The use of vegetative propagation in forestry is the fastest, most flexible and effective way to produce enough genetically improved material to meet future wood demands (Lelu-Walter et al., 2013). Somatic embryogenesis (SE) is currently considered one of the successful biotechnological techniques for mass propagation, enabling multi-varietal forestry (Park et al., 2016) and for cryopreservation of embryonal masses (EMs) from elite genotypes (Corredoira et al., 2006).

In the last years, several improvements in different aspects of Pinus spp. SE process, such as initiation (Montalbán et al., 2012), maturation (Montalbán et al., 2010) and germination (Montalbán and Moncaleán, 2018) have been carried out in our laboratory. Traditionally, the adjustments have been focused on modifications of basal medium components (Salajova and Salaj, 2005; Zhang et al., 2007), exposure to different types and concentrations of exogenous plant growth regulators (Choudhury et al., 2008) and the use of different initial explants (Hargreaves et al., 2017).

One of the main bottlenecks of the SE in conifers is the progression from immature embryogenic cultures into mature cotyledonary embryos able to develop well-growing plants (Harvengt, 2005). Several external stimuli, such as plant growth regulators (Ayil-Gutiérrez et al., 2013), osmotic agents (Kikuchi et al., 2006), nutritional components (Pěnčík et al., 2015), amongst others, have been recognized as essential factors in determining both the hormone biosynthesis and the developmental fate of explant cells (Grzyb et al., 2017). In conifers, maturation is stimulated by reduction of water availability by means of raising the osmoticum of the medium (Montalbán et al., 2010) or the concentration of gelling agent (Klimaszewska and Smith, 1997). Modifying gellan gum concentration in maturation media has been studied in a few species of angiosperms (Márquez-Martín et al., 2011) and gymnosperms (Morel et al., 2014); these authors reported an improvement in the development and maturation of somatic embryos (Se) of P. pinaster by increasing gellan gum concentration, pointing out that the effect of gellan gum concentration on plant response is provoked by changes of water availability in the culture medium. In our laboratory, we demonstrated in previous studies that the agar concentration and the temperature had a significant effect on initiation step of radiata pine SE; and in the case of temperature, in the subsequent phases of the process (García-Mendiguren et al., 2016). As far as we know, the effect of different temperatures in maturation step has not been tested in Pinus spp., only in some angiosperms (Wang et al., 2014).

The active cytokinin (CK) pool is regulated during development by biosynthesis, uptake from extracellular sources, metabolic interconversions, inactivation, degradation, and transport (Kakimoto, 2003). The relative abundance of different CKs can vary greatly between plant species, tissues and developmental stages, and depends on the environmental conditions (Frébort et al., 2011). Some studies have examined the effect of varying the culture atmosphere on endogenous CK levels after organogenesis in angiosperms (Jiménez and Bangerth, 2001; Moncaleán et al., 2003) and gymnosperms (Moncaleán et al., 2005; Montalbán et al., 2012). On the other hand, the role of indole-3- acetic acid (IAA) and CKs in the SE process have been analyzed in cotton (Zeng et al., 2007; Jin et al., 2014) and wheat (Hess and Carman, 1998; Jiménez and Bangerth, 2001) and fern (Grzyb et al., 2017). However, up to know a deep analysis of endogenous cytokinins and IAA concentration in SE process in pines has not been determined. In this sense, and as it has been described previously, IAA and other phytohormones have an important role in embryo differentiation events (Silveira et al., 2004; Vondrakova et al., 2018); thus, a better understanding of the phytohormone profile in somatic embryos under different environmental conditions could enable us to manipulate this stage of SE to increase its efficiency. Taking into account all the above mentioned studies, the main objective of this study was the evaluation of how the environment of cultures (gellan gum concentration and temperature) along maturation stage affects the success of radiata pine SE processes as well as to analyze its effect in CK and IAA endogenous levels.

The inductive conditions which allow differentiated somatic cells to develop into competent dedifferentiated cells can be achieved by the addition of specific phytohormones, and/or by exposing the tissue to different stress factors (Leljak-Levanič et al., 2016). Moreover, water relations between the embryo and its environment play a regulatory role in embryo development, particularly during maturation (from Márquez-Martín et al., 2011) and for this reason, the water availability of maturation media was determined in our experiments. Radiata pine somatic embryos cultured in EDM medium supplemented with 10 gL^-1 and stored at 28°C grew in an environment with the lowest water availability in comparison to the other conditions assayed. To this respect, in Douglas-fir, a reduction in water availability of the culture media provoked an increase in Se quality (Lelu-Walter et al., 2018) probably because it has been described that the reduction in water availability leads to obtain embryos with low water content similar to their zygotic counterparts (Morel et al., 2014).

The culture of embryogenic calli in media with an increased gelling agent concentration is a common procedure for the production of high quality mature embryos (Krajňáková et al., 2009; Troch et al., 2009; Buendía-González et al., 2012). Our results in P. radiata are in agreement with those obtained in P. strobus where the best maturation results were obtained increasing gellan gum concentration due to a shift in the developmental program of the culture, from proliferation of cells and early somatic embryos to the production of embryos at advanced stages (Klimaszewska et al., 2000).

Respect to the number of embryos developed, the temperature and gellan gum concentration affected significantly the number of Se per gram of EM. In this sense the temperature (28°C) and the agar concentration (10 gL^-1) that provoked the lowest water availability led to the lowest and the highest number of Se, respectively. For this reason, it seems that water availability was not related to the success of the process at this stage. In this sense, there are two studies suggesting that under sub-optimal water availability, plants may take to a premature drought stress response and arrest or drastically reduce growth due to the effect on source–sink relationships through preferentially inhibiting nutrient acquisition processes more and earlier than necessary (Skirycz et al., 2011). It is clear that the water availability registered in our experiments is not sub-optimal for the maturation of EMs. In our laboratory, the effect of different physical and chemical conditions at the initiation stage of P. radiata and P. halepensis (García-Mendiguren et al., 2016; Pereira et al., 2016) as well as in proliferation stage of P. halepensis (Pereira et al., 2017) was studied and we found that the highest temperature seemed to produce a selective pressure as pointed by Fehér (2015); In this regard, Bonga et al. (2010) suggested that reducing or increasing temperatures may improve initiation and proliferation since temperature stress may promote cellular reprogramming. Our results are in agreement with de Almeida et al. (2014) where temperature had a significant influence on the direct SE capacity in coffee.

Regarding gellan gum concentration, some reports suggested that increasing the concentration of a gelling agent from 4 to 8 gL^-1 (Teyssier et al., 2011) or from 4 to 9 gL^-1 (Morel et al., 2014) promoted the maturation of the Se of some conifers. In this sense, Garin et al. (2000) demonstrated that the phytagel concentration was critical for the maturation of Se of five tested ECLs of Pinus strobus. The increase of phytagel concentration in the maturation media allowed obtaining well-developed Se capable of germination in various species of pines (Klimaszewska et al., 2009) and hybrid larch (Lelu-Walter and Pâques, 2009) probably due to the action in endogenous ABA modulation in mature somatic embryos (Morel et al., 2014). In summary, our results agree with abovementioned studies pointing that increasing gelling agent concentration is a good strategy to improve the Se yield in radiata pine.

The highest germination rates (above 90%) were obtained in Se coming from EMs maturated at 18 and 23°C, whereas maturation at 28°C led to lower germination rates, particularly when it was combined with 9 or 10 gL^-1 gellan gum. Our results are not in agreement to Prewein et al. (2004) when they described that desiccation of the embryos due to decreased water availability in maturation media improves the germination frequency, probably by reducing the endogenous ABA level or by changing the sensitivity to ABA (Jiménez, 2005). Unlike observed by Morel et al. (2014) in maritime pine, we found no differences in germination rates with increasing concentration of agar, however, this could be due to differences in the range of concentrations used by these authors (4 and 9 gL^-1 Phytagel) and us (8, 9, and 10 gL^-1 Gelrite).

An improvement in plant conversion when the gellan gum concentration was increased at all temperatures tested was observed. However, 23°C showed the best results improving by 132% the plant conversion process. Summarizing, increasing the gellan gum concentration in the maturation stage improved the SE process in all radiata pine cultures independently of the temperature of storage obtaining the best results when somatic embryos were submitted to maturation process at control temperature (23°C). So, higher temperature and gellan gum concentration applied at maturation stage did not result in an increment of the efficiency of SE process like it was observed when this stressful condition was applied at initiation stage (García-Mendiguren et al., 2016). Moreover, ex vitro survival did not show significant differences although the highest percentages were obtained in somatic plants coming from EMs maturated at 28°C.

There are only a few reports describing how the contents of different endogenous hormones are altered by different environmental conditions. The relative abundance of different CKs can vary greatly between plant species, tissues and developmental stages, and depends on the environmental conditions (Frébort et al., 2011). In our experiments, the prevailing isoforms of CKs were cis-isomers of Z and ZR according with Zur et al. (2015) in microspore embryogenesis in triticale. Based on studies with Arabidopsis, cis-isomers were regarded as CK derivatives without any or with low biological activity. Further studies with non-model plants, like pines, showed that cis-isomers can be the dominant form of CK in specific plant organs and/or stages of development (Emery et al., 1998: Vyroubalová et al., 2009; Kudo et al., 2012). To this respect, Gajdošová et al. (2011) proposed that cZ can be qualified as a regulator of CK responses in plants under growth-limiting conditions. Our results only showed significantly differences in cZ in Se cultured with 9 gL^-1 of gelrite at the highest temperature (28°C) with the others analyzed.

Highest amount of bases, the active forms of cytokinins (Klemš et al., 2011), were found in Se developed under 18 and 23°C and low gellan gum concentration, treatments showing the highest water availability. So, our results are in agreement to other authors describing that a decrease in water availability could be related to the rapidly utilization of CK bases and tightly regulated by mechanisms that prevent their accumulation at high levels (Moncaleán et al., 2005; Stirk et al., 2005).

CK ribosides, mainly represented by cZR and iPR, were higher in the treatment that led to the highest somatic embryo production and success rate (23°C and 10 gL^-1). In terms of metabolic groups, ribosides were the predominant group of cytokinins and its percentage respect to bases and O-glucosides increased in parallel with the decrease of water availability. Nevertheless, in peach seeds, phosphates and ribosides of the DHZ and iP-type are the predominant CKs and functions like embryo formation and growth, division of endosperm primary nucleus, endosperm cellularization and endosperm cell division have been attributed to them (Arnau et al., 1999). So, opposite to Pacheco de Freitas et al. (2016) active CK levels seem to be important during the later stages of embryo development and maturation showing large amounts of these phytohormones in treatments with high production of mature embryos.

CK O-glucosides (reversible forms of storage), particularly cZOG, were highest in those embryos coming from EMs maturated in the most stressful conditions (28°C and 10 gL^-1 gelrite, lowest water availability). However, CK O-glucosides were found in minute amounts (below 4 pmolg^-1 DW), for instance Montalbán et al. (2011) reported O-glucoside concentrations above 240 pmolg^-1 DW in P. radiata vegetative buds.

Low content of IAA and iP were obtained in Se treated with the highest gelrite concentration at the standard temperature (23°C). In this sense, in cotton a high content of both groups of phytohormones were related to the redifferentiation stage that leads to embryogenesis induction (Zeng et al., 2007). Somleva et al. (1995) and Sáenz et al. (2003) have found an inverse relationship between cytokinins and the embryogenic response; in this sense, the treatment that provoked the highest success in the SE process (23°C, 10 gL^-1 gelrite) showed a low level of IAA/bases+ribosides CKs relation. So it seems that IAA can play an important role in radiata pine SE maturation process; this conclusion is in agreement to the postulation that auxin metabolism (biosynthesis, conjugation, degradation), intercellular transport, and signaling is crucial in coordinating the morphogenesis and development of plant reproductive organs, including the formation of embryos (Benková et al., 2003; Bernardi et al., 2012). The accumulation of endogenous IAA in response to abiotic stress may have a mediating role during SE (Pěnčík et al., 2015; Nic-Can et al., 2016).

It seems that the differences found in specific CKs analyzed did not provoke changes in maturation rates. Our results suggest that none of phytohormones found acts alone in the acquisition of embryogenic maturation capacity; it seems that the dynamic and complementary actions of the auxin and cytokinin pathways regulate several developmental processes and their ability to crosstalk makes them ideal candidates for mediating stress-adaptation responses according with Bielach et al. (2017) as well as initiating various signal transduction pathways (Kohli et al., 2013). In this regard, comparing the endogenous levels of IAA, Z, ZR, iP, and iPR in several embryogenic habituated callus lines of Asparagus officinalis, no correlation between embryogenic potential and endogenous hormone levels could be found (Limanton-Grevet et al., 2000). These and other findings demonstrate that there is no direct correlation in endogenous hormone content and embryogenic competence of different genotypes, or between competent and non-competent genotypes, which suggests the involvement of additional, hormone-related internal mechanisms for embryogenic competence regulation and regeneration ability in vitro (Leljak-Levanič et al., 2016). However, from our results, it seems clear that low levels in the relation between IAA and CK bases+ribosides are related to highest success of SE process in radiata pine.

Further studies increasing the temperatures during shorter periods of time can help us to go further in the understanding of the mechanisms involved in SE in Pinus radiata as well as the role of phytohormones in this process.

PM and IM conceived and planned the experiments. PM, OG-M, and IM carried out maturation experiments. OG-M, ON, and MS carried out phytohormone analyses. IM, MU, and TG carried out the statistical analyses. All authors provided critical feedback and helped to shape the research, analyses, and manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.