in a 3:1, 2:2, and 1:3 ratios, and white LEDs (Table 1). Seedlings in the control group were incubated with a Phillips white fluorescent 39 W lamp. Each treatment was repeated six times and the experimental unit was a jar with 30 seedlings.

Wavelengths emitted by the light sources were measured with an ASD^® (Analytical Spectral Devices) field spectrum radiometer (Field Spec Pro). Blue and red LEDs emitted spectrums featuring a single emission peak centered at 458 and 636 nm respectively, of Gaussian appearance and with a standard deviation of 13 nm. On the one hand, white LEDs had a 444 nm emission peak of (also with a 13 nm standard deviation) and a wide area of emission centered around 553 nm (but with a 50 nm dispersion). On the other hand, fluorescent lamps had a series of close peaks at 436, 490, 545, 586, and 612 nm (Figure 1). Light intensity was determined with the radiation sensor of a portable photosynthesis system (IRGA6400, LI-COR^®); the resulting intensity was 31±2 µmol m^-2 s^-1 in all treatments, with a photoperiod of 16 h of light and 8 h of dark and an average temperature of 25 °C.

Figure 1 Characteristics of the wavelengths emitted by LED lights (blue, red, and white) and fluorescent lamp, measured with an ASD ^® (Analytical Spectral Devices) field spectrum radiometer (Field Spec Pro). Normalized emission is equivalent to the percentage of total visible energy found on each 1 nm band. 

Measured variables

Seedling and root length (cm) was measured from the base and the apex of the plantlet or the root, respectively, using graph paper, on 10 seedlings for each treatment and repetition.

The number of leaves and roots was also registered in 10 seedlings per treatment and repetition.

Seedlings’ fresh weight (mg), 10 seedlings of each treatment were selected and weighted in an OHAUS^® analytical electronic balance with a 0.01 mg readability.

Seedlings’ dry matter (%). This variable was measured in a 100 mg sample of seedling’s fresh matter, one for each treatment and repetition, using the gravimetric method. Drying was performed in a FELISA oven, at 70 °C during 72 h. Dry samples were weighted in an OHAUS^® analytical electronic balance and results are shown as percentage of dry matter with respect to its fresh weight.

Seedlings’ pigment content (mg g^-1). a, b, and total chlorophyll concentration was determined in 100 mg samples of each treatment and repetition, extracting the leaves and measuring them through spectrophotometry (Porra et al., 1989). Leaves were placed in glass jars with 3 mL of N-dimethylformamide covered with aluminum foil. The jars were stored at 4 °C during 48 h. Afterwards, a 1 mL sample was placed in a cell to testmeasure absorbance in a JENWAY 6320D spectrophotometer; standard deviation of 13 nm. On the one hand, white LEDs had a 444 nm emission peak of (also with a 13 nm standard deviation) and a wide area of emission centered around 553 nm (but with a 50 nm dispersion). On the other hand, fluorescent lamps had a series of close peaks at 436, 490, 545, 586, and 612 nm (Figure 1). Light intensity was determined with the radiation sensor of a portable photosynthesis system (IRGA6400, LI-COR^®); the resulting intensity was 31±2 µmol m^-2 s^-1 in all treatments, with a photoperiod of 16 h of light and 8 h of dark and an average temperature of 25 °C.

Measured variables

Seedling and root length (cm) was measured from the base and the apex of the plantlet or the root, respectively, using graph paper, on 10 seedlings for each treatment and repetition.

The number of leaves and roots was also registered in 10 seedlings per treatment and repetition.

Seedlings’ fresh weight (mg), 10 seedlings of each treatment were selected and weighted in an OHAUS^® analytical electronic balance with a 0.01 mg readability.

Seedlings’ dry matter (%). This variable was measured in a 100 mg sample of seedling’s fresh matter, one for each treatment and repetition, using the gravimetric method. Drying was performed in a FELISA oven, at 70 °C during 72 h. Dry samples were weighted in an OHAUS^® analytical electronic balance and results are shown as percentage of dry matter with respect to its fresh weight.

Seedlings’ pigment content (mg g^-1). a, b, and total chlorophyll concentration was determined in 100 mg samples of each treatment and repetition, extracting the leaves and measuring them through spectrophotometry (Porra et al., 1989). Leaves were placed in glass jars with 3 mL of N-dimethylformamide covered with aluminum foil. The jars were stored at 4 °C during 48 h. Afterwards, a 1 mL sample was placed in a cell to testmeasure absorbance in a JENWAY 6320D spectrophotometer; measurement was performed with the following wavelengths: 664 nm for chlorophyll a and 647 nm for chlorophyll b. Absorbance data was used to calculate the concentration of each chlorophyll through the following equations: chlorophyll a=12.70 (A₆₆₄)-2.97 (A₆₄₇); chlorophyll b=20.70 (A₆₄₇)-4.62 (A₆₆₄) and total chlorophyll=17.90 (A₆₄₇)-8.08 (A₆₆₄).

Percentage of necrotic and albino plants was only measured for O. tigrinum. Each experimental unit (jar) was divided into quadrants to facilitate counting. Then, normal, necrotic or albino plants were counted by quadrant and added to obtain the total number of each condition per jar. With the total sum of normal, necrotic, and albino plants per experimental unit, the total number of plants per experimental unit was obtained. This number was then used as denominator to calculate the percentage of necrotic or albino plants with the formula: percentage of necrotic or albino plants per experimental unit=(Number of necrotic or albino plants per jar/Total number of plants in the jar)×100.

Variable recordings for O. tigrinum were performed 30 d after treatments started, while variable recordings for L. autumnalis were performed after 90 d.

Data analysis

With the data a variance analysis and a Tukey test (p≤0.05) were performed with SAS 9.0. To better illustrate mean comparisons between treatments, R software 3.1.1 was used, which provides a graphic representation of the differences. The Shapiro test was used to verify distribution normality in the data (^{R Core Team, 2014}).

Oncidium tigrinum development

Oncidium tigrinum seedlings did not have significant differences between treatments in terms of fresh matter weight, total leaf number, or a, b or total chlorophyll. The exposure to the combination of red and blue spectrums was expected to increase the concentration of pigments and fresh weight, given that: 1) a and b chlorophylls reach their maximum absorbance in the blue and red range (Pimentel et al., 2007); 2) light has a strong influence in cellular multiplication and in in vitro cultured plant tissue growth (^{Araujo et al., 2009}); 3) an increase in biomass and photosynthetically active pigment concentration (a, b, and total chlorophyll) as a result of the exposure to the combination of red and blue light (4:1 and 7:3) had already been documented on seedlings of another species of Oncidium (^{Mengxi et al., 2011}).

In our study significant differences in seedling length, root number, and size were found, as well as differences in dry matter, albino, and necrotic plant percentages. The longest seedlings (1.64 cm) were obtained with the combination of red and blue light (1:3); however, they were statistically similar to other treatments except T2, where the size of seedlings was reduced by 28 % (Figure 2). T2 also inhibited root production and length (Figure 3A). T1 only reduced root length (Figure 3B).

Figure 2 Effects of T1 (3:1 red and blue LEDs), T2 (2:2 red and blue LEDs), T3 (1:3 red and blue LEDs), T4 (white LEDs), and T5 (white fluorescent light) light treatments on in vitro cultured Oncidium tigrinum seedlings’ length. Different letters on each figure indicate significant differences (Tukey; p≤0.05). 

Figure 3 Effect of T1 (3:1 red and blue LEDs), T2 (2:2 red and blue LEDs), T3 (1:3 red and blue LEDs), T4 (white LEDs), and T5 (white fluorescent light) on root number (A) and length (B) of Oncidium tigrinum in vitro cultured seedlings. Different letters on each figure indicate significant differences (p≤0.05). 

The reduced root length on T1 and T2 (larger and equal ratio of red versus blue light, respectively) could be attributed to what was noted by ^{Correll and Kiss (2005)}. They claim that red light receptors (A and B phytochrome) have an inhibitory effect on root elongation. White light -which has less radiation ratio in the red region of the spectrumstimulates rooting in Cattleya loddigessi seedlings (^{Araujo et al., 2009}) and, together with green light, boosts bud rooting of Dendranthema×grandiflora ‘Lilac Wonder’ (^{Miler and Zalewska, 2006}). Our study confirms that white light from LEDs and fluorescent lamps promoted rooting of in vitro O. tigrinum seedlings and that these light sources emitted the least amount of radiation in the red region, compared to all other studied spectrums.

Micropropagated seedlings had good quality in all lighting treatments and degenerative changes in leaves and pseudobulbs were not found (Figure 4), apart from some cases with total chlorophyll absence. The highest incidence of albino seedlings (33.7 %) was found on T5 (fluorescent light) and was similar to that of T4 (15.8 %), but significantly different from the other treatments, which showed an average of 2.16 % of albino seedlings (Figure 5). These findings seem to indicate that fluorescent lamps and white LEDs light emission on the red and blue regions of the spectrum are not enough to ensure that some O. tigrinum seedlings can perform normal photosynthesis.

Figure 4  In vitro Oncidium tigrinum seedlings lighted during 30 days with different light combinations T1 (3:1 red and blue LEDs), T2 (2:2 red and blue LEDs), T3 (1:3 red and blue LEDs), T4 (white LEDs), and T5 (white fluorescent light). 

Figure5. Effecto flight treatments T1 (3:1 redand blue LEDs), T2 (2:2 red and blue LEDs), T3 (1:3 red and blue LEDs), T4 (white LEDs) and T5 (white fluorescent light) on the incidence of albino seedlings of in vitro cultured Oncidium tigrinum. LSD: Least significant difference. In the figure, different letters indicate significant differences (Tukey; p≤0.05). 

Albinism is an evolutionary genetic characteristic that has not been documented on the Oncidium genus, although it occurs in some individuals of orchid populations of the Epipactis (^{Stöckel et al., 2011}) and Cephalanthera (^{Roy et al., 2013}) genera, which grow under canopies. In any other green plants the photosynthetic process is decreased due to the reduced ambient light or the low efficiency of their photosynthesis apparatus or both (^{Roy et al., 2013}). Furthermore, during the first stages (protocorm formation), Oncidium sphacelatum seedlings are completely myco-heterotrophic, i.e., they need to establish a symbiotic relationship with a mycorrhizial fungus that provides them with a carbon source (^{Valadares et al., 2014}).

The largest percentage of dry matter (12.2 %) was obtained with T1; it was statistically larger than T3 -the latter had 4% less biomass-, but it was similar to the other treatments (Figure 6A).

Figure 6 Effect of light treatments T1 (3:1 red and blue LEDs), T2 (2:2 red and blue LEDs), T3 (1:3 red and blue LEDs), T4 (white LEDs), and T5 (white fluorescent light) on dry matter accumulation (A) and in the incidence of necrotic seedlings (B) of in vitro cultured Oncidium tigrinum. Different letters on each figure indicate significant differences (p≤0.05). 

Similar responses have been found in in vitro conditions with light intensities of 70 µmol m^-2s^-1. In Lilium oriental hybrid ‘Pesaro’, blue+red (1:1) LED lighting and fluorescent light increased bulbils’ dry matter (^{Lian et al., 2002}) and, in Dendrobium officinale, red+blue (1:1) light improved biomass concentration in protocorms (^{Lin et al., 2011}). Also, blue-only lighting reduces dry matter (^{Lian et al., 2002}); which is why in our study, the lowest registered value from T3 might be the result of the highest proportion of blue light versus red light. T3 had the largest percentage of necrotized seedlings (57.7 %), and it may be possible that tissue necrosis is due to photoinhibition or photooxidation related damage, as ^{Araujo et al. (2009)} and ^{Casierra-Posada et al. (2011)} suggest. However, this percentage was similar (p>0.05) to the rest of the treatments, except T4 that significantly decreased the incidence of necrotized seedlings. i.e., with 100 % white LED light there was almost no observable damage in the seedlings (Figure 6B).

Laelia autumnalis development

Significant differences on seedling and root length, number of leaves and roots, in seedling’s fresh and dry weight, and in a, b, and total chlorophyll content were found in this species. The longest seedlings were obtained in treatments with the highest emissions of light in the blue region: T3 (3.26 cm), T1 (3.04 cm), and T2 (2.93 cm), statistically similar, but superior to T4 (2.65 cm), and T5 (1.88 cm) (Figure 7A). However, this elongated plants presented degenerative morphological changes, with thinner stems and leaves than the other treatments (Figure 8). This could be explained partly because blue light increases plants’ water consumption to their maximum cellular capacity, which can produce increased tissue length (^{Ribeiro et al., 2009}). Contrary to our study’s results, ^{Ribeiro et al. (2009)} did not find differences in Zantedeschia aethiopica seedlings length as a result of the quality of white, red, blue, and green light; however, in Paphiopedilum sprouts, ^{Lee et al. (2011)} observed apical dominance loss (small plants) as a result of blue light.

Figure 7 Effect of light treatments T1 (3:1 red and blue LEDs), T2 (2:2 red and blue LEDs), T3 (1:3 red and blue LEDs), T4 (white LEDs), and T5 (white fluorescent light) on leaf length (A) and number of leaves (B) of in vitro cultured Laelia autumnalis seedlings. Different letters on each figure indicate significant differences (p≤0.05). 

Figure 8  In vitro Laelia autumnalis seedlings irradiated with different combinations of light during 90 days: T1 (3:1 red and blue LEDs), T2 (2:2 red and blue LEDs), T3 (1:3 red and blue LEDs), T4 (white LEDs), and T5 (white fluorescent light). 

Average leaf number in all treatments was over six, except in T1 where the average was just five, although only T2 was statistically different with 8.23 leaves per plant (Figure 7B). This could be a reflection of the low proportion of blue light versus red light in T1 (1:3), compared with T2 (2:2). Lateral bud and leaf imposition in Anthurium seedlings (^{Budiarto, 2010}) and Dendrobium officinale PLBs (^{Lin et al., 2011}) are stimulated when exposed to more blue light than red light. However, this is not a viable option to increase leaf production in L. autumnalis seedlings, given the degenerative morphological changes produced by excessive blue light.

The greatest fresh matter weight was obtained with T3 (red and blue 1:3) and T2 (red and blue 2:2), which were superior to the other treatments (Figure 9A). These results are similar to those obtained by ^{Wu and Lin (2012)} who reported that red and blue light (1:1) at 50 µmol m^-2s^-1 intensity increases the accumulation of Protea cynaroides seedling’s fresh weight; but this finding differs from what was reported in Cymbidium, where PLBs’ fresh weight increase is attributed to green light (^{Nahar et al., 2012}). Red and blue light affect stomatal opening, but ^{Squeo and León (2007)} consider that only blue light is responsible for this mechanism, and therefore plants exposed to this light spectrum consume more water, which could help to explain our results (larger fresh weight in T3).

Figure 9 Effect of light treatments T1 (3:1 red and blue LEDs), T2 (2:2 red and blue LEDs), T3 (1:3 red and blue LEDs), T4 (white LEDs) and T5 (white fluorescent light) on fresh (A) and dry (B) weight of in vitro cultured Laelia autumnalis seedlings. Different letters on each figure indicate significant differences (p≤0.05). 

The largest percentage of dry matter (10.83) was registered with T4, larger than T1, T3, and T5 (8.83) (Figure 9B). These results differ from those of ^{Lin et al. (2011)}, who point out that dry weight accumulation in Dendrobium officinale seedlings was favored with 70 µmol m^-2s^-1 of red and blue light (1:2) and is halved when the proportion is even (1.1).

White LED light (T4) and control (T5) increased the accumulation of a chlorophyll (Figure 10A), b chlorophyll (Figure 10B), and total chlorophyll (Figure 11) compared with the other treatments, with up to 16.6 mg g^-1 more total chlorophyll (in T1). This could be the result of the spectral composition of white light (400-700 nm) that includes blue (430480 nm), green (495-570 nm), and red (590-670 nm) wavelengths that absorb photosystems II (≤680 nm) and I (≤700 nm), because maximum efficiency requires both photosystems to be engaged (Solarte et al., 2010). The fact that leaves mainly absorb photons of the blue and red spectrums -and that they absorb a lesser amount of green photonsmust be taken into consideration, although the plants reflect most of those photons as diffuse radiation (Lazo and Ascencio, 2010). Lee et al. (2011) observed a decrease in total chlorophyll content in Paphiopedilum leaves, under blue, red, and red plus blue (9:1) light conditions, which matches the results of our study; while ^{Lin et al. (2011)} found that in Dendrobium officinale the blue spectrum increased more than twofold the concentration of photosynthetically active pigments.

Figure 10 Effect of light treatments T1 (3:1 red and blue LEDs), T2 (2:2 red and blue LEDs), T3 (1:3 red and blue LEDs), T4 (white LEDs), and T5 (white fluorescent light) on the accumulation of a chlorophyll (A) and b chlorophyll (B) of in vitro cultured Laelia autumnalis seedlings. Different letters on each figure indicate significant differences (p≤0.05). 

Figure 11 Effect of light treatments T1 (3:1 red and blue LEDs), T2 (2:2 red and blue LEDs), T3 (1:3 red and blue LEDs), T4 (white LEDs), and T5 (white fluorescent light) on the accumulation of total chlorophyll of in vitro cultured Laelia autumnalis seedlings. Different letters on each figure indicate significant differences (p≤0.05). 

The highest number of roots per seedling (0.90) was found in the fluorescent light treatment (T5) and was similar to LED white light (T4); whereas that combinations of red and blue light (3:1, 2:2, and 1:3) did not induce root formation (Figure 12). This rhizogenesis inhibition could be related to an increase in the blue spectrum radiation in T1 (3:1), T2 (1:1), and T3 (1:3). Results of Chrysanthemun seedlings (^{Kurilčik et al., 2008}) match this tendency, where root formation is inhibited when blue light increases with respect to red and far red light; they, however, differ with observations in Paphiopedilum, where root development was brought about with red and blue light, and where fresh and dry weights reached up to 387.7 and 31.2 mg, respectively, more than with white light (^{Lee et al., 2011}).

Figure 12 Effect of light treatments T1 (3:1 red and blue LEDs), T2 (2:2 red and blue LEDs), T3 (1:3 red and blue LEDs), T4 (white LEDs) and T5 (white fluorescent light) on the number (A) and length (B) of in vitro cultured Laelia autumnalis seedlings’ roots. Different letters on each figure indicate significant differences (p≤0.05). 

Blue light stimulates phenolic compound biosynthesis, and therefore reduces rooting of in vitro sprouts of Protea cynaroides; furthermore, combined with red light, it increases the accumulation of 3,4-dihydroxybenzoic acid, malic acid, and ferulic acid. Conversely, root formation in seedlings lighted with red LEDs is the result of a low endogenous concentration of such phenols (^{Wu and Lin, 2012}). During simultaneous lighting with blue and red lights, a synergic interaction between cryptochromes and phytochromes (blue and red photoreceptors, respectively) takes place, and those pigments could be responsible for the rhizogenesis process perception and activation. Phytochromes could possibly regulate the rhizogenesis process via phytohormone systems (^{Kurilčik et al., 2008}).

Based on our results, it is feasible to light in vitro cultures of O. tigrinum and L. autumnalis with white LED light, because it produces similar developmental effects than those produced by white fluorescent light. Furthermore, LED lights produce little heat and can be placed directly over the plants (^{Schroeter-Zakrzewska and Kleiber, 2014}). This solution allows installing several levels in a shelf, increasing the number of seedlings per area unit, and reducing energy expenses due to lighting and cooling. However, it is still necessary to study the effect of photoperiod and photosynthetic photon flux density, which vary among species, and represent a particular ecologic adaptation (^{Kurilčik et al., 2008}). For instance, strawberry seedling development requires 60 µmol m^-2 s^-1 (^{Nhut et al., 2003}), while chrysanthemum explants only need 40 µmol m^-2 s^-1 (^{Kurilčik et al., 2008}).