Introduction

The Pseudotsuga genus (Pinaceae) known as fir or Douglas fir is distributed in Western North America and Southeast Asia (^{Farjon, 1990}). Pseudotsuga menziesii (Mirb.) Franco develops in varied environments and its geographical distribution is wide, forming extensive natural forests in the Western United States of America and Canada (^{Hermann and Lavender, 1999}).

In Mexico, the natural distribution of P. menziesii is geographically extensive, although it is found in small populations or isolated patches along the Sierra Madre Occidental and in the Northern part of the Sierra Madre Oriental and the Sierra de Querétaro (^{Rzedowski, 1978}). In addition, it forms small stands along the Neovolcanic Axis in the center of the country (^{Villagómez and Bello, 2015}), in Tlaxcala and Puebla (^{Domínguez et al., 2004}). Also, two populations are cited in the state of Oaxaca: one in Santa Catarina Ixtepeji (^{Debreczy and Rácz, 1995}) and another in the Sierra Madre del Sur, both representing the Southern natural limit of the species (^{Del Castillo et al., 2004}).

In the national territory, P. menziesii is highly appreciated for its potential as a Christmas tree. Therefore, germplasm is required to establish commercial forest plantations (CFP) and at the same time to conserve the species in its natural populations.

Currently, planters demand a large quantity of seeds for Christmas tree CFPs, mainly in the central area of the country (Puebla, State of Mexico and Tlaxcala). Some alternatives to solve this supply problem consist of having a better knowledge of the factors that regulate flowering and seed production in trees. To this end, there are cultivation practices and treatments that have been used successfully to accelerate the production of cones and seeds in juvenile trees, based on practices such as trunk girdling and the application of growth regulators, root pruning, stress temperature and fertilization (^{Ross and Bower, 1989}; ^{Cherry et al., 2007}; ^{Muñoz-Gutiérrez et al., 2010}).

Floral induction through the addition of growth regulators that accelerate cone and seed production in a wide range of conifers is well known, particularly with gibberellic acid (GA). The combination of GA₄ and GA₇ (GA_4/7) is the most effective and is used in the induction of reproductive structures in different conifer species (^{Ross, 1983}; ^{Greenwood, 1987}; ^{Owens et al., 2001}; ^{Aderkas et al., 2004}; ^{Cherry et al., 2007}). However, these treatments, in addition to promoting the formation of reproductive structures, may modify the periods of pollen receptivity and dispersal (^{Vargas-Hernández and Vargas-Abonce, 2016}), a modification that can reduce or eliminate the contribution of some individuals in the seed production (^{Muñoz-Gutiérrez et al., 2010}). The above is due to the lack of phenological synchrony between individuals from the same plantation, which increases the probability of inbreeding, and therefore the number of wasted seeds (^{Burczyk and Chalupka, 1997}).

Although there are already precedents in Mexico on floral induction work in P. menziesii (^{Muñoz-Gutiérrez et al., 2010}), there are no studies on its reproductive phenology in natural populations or in Christmas tree plantations with young trees treated to stimulate early flowering. Therefore, the general objective of the present work was to determine the periods of receptivity of female strobili and pollen dispersal in a plantation of young P. menziesii trees that were treated with gibberellic acid (GA_4/7) to induce the formation of reproductive structures. Two groups of trees with different levels of response to the formation of female and male reproductive structures were used; it was proposed to analyze, in particular, the differences between groups of trees in terms of the beginning, end and duration of the receptivity of the female strobili and pollen dispersal, in addition to their relationship with the accumulation of degree-days, and determine the degree of phenological overlap between individuals in the same plantation. The hypothesis raised is that the synchrony between female receptivity and pollen dispersal is different between two groups of young P. menziesii trees due to the differential response to the formation of male and female reproductive structures.

Materials and Methods

Phenology of reproductive structures

Monitoring was carried out on 12 trees with the formation of female and male reproductive structures during the months of February and March 2022, regardless of the type of treatment. For each tree, three branches located between the middle and upper part of the crown were marked, since it is the section with the greatest presence of female reproductive structures, and three branches for monitoring the male structures, which were located between the middle and lower part of the crown, where the male strobili mainly develop.

Observations of phenological progress were made with a frequency of four to five days from the moment the reproductive structures were visualized in the buds, until the female structures were no longer receptive, and when pollen dispersal in the male strobili ended.

To determine the phenological stages, the methodology described by ^{Webber and Painter (1996)} was followed. For the female strobili, five stages were defined: (E1) The female reproductive structures are distinguished, not yet emerging, in the bud, (E2) Elongation of the strobile, burst bud and the bracts of the reproductive structure are visible, (E3) 30 to 40 % of the bracts are exposed, the female structure is already receptive, (E4) Maximum receptivity, and (E5) Formation and elongation of the female strobile (Figure 1).

Figure 1 Phenological stages of female strobili of Pseudotsuga menziesii (Mirb.) Franco. 

For male strobili, five stages were considered too: (E1) Male reproductive structures are distinguished, not yet emerging in the buds, (E2) The buds swell and the pollen sacs are visible, (E3) Elongation of the strobile, breaks the scales of the bud, (E4) The strobile reaches maximum size and maturation, a phase in which the pollen is dispersed, and (E5) Dispersion ends and the strobile begins to dry (Figure 2).

Figure 2 Phenological stages of male strobili of Pseudotsuga menziesii (Mirb.) Franco. 

From the phenological data of each sampled branch, the beginning and end date in Julian days (number of days from January 1^st) was determined, as well as the duration of the period of receptivity of the female structures and pollen dispersal.

Results

Phenological overlap

The receptivity of the female strobili began on day 52 (February 21^st) in 20 % of the trees; by day 61, 80 % of the individuals were at their maximum receptivity (E4) for a period of approximately one week, at the same time that 50 % of the trees were dispersing pollen (Figure 3). It was observed that the periods of maximum female receptivity and maximum pollen dispersal coincided phenologically for a six-day period, approximately.

Figure 3 Phenological overlap of female receptivity and pollen dispersal events. 

Discussion

There were no differences between groups of trees, regarding the periods of pollen receptivity and dispersal, but the latter was of shorter duration. Possibly, the stress conditions caused by the girdling of the trunk affected the periods of differentiation of the floral buds, initiation and subsequent development of the male strobili (^{Webber, 2004}).

According to the results, female receptivity begins before pollen dispersal in the same tree, as in Pinus patula Schltdl. & Cham. (^{Hernández et al., 2016}), P. radiata D. Don. (^{Griffin, 1984}; ^{Kang et al., 2001}) and P. sylvestris L. (^{Burczyk and Chalupka, 1997}; ^{Sarvas, 1962}). This behavior is known as “metandry” (^{Pulkkinen, 1994}), which is common in species of the Pinus genus. The average duration of receptivity of the female strobili of Pseudotsuga menzeisii in the two groups was 14 days, less than that recorded for the same species by ^{El-Kassaby (1989)} in sexually mature and productive trees present in a seed orchard, who documented a duration of 24.3 days in natural conditions and 17.3 days in trees that were under cooling treatment. However, it is shorter than the duration indicated by ^{Owens et al. (1981)}, whose values for female receptivity varied between four and six days.

For pollen dispersal, in the same study (^{El-Kassaby, 1989}) it was documented that under natural conditions and at the population level, this phase lasted 27.7 days to 18.3 days in trees under cooling treatment. The few existing records indicate that the optimal duration of the pollen dispersal period is three to five days (^{Barner and Christiansen, 1962}; ^{El-Kassaby, 1989}).

In species such as Pinus greggii Engelm. ex Parl. (^{Ruiz-Farfán et al., 2015}) and P. nigra J. F. Arnold (^{Alizoti et al., 2010}), a phase shift was recorded in the maximum peaks of both phenological events, but the dispersion periods were shorter than those of receptivity. The duration of pollen dispersal can be shortened by excess humidity, due to continuous periods or short events of abundant rain, which causes compaction of the pollen within the strobile and it quickly loses viability. The opposite occurs during droughts when pollen dispersal is faster (^{Ruiz-Farfán et al., 2015}).

The duration of the pollen dispersal period in Pseudotsuga menziesii (^{Erickson and Adams, 1989}) and in Pinus radiata (^{Codesido et al., 2005}) depends on humidity, temperature and DD accumulated during the differentiation process and later development of the reproductive structures (^{Huusko and Hicks, 2009}). For P. menziesii, between 800 and 900 DD above 10 °C are required to initiate dispersal at an altitude of 305 m, while at altitudes between 762 and 1 067 m, 400 DD are needed (^{Ebell and Schmidt, 1964}). However, in this study 348 DD were accumulated. Possibly, the lower number of degree-days required in the plantation that is located at a higher altitude (2 640 m) suggests an adaptive value of the species to initiate the flowering processes.

On the other hand, not all the trees in Group 1 developed both types of reproductive structures in the same tree; all of them presented female strobili, but only in 50 % were male strobili observed. Therefore, the participation of not all individuals in the formation of male strobili and pollen dispersal is a limitation for the production of viable seeds, since only few trees act as pollinators or pollen dispersers. On the contrary, in Group 2 in which the trees developed female and male strobili on the same tree, they had the possibility of crossing each other, with slightly higher values regarding the average synchronization index (PS=0.64).

The average PS between female receptivity and pollen dispersal in both groups was 0.60 and 0.64, which is interpreted as moderate to high synchronization. This is possibly explained by the little variability in the periods of pollen receptivity and dispersal, however, the pollen from the trees in Group 2 had a greater probability of pollinating all individuals. These values are higher than those recorded in other conifers such as P. patula and P. greggii in 2012, which were 0.48 and 0.57, respectively (^{Ruiz-Farfán et al., 2015}; ^{Hernández et al., 2016}).

The correlations between the periods of female receptivity and pollen dispersal (0.79≤r≤0.87) indicate that the phenological events are not independent, since there are probably effects associated with the treatments to promote early flowering that affect its subsequent development; however, in the same plantation the formation of reproductive structures was observed in untreated trees of the same age. Lower values were recorded in Pseudotsuga menziesii (^{El-Kassaby et al., 1984}), Pinus patula (^{Muñoz-Gutiérrez et al., 2019}) and P. taeda L. (^{Askew, 1988}) with positive correlations between 0.45 to 0.67. The fact that phenological events are not independent favors self-fertilization, which negatively affects seed production (^{Muñoz-Gutiérrez et al., 2019}).

Variation in strobile production and synchronization has been observed in species of the Pinus genus and other conifers regularly at early ages, and in first generation seed orchards (^{El-Kassaby et al., 1988}; ^{Matziris, 1994}). In these cases, the use of growth regulators and controlled crossings are alternatives to maximize the participation of the individuals of interest.

Conclusions

There are no broad differences in reproductive phenology between groups of trees, particularly, the quantity and maturation of male strobili are not equal. This fact gives the opportunity for a group of trees to pollinate several trees with varied receptivity, since they are moderately synchronized, despite the fact that the female receptivity stage (E3) begins earlier and lasts longer than pollen dispersal (E4).

The PS and the correlation values between the phenological events of the development of both reproductive structures in the studied groups are high, if the origin of the plantation material (seed) is considered.

The use of growth regulators and cultural practices to promote early flowering in young trees are useful techniques to reduce seed production times, and do favor the phenological overlap of reproductive structures between trees. However, trees within the same plantation, not included in the present investigation, were detected with a certain precocity to the formation of female and male reproductive structures, similar to the evaluated groups; therefore, complementary management activities, such as artificial pollination, are operationally required.