Conventional forest genetic improvement

Conventional forest genetic improvement is based on the selection of individuals with outstanding and desirable phenotypic characteristics that are part of a given population when they are in their reproductive stage. For the selection process to be effective, the existence of genetic variation within its range is essential, as it is essential to identify that the trait of interest is heritable and has socioeconomic value. These characteristics must have the potential to increase the productivity of timber or non-timber goods (resins, metabolites, fruits, etc.) obtained from forest resources or to respond favorably to adverse conditions of biotic or abiotic stress and, in this way, keep its evolutionary potential (^{Zobel and Talbert, 1988}).

In line with the above, the data obtained from the selection of individuals from a given population, offer a rigorous description of the phenotypic plasticity. However, this information is insufficient and, as genetic improvement programs progress, it becomes necessary to use assessment methodologies that allow the estimation of genetic parameters (^{Torres, 2000}). Thus, the phenotypic information is structured in databases and is related to its geographical origin to define the provenance of the germplasm. Subsequently, seeds are collected (NMX-AA-169-SCFI-2016) and provenance tests are established, that is, the planting of genotypes with different geographical origin in a common site with homogeneous physiographic and environmental characteristics to evaluate their behavior over time (^{Callaham, 1964}). In addition, forest geneticists carry out progeny tests (the field evaluation of individuals descended from selected trees) in order to predict with statistical tools the genetic values of the parents based on the performance of their descendants (^{White and Hodge, 1989}).

The aim of provenance and progeny trials with respect to the improvement of forest species is to obtain genetic gain. That is, to have a superior performance of the evaluated phenotypic characteristic concerning the average performance in the natural population. This requires the use of outstanding genotypes, the quantification of phenotypic variables, and the estimation of the heritability of the characters (^{Flores et al., 2014}). However, despite the successful cases of different forest plantations in obtaining genetic gain, conventional genetic improvement has limitations; among them, the existence of biases in the estimation of the genotype, since the expression of the phenotype is affected not only by genetic variation but also by the environment and the effects of dominance and epistasis (^{White et al., 2007}). On the other hand, provenance and progeny trials in trees involve a significant investment in economic resources, logistics, and very long selection cycle times before achieving significant results (^{Lynch and Walsh, 1998}). For this reason, it has been necessary to create efficient selection methods in order to reduce time, costs and to estimate heritability and genetic gain more accurately.

Early selection is a type of indirect selection in which multiple traits, that manifest in premature ages, are quantified and correlated with a characteristic of economic interest observed only in mature stages of the individual (^{Vargas-Hernandez and Adams, 1992}; ^{Wu, 1998}; ^{White et al., 2007}). However, this selection technique poses two main challenges: 1) Defining the optimal age for selection of the next generation, and 2) Implementing the use of early assessment data to improve selection. For this, it is necessary to create robust correlation matrices to calculate the selection index with greater fidelity (^{Burdon, 1989}).

Starting in the 1960s, the first germplasm collections in coniferous species in Mexico were established. In 1963, the INIF initiated a program for the establishment of seed areas for Pinus spp., Abies spp. and Pseudotsuga spp.; although it was not until 1985, with the creation of the Centro de Genética Forestal A. C. (Center for Forest Genetics A. C.), that the efforts to improve forest genetics increased significantly. At the same time, the Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias (INIFAP), resulting from the merger in 1985 of the National Institutes for Forest Research (INIF), Agriculture (INIA) and Livestock (INIP), resumed research and actions aimed at this purpose, with the program of genetic improvement of woody species of commercial value. In 1993, INIFAP implemented activities to collect germplasm of red cedar (Cedrela odorata L.), mahogany (Swietenia macrophylla King), sac-chacá (Dendropanax arboreus (L.) Decne. et Planch.), and boxwood (Cordia alliodora Ruiz & Pav.), in order to obtain improved lines of species with high productive potential for central and southern Veracruz State (^{Sánchez et al., 2003}).

In addition, with the financial support of some state governments, producer associations and private companies, forest genetic improvement programs began in several states in the north and center of the country. Unfortunately, after a few years of operation, most of these programs were canceled temporarily or permanently, due to lack of financing and association with the productive sector. This situation prevented efforts from resulting in orchards that could provide genetically improved seed for use in intensive reforestation efforts or commercial plantations (^{FAO, 2012}).

On the other hand, with the creation of the National Forest Commission (Conafor) in 2001, specific actions focused on the production of forest germplasm of known and genetically improved origin were established. Simultaneously, actions initiated to create a National Forest Genetic Resources System that focused on integrating and promoting forest genetic improvement and production (^{FAO, 2012}).

These genetic improvement actions have been concentrated in few species of the following genera: Pinus, Cupressus, Taxus, Eucalyptus, Gmelina, Hevea, Cedrela and Swietenia; based upon the economic value of the production of sawn wood, many of them are already being assessed for their possible commercial and environmental use (^{FAO, 2012}). However, most commercial forest plantations in Mexico are small and do not have genetic improvement programs. Therefore, the results in the progress of the improvement activities show a different degree of success in each of the species involved and with different perspectives of long-term permanence (^{Conafor, 2009}).

Selection assisted by molecular markers

A different way of selecting superior genotypes is through the evaluation of genetic information derived directly from the study of hereditary material contained in deoxyribonucleic acids (DNA) and ribonucleic acids (RNA). Unlike the variation found in phenotypes, DNA variability within or between populations is not affected by environmental conditions, but is the result of evolutionary forces that have influenced the adaptation of the species to its area of distribution (^{Francia et al., 2005}). The development of the polymerase chain reaction (PCR) advanced the study and understanding of the main molecule behind the inheritance of characters (^{Bartlett and Stirling, 2003}). The PCR technique allows the use of DNA markers (nuclear, chloroplastic or mitochondrial) to quantify the genetic diversity in genetic improvement programs in perennial species (^{Porth and El-Kassaby, 2014}).

For example, in the Tecomán Experimental Field of INIFAP, with the collaboration of the Colegio de Postgraduados, SSR (single sequence repeats) markers for the identification of Mexican lemon hybrids have been implemented (^{Carrillo-Medrano et al., 2018}). SSRs are non-randomly distributed regions throughout the genome, characterized by being tandem repeats of nucleotide sets (^{Carneiro et al., 2016}). In their study, ^{Carrillo-Medrano et al. (2018)} were able to successfully identify and differentiate 13 crosses of Mexican lemon with other varieties using four molecular markers. The above, is part of the INIFAP-Tecomán Mexican lemon genetic improvement program that searches for genotypes with greater resistance to diseases.

On the other hand, some researchers from INIFAP's Uruapan-CIR-Pac Experimental Field developed a protocol to describe the genetic diversity of Pinus pseudostrobus Lindl. From the seed area of the hydrological region (RH) Lerma-Santiago (^{Orozco et al., 2010}). This RH is one of the most complex in the country since it has a great diversity of hydrological, geographical, economic, ethnic and historical characteristics, in addition to its abundant population. Therefore, obtaining quality germplasm is essential for management and reforestation plans. The AFLPs (Amplified Fragment Length Polymorphism) obtained by ^{Orozco et al. (2010)} are the first step for a seed quality assessment program since these molecular markers detect multiple loci and allow estimating different parameters of population genetics.

Another interesting example from INIFAP regarding the use of molecular markers in the forestry sector was the genotyping of the rubber tree (Hevea brasiliensis (Willd. Ex A. Juss.) Müll. Arg.), by means of a high-resolution dissociation analysis (HRM) (^{Ortiz et al., 2017}). This automated analysis allows the differentiation, with low cost and great precision, of the existing variation between two or more polymorphic sequences. In the study, carried out by the Cenid Comef Forest Biotechnology Laboratory, three genetic markers allowed the differentiation of ten clonal lines, a research further enriched by the morphological characterization of the clones. The precedent is relevant for the improvement and selection of the rubber crop, because it provides certainty during the production of the plant in the nursery and the establishment of multiplication gardens and commercial plantations. It also facilitates the selection of plants based on their susceptibility or tolerance to pests and diseases, as well as the preservation of their germplasm (^{Ortiz et al., 2017}).

The common denominator of the three previous examples is the use of the PCR technique to identify some genetic characteristics typical of different varieties or populations of individuals. However, these techniques are often complicated when trying to make a more detailed association between genetic and phenotypic variations. This is due to three limitations: 1) The amount of molecular markers is small in relation to the size of the genome, which implies a significant information bias; 2) Much of the variation is located in unknown or non-coding regions, which prevents the understanding of the genes and physiological mechanisms involved in the phenotype; and 3) In doing studies that involve many populations and a large number of individuals it becomes expensive and difficult to operate and analyze. To counteract the above, in recent years the use of next-generation massive sequencing technologies and bioinformatics tools has been promoted to accelerate the domestication process of trees (^{Harfouche et al., 2012}; ^{Isik, 2014}).

An example of the above is the study of the Mexican lemon variety of INIFAP in response to the infection of Candidatus Liberibacter asiaticus, which causes the Yellow Dragon disease (also called Huanglongbing or HLB) that is responsible for huge economic losses in citrus plantations worldwide. By means of messenger RNA sequencing (RNA-seq), 41 522 transcripts were obtained. This data set represents the first approach to the use of the RNA-seq technique for the Mexican lemon and reveals valuable information for future breeding programs, and will be an essential tool to facilitate the design of new molecular markers to diagnose and study the HLB disease (^{Arce-Leal et al., 2020}).

Vaccinium macrocarpon Aiton (large cranberry) is a small perennial shrub that belongs to the Ericaceae family. It is an example of a non-timber forest resource that, together with other berries, has become one of the most important agricultural/forest products in the country due to its exportation value. ^{Diaz-Garcia and collaborators (2020)} from the Pabellón Experimental Field of INIFAP performed a genotyping by sequencing analysis to identify the domestication status of this species. This research used 121 native (wild) and 111 selected commercial accessions. As a result, more than 21 000 single nucleotide polymorphisms (SNPs) were detected throughout its ten chromosomes. Analysis of the genetic data revealed a clear directional and gradual transition from wild to cultivated forms. However, the authors also mentioned that the recurrent introgression of wild germplasm into commercial cultivars and the reduced number of breeding cycles has slowed down the genetic differentiation between wild and cultivated large cranberry materials. This shows that cutting-edge innovations are required to accelerate the genetic improvement process in species with long life cycles (^{Diaz-Garcia et al., 2020}).

Vegetative propagation

As mentioned above, domestication of trees and other long-lived perennials (including bananas and palms) occurred at a later period than most cultivated annual species. Few trees have undergone this process for more than 4 000 years and, in most cases, only during the last 2 000 years (^{Meyer et al., 2012}). The origin of tree cultivation is closely related to the development and management of vegetative propagation techniques, in particular cuttings and grafts (^{Meyer et al., 2012}). This type of plant reproduction makes it possible to multiply, in a practical way, a genotype/phenotype of particular interest and is an essential part of forestry practices (^{Libby and Rauter, 1984}). INIFAP has contributed to the development of asexual propagation, characterization and establishment of clonal orchards of tropical and fruit trees. Researchers from the forest area in the states of Veracruz and Quintana Roo have achieved the standardization of propagation protocols by grafting of red cedar (Cedrela odorata) and mahogany (Swietenia macrophylla) for the establishment of clonal banks and asexual seed orchards. This has been carried out at the El Palmar Experimental Field, Veracruz and at the San Felipe Bacalar Experimental Site, Quintana Roo ((^{Sánchez et al., 2003}; ^{Curti-Díaz et al., 2012}; ^{Salazar-García et al., 2016}; ^{Ortiz et al., 2017}; ^{Avendaño-Arrazate and Cueto-Moreno, 2018}; ^{Sampayo-Maldonado et al., 2019}; ^{Cob et al., 2020}).

On the other hand, in vitro culture also allows the multiplication of selected material. This technique, founded on the totipotency of the plant cell, allows the regeneration of a complete individual from a fragment of the plant. In the Forest Biotechnology Laboratory of Cenid Comef of INIFAP, has developed several in vitro culture protocols for forest species. An example of the above is the propagation for mahogany (S. macrophylla) through in vitro germination, proliferation of axillary buds and callus, that can be used for the regeneration of complete plants, as well as for the development of research on biochemical routes and the production of natural metabolites with defense potential against pests (^{Wegier et al., 2013}).

Furthermore, the Mexican avocado (Persea americana Mill. var. drymifolia) which belongs to INIFAP germplasm bank, has been successfully reproduced in Murashige y Skoog (MS) basal medium. The developed protocol generated acclimatized plants under greenhouse conditions. This multiplication under laboratory conditions allows the production of plants resistant to an adverse environment such as drought, flood, salinity or the absence of pathogens (^{Cortés-Rodríguez et al., 2011}).

Micropropagation can also be useful to induce genetic variability that does not exist in nature. The use of ionizing radiation such as X-rays, gamma rays and neutrons, or the use of chemical mutagens, produce changes in the genetic material. When this technique is applied to in vitro grown vegetative material subjected to a certain selection pressure (for example salinity or drought), new variants with improved characteristics can be obtained (^{Ahloowalia and Maluszynski, 2001}). In this way, in collaboration with the National Institute for Nuclear Research, INIFAP forest biotechnology researchers have generated a proof of concept to obtain mutants of Psudotsuga menziesii (Mirb.) Franco through irradiation with different doses of gamma rays. This work represents the first effort of INIFAP for the artificial induction of genetic variability in a conifer (^{Castillo-Martínez et al., 2018}).

Genetic engineering

The genetic improvement of tree species by means of recombinant DNA technology is still incipient. The combination of modern biotechnologies such as DNA synthesis, vertical transfer of genetic material, CRISPR-based editing and next-generation high-throughput sequencing, increasingly allow precise manipulation of genes and genomes (^{Simon et al., 2019}). In recent years, the development of efficient protocols for the precise editing of tree genomes has begun and is expected to be a powerful tool to accelerate the genetic improvement process in species with a long life cycle (^{Van Zeijl et al., 2018}; ^{Fan et al., 2020}). Although in Mexico there are no published studies involving the genetic manipulation of forest species, there is an enormous potential of both natural and human resources that can sustain scientific exploration and should not be neglected as the basis for its own homegrown technologies (^{Vallejo et al., 2019}). However, research carried out in this regard must be founded on human and environmental biosafety guidelines, local and national needs, and the characteristics of the species to be studied, so that they can be accepted by society and be successful in solving a very particular problem in the Mexican forest sector.