Introduction

Avocado production in Mexico is based on the use of vigorous rootstocks that generate trees that make agronomic management difficult due to their size. For this reason, several research projects are focused on finding dwarf cultivars that are productive and have good fruit quality, in order to establish high-density plantations and reduce production costs (^{Sánchez-Colín, Rubí-Arriaga, & de la Cruz-Torres, 1992}).

Currently, in breeding programs, the physiological efficiency of new avocado cultivars is important, which can be related to leaf efficiency, for example, water-use efficiency (^{Acosta-Rangel et al., 2018}), when considering that it responds to soil conditions such as root hypoxia (^{Gil, Gurovich, Schaffer, García, & Iturriaga, 2009}), and salinity (^{Acosta-Rangel et al., 2019}), among other factors. In addition, unlike roots that require digging or the use of rhizotrons to obtain and study them, direct access is available with leaves.

The primary processes that determine plant growth are, fundamentally, those involving gas exchange with the surrounding air, such as photosynthesis, respiration and transpiration (^{Taiz & Zeiger, 2006}), and they occur mainly in leaves. Within the epidermal tissue of the leaf are the stomata, formed by an ostiole or pore surrounded by two kidney-shaped occlusive cells (called guard cells), and may be accompanied by cells called subsidiary or support cells (^{Prabhakar, 2004}).

The presence of stomata on the upper and lower sides of the leaf blade are distinctive characteristics among species, which are classified as amphistomatic (which have stomata on both surfaces), hypostomatic (with stomata on the abaxial or lower surface) and epistomatic (with stomata on the adaxial or upper surface) (^{Fricker & Willmer, 2012}). The distribution of stomata in avocado leaves has been classified as hypostomatic (^{Kofidis, Bosabalidis, & Chartzoulakis, 2004}).

Gas exchange is mainly carried out by stomata, where gas conductance is controlled by changes in the turgor of the guard cells and responds to various environmental factors, including light, humidity, and CO₂ concentration (^{Assmann & Gershenson, 1991}). The relationship between CO₂ assimilation and stomatal conductance varies among species, and may be absent, resulting in internal CO₂ accumulation (^{Gil et al., 2009}).

The number of stomata in plant species is highly variable. ^{Fricker and Willmer (2012)} observed a variation from 14 to 2,200 stomata·mm^-2 and found that the frequency is affected by a number of factors. Some researchers found, in various citrus and Annona species, that the number of stomata on the leaf varies with respect to the rootstock used (^{Schoch, Zinsou, & Sibi, 1980}; ^{Cañizares, Sanabria, Rodríguez, & Perozo, 2003}; ^{Parés-Martínez, Arizaleta, Sanabria, & Brito, 2004}), and in avocado it is modified by intergrafting (^{Ayala-Arreola, Barrientos-Priego, Colinas-León, Sahagún-Castellanos, & Reyes-Alemán, 2010}). It has been reported that avocado has from 100 to 610 stomata·mm^-2 in seed-derived plants (^{Kadman, 1965}), and that it can vary with the seedling profile, avocado race (^{Barrientos-Priego, Borys, Trejo, & López-López, 2003}) and genotype (^{Barrientos-Priego & Sánchez-Colín, 1987}).

In avocado, ‘Colín V-33’, which has low vigor and a sprawling growth habit (^{Roe, Conradie, & Köhne, 1995}), has been used as an intermediate graft to reduce the height of trees of the ‘Fuerte’ (^{Barrientos-Priego, López-Jiménez, & Sánchez-Colín, 1987}) and ‘Hass’ varieties (^{Barrientos-Villaseñor, Barrientos-Priego, Rodríguez-Pérez, Peña-Lomelí, & Muñoz-Pérez, 1999}). Likewise, it has been used in the selection of dwarfing rootstocks to meet the needs of new production systems, where stomatal density has been proposed as a possible pre-selection index (^{Barrientos-Pérez & Sánchez-Colin, 1982}), and it has been found that the higher the density of stomata in the leaves, the smaller the tree size (^{Barrientos-Priego & Sánchez-Colín, 1987}).

The use of intergrafting can modify the stomatal density of ‘Hass’, ‘Fuerte’ and ‘Duke 7’ avocado leaves, and according to ^{Ayala-Arreola et al. (2010)}, the intergrafting of ‘Colín V-33’ significantly increased the stomatal density of the final graft. Higher stomatal density has been related to smaller stomatal size, which can react quickly to unfavorable conditions (^{Drake, Froend, & Franks, 2013}).

Because stomatal density is a characteristic that requires time to be defined, pre-selection of genotypes at early stages could be done by assessing porometer-measured stomatal conductance, as these characteristics are related (^{Taylor et al., 2012}). Therefore, the aim of this research was to evaluate the relationships between stomatal and gas exchange characteristics of the leaf blade in avocado seedlings derived from ‘Colín V-33’ seed.

Materials and methods

Results and discussion

According to the observed stomata distribution, avocado leaves were classified as hypostomatic, which agrees with ^{Kofidis et al. (2004)}. This characteristic is common in fruit and forest trees, as it is a protection mechanism for the plant's stomata to avoid excess transpiration. According to ^{Camargo and Marenco (2011)}, the hypostomatic characteristic is considered a primitive trait in plants, while ^{Rendón-Anaya et al. (2019)} considered avocado a basal lineage species, close to the origin of flowering plants.

The distribution of stomata on the abaxial surface had no specific arrangement, as they were observed scattered throughout the underside of the leaf (Figure 1), a trait that is also considered primitive (^{Camargo & Marenco, 2011}) and, in the case of avocado, they have been found distributed on the abaxial surface, with no differences between the apical, middle and basal part of the leaf (^{Morales-Escobar, Barrientos-Priego, Barrientos-Pérez, & Martínez-Damián, 1992}). According to what was obtained in the t-test, there is no difference between the two sides of the leaf in the abaxial part (P = 0.615).

Figure 1 Frontal view of the abaxial surface of the leaf blade of a Persea americana Mill. ‘Colín V-33’ seedling: a) distribution of stomata (e) and typical epidermal cells (ce), and b) detail of an anomocytic stoma. 

The arrangement of epidermal cells around the guard cells (Figure 1) allowed classifying the stomata as anomocytic, whose main characteristics are that they do not have adjoining or accompanying cells and that they have four or more subsidiary cells that vary in position, shape and size (^{Prabhakar, 2004}). That is, the cells surrounding the stoma do not differ from other epidermal cells.

On average, stomata length ranged from 13.11 to 19.64 µm, being 16.35 µm on the right side of the underside of the leaf blade and 16.06 µm on the left side (Table 1), with no statistical differences between the two sides of the leaf blade (P = 0.757). ^{Drake et al. (2013)} state that under unfavorable conditions small stomata have a faster response than large stomata, and combined with high stomatal density show high conductance. Thus, these characteristics can be considered in a comprehensive selection for high stomatal conductance, probably accompanied by a high photosynthetic capacity. These principles can be applied to plant selection in plant breeding.

The density of stomata in leaf number eleven of ‘Colín V-33’ plants varied between 105.88 and 317.65 stomata·mm^-2. This interval agrees with that reported by ^{Barrientos-Priego et al. (2003)}, who stated that in leaf 11 a higher stomatal density is detected in three avocado races, ranging from 100 to 610 stomata·mm^-2 in various races and cultivars. These values are within those established for the underside of leaves of C₃ plants (^{Leegood, 1993}). The variation found gives the opportunity to select contrasting genotypes with respect to stomatal density, and may be useful for dwarfing, as proposed by ^{Barrientos-Priego and Sánchez-Colín (1987)}, who observed that higher stomatal density is associated with smaller avocado tree size.

Stomatal size and density are the variables most sensitive to changes in environmental conditions (^{Hetherington & Woodward, 2003}), although they also respond to other factors such as the type of intergraft (^{Ayala-Arreola et al., 2010}). It has been proposed to pre-select avocado genotypes at the seedling stage based on stomatal density (^{Barrientos-Pérez & Sánchez-Colin, 1982}); however, it is important to consider the position of the leaf in the seed-derived seedling, since stomatal density varies, although it has been indicated that it stabilizes in leaf 11 (^{Barrientos-Priego et al., 2003}).

The average stomatal index (SI) on both sides of the leaf varied between 11.78 and 26.16 % (Table 1). The change in SI depends on the radiation received and the environmental conditions occurring during the days preceding the differentiation of the leaf stomata (^{Fricker & Willmer, 2012}). The SI also changes due to the effect of the rootstock, as in ‘Persian’ lime grafted on ‘Carrizo’ citrange (^{Cañizares et al., 2003}), in ‘Valencia’ orange grafted on rootstocks tolerant to the citrus tristeza virus (^{Arrieta-Ramos et al., 2010}) and in avocado when having an intergraft (^{Ayala-Arreola et al., 2010}). For this reason, there are modifications in the SI in the grafted variety due to the effect of the rootstock.

The SI is considered a diagnostic characteristic used in plant systematics, because it generally remains unaltered. Such is the case of seed-derived avocado seedlings according to their leaf position, despite the great differences in stomatal density (^{Barrientos-Priego et al., 2003}), which constitutes an attribute for the identification of genotypes in early stages of development under optimal growing conditions. However, it should be considered that the SI is affected by stressful conditions, both environmental and nutritional (^{Salas, Sanabria, & Pire, 2001}), so the possible effects caused by stress conditions should always be considered.

^{Taylor et al. (2012)} concluded that the density and SI of a species are important, as it has been shown that maximum stomatal conductance is determined by stomatal size and density, being lower in C₄ grass species than in C₃ species. The stomatal system controls CO₂ assimilation completely, and high temperatures inhibit its functioning due to stomatal closure caused by increased internal CO₂; however, if closure is slow, transpiration can continue. In the case of avocado, the stomatal system is highly sensitive to plant water stress (^{Barrientos-Priego & Rodríguez-Ontiveros, 1994}). Another factor that modifies this system in avocado is the position of the leaf within the plant canopy, since the opening of the stomata is affected by light; that is, it is slower if the leaf has developed under shade (^{Heath & Arpaia, 2004}).

In this study, the transpiration rate ranged from 0.32 to 2 mmol·m^-2·s^-1 of H₂O (Table 1), with a coefficient of variation of 45.32 %, denoting moderate variation. In addition, a positive and significant correlation was found between the variables stomatal conductance and transpiration rate (r = 0.96**; P ≤ 0.01; Table 2). Conductance refers to the control exerted by the stomata on the transpiration rate, which represents the ability of a water molecule to diffuse through the leaf per unit of time. A low stomatal conductance indicates lower CO₂ assimilation, and in turn a lower transpiration rate (^{Fricker & Willmer, 2012}). The most important environmental factors affecting transpiration are solar radiation, air vapor pressure deficit, temperature, wind speed, CO₂ concentration, and soil moisture and nutrient availability (^{Sánchez-Díaz & Aguirreolea, 2008}; ^{Pritchard & Amthor, 2005}).

There is a moderate negative correlation between the variables CO₂ assimilation rate and internal CO₂ concentration (r = -0.64**), because the cells have a certain metabolic capacity to process the CO₂ that the plant can capture. When the CO₂ concentration is very high, the stomata close, stopping photosynthesis in the plant and saturating its metabolic machinery, which prevents the processing of more molecules and is known as the saturation point (^{Azcón-Bieto, Fleck, Aranda, & Gómez-Casanovas, 2008}). Thus, as the CO₂ concentration increases, the net assimilation rate also increases until the saturation level is reached, with kinetics specific to each species according to its photosynthetic mechanism (C₃ or C₄) (^{Azcón-Bieto et al., 2008}). Considering that avocado is a species with C₃ photosynthetic behavior, this explains its better adaptation to cool and humid environments.

The increase in the internal CO₂ concentration indicated no stomata opening, which makes CO₂ assimilation impossible due to low gas exchange, and may generate a decrease in plant growth and development (^{Giuliani et al., 2016}). Avocado leaf CO₂ assimilation is closely correlated with stomatal conductance under most conditions (^{Liu, Mickelbart, Robinson, Hofshi, & Arpaia, 2002}), possibly because CO₂ assimilation is mainly determined by stomatal conductance (^{Scholefield, Walcott, Kriedemann, & Ramadasan, 1980}). Nevertheless, according to what was found, there is a correlation between these variables (r = 0.35*, P ≤ 0.05) (Table 2), although it is considered a very weak relationship despite its significance (^{Schober, Boer, & Schwarte, 2018}) and caution should be exercised in considering it as a true existing relationship (^{Taylor, 1990}).

In the anatomical variables, a moderate positive correlation was observed between stomatal density and epidermal cell density (r = 0.59**, P ≤ 0.01), similar to the correlation coefficients found by ^{Barrientos-Priego et al. (2003)}, which showed a positive and significant association between them. These authors also found a negative correlation between epidermal cell density and stomatal index, which was not corroborated in this study, as the coefficient was very low (r = -0.25*, P ≤ 0.05).

Stomatal density and stomatal size are considered key ecophysiological parameters in influencing stomatal conductance (^{Holland & Richardson, 2009}). Based on the above, it is possible to consider the use of stomatal conductance as a quick test in the selection of genotypes, as an indirect measure of stomatal density, since this can be a way of separating short-sized genotypes as proposed by ^{Barrientos-Priego and Sánchez-Colín (1987)}. However, no association was found between conductance and stomatal density (r = -0.02).

Conclusions

The leaves of ‘Colín V-33’ avocado seedlings have hypostomatic stomata and are classified as anomocytic.

The lack of association between stomatal characteristics (stomatal index and density) and gas exchange in leaf 11 of ‘Colín V-33’ avocado seedlings prevents recommending these variables as early selection criteria for breeding.

Due to the variation found in stomatal density and stomatal index, it is possible to select contrasting genotypes that could have an impact on grafting when used as rootstocks, given the evidence from other studies where there are modifications in these variables.