Introduction

Carbon distribution refers to the carbon allocated to the production of different parts of the plant (^{Friend et al., 1994}); two theories have been proposed: the optimal (^{Bloom et al., 1985}), which states that a plant directs it to the organ that acquires the most limiting source, and that of allometric biomass (^{Niklas and Enquist, 2001}) which sends it to the aerial section or the root, depending on its size, calculated by dry weight or length. The amount of carbon devoted to each of these two structures depends on growth conditions and life history, and is important because it influences the survival capacity of pines after planting (^{South, 2000}). Therefore, it is convenient to know this information in species used in reforestation and conservation.

In relation to growth conditions, a decrease in shoot mass and root mass in Pinus pinaster Ait., P. pinea L., P. canariensis C. Sm. ex DC., P. halepensis Mill. occurred in growth chamber where situations of low availability of water prevailed (^{Chambel et al., 2007}) and in greenhouse (^{Aranda et al., 2010}); in the latter environment the same behavior has been observed in cultivated hardwoods (^{Guarnaschelli et al., 2003}).

On the other hand, plants confer a larger amount of carbon to build lateral roots in infertile (^{Paz, 2003}) or humid places (^{Markesteijn and Poorter, 2009}), as well as to make up a deep root if the seeds come from sites with low water availability and develop in greenhouses with sufficient water and nutrients (^{Climent et al., 2008}).

Thus, the carbon distribution is subject to factors such as the provenance or scope of growth and it is feasible to evaluate it indirectly through variables associated with it, which facilitates the identification of species with better response under natural conditions. Therefore, in the hypothesis of this work it was considered that there is significant variation in the carbon distribution in species with different survival in the field: Pinus michoacana Martínez with 76 % (^{Gómez-Romero et al., 2012}); P. pseudostrobus Lindl. from 2 to 68 % (^{Muñoz et al., 2011}; ^{Gómez-Romero et al., 2012}) and P. martinezii E. Larsen, for their conservation value.

Materials and Methods

The variables that were considered were the fresh and dry weight of the shoot and root, specific length of the root and their morphology in seedlings of the selected species that were grown at the nursery.

A total of 6 960 Pinus martinezii seeds were collected in Los Azufres, Ciudad Hidalgo municipality, Michoacán State, México (^{Morales, 2014}) equivalent to 100 g and the same volume for P. michoacana Martínez and P. pseudostrobus Lindl. seeds, which were provided by the Comisión Forestal del Estado de Michoacán (Forest Commission of the State of Michoacán).

From this material, one hundred seeds per species were taken for this study. A pregerminative treatment of stratification was applied, which consists of subjecting them to a temperature of 4 °C for 15 days. The pre-treated seeds were seeded in unicell cavities in peat-agrolite substrate in a ratio of 1: 2 (volume: volume), then placed in a germination-growing chamber under conditions of 25 °C, 75 % of RH and 14 hours of photoperiod.

From the seedlings obtained (with a percentage of germination between 85 and 95 %), 20 of 6 cm high were selected per species, which were transplanted to a mini-rhizotron system (^{Climent et al., 2008}). These mini-rhizotrons were formed from polyvinyl chloride (PVC) tubes 10 cm in diameter and 35 cm in length, which were cut by the center, resulting in two mini- rhizotrons from each tube (Figure 1).

P = Clear cell polycarbonate; PVC = Pilivinyl chloride; C = Plywood plywood 3 mm thick.

Figure 1 Mini-rizotron for the growth of Pinus pseudostrobus Lindl., Pinus martinezii E. Larsen and Pinus michoacana Martínez plants. 

150 days after transplantation, a destructive analysis was performed with 10 plants per species selected at random. From each of the structures studied, the fresh weight and dry weight were recorded by analytical balance (Ohaus l PA214 mode).

From the root, the specific length (LER) (ratio between root length and root dry weight in cm g-1), density (root dry weight ratio and fresh volume in g cm-3 and the fresh volume (water displacement method in 1 ml graduated pipette), all lateral roots were evaluated for number, length (cm) and diameter (mm) effects, and their classification was based on the main root as The total length (LTRL) was calculated by summing all of them.

After the fresh weight measurements, the seedlings were dried in an oven (Terlabo TE-H35D model) at 60 °C for seven days until reaching constant weight (g) in each section.

The analysis of the radical system was carried out by the ImageJ 1.4 program (Macinstosh). An ANOVA was used by a completely randomized design with the S-pluss package (Tibco Software Inc., 2000) with a P value <0.05 for significance.

Results and Discussion

Results indicate that Pinus michoacana reached significantly higher values in fresh and dry weight of the studied parts compared to P. martinezii and P. pseudostrobus. The growth of the first species was 72 to 97 % in the aerial part and 70 to 140 % in the root. There was no significant difference in these variables between P. martinezii and P. pseudostrobus (Table 1).

The growth of the three species has not been recorded from the same system. These results suggest that P. pseudostrobus and P. martinezii come from similar sites, as has been observed in other pines (^{Salazar et al., 1999}) and that the survival of these two species could behave in the same way.

In addition, P. michoacana has a greater potential for survival than P. pseudostrobus and P. martinezii, as it has a larger water storage capacity, as indicated by the fresh weight of shoot and root, and incorporates an amount of assimilated carbon as tissues (dry weight) under similar growth conditions. This coincides with the findings by ^{Gómez-Romero et al. (2012)} who observed greater survival of P. michoacana plants than P. pseudostrobus plants at a degraded site.

The values of the aerial part / root ratio (A / R) were between 1.58 and 1.96 in all three species. Previous studies have shown that conifers allocate more carbon to build up the area than the root (^{Levy et al., 2004}; ^{Mokany et al., 2006}), which agrees with the results of this experiment. This carbon distribution (A / R ratio) coincides with the allometric biomass distribution theory (^{Niklas and Enquist, 2001}) and indicates that all three species come from similar sites in relation to water availability. In this sense, ^{Cregg (1994)} observed greater carbon distribution to shoot in Pinus ponderosa Dougl. ex Laws plants. from different locations with high availability of water.

The variables of the root showed no significant difference between P. martinezii and P. pseudostrobus (Table 2), while the difference between these two species and P. michoacana was significant, which is similar to that observed in root weight and the aerial part. Carbon distribution measured through root morphological variables has not been recorded under conditions of sufficient water availability. The specific length of this structure (LER) was different among the species evaluated in this experiment and was greater than 100 % in P. pseudostrobus and P. martinezii in relation to P. michoacana. These results coincide with those of ^{King et al. (1997)}, who observed a significant difference of LER in two pine species. LER showed an inversely proportional relation with LTRL, LPRL, root density (DR) and root volume (VR). The difference between P. pseudostrobus and P. martinezii in relation to P. michoacana was between 40 and 90 %. The inverse relationship between these variables coincides with previous studies. ^{Ostonen et al. (2007)} found plants with lower LER and greater soil exploration capacity. Therefore, P. michoacana could have a better survival compared to the other species for its ability to explore more soil volume because of its lower RSI. Such behavior in these species of conifers does not depend on the total number of lateral roots, but the exploration depends on the length of the lateral roots of the first order.

The values of root morphology show a clear difference in the carbon distribution in the three species (Table 3). The length of the main root and the number of first order roots was significantly higher in P. martinezii than in P. pseudostrobus and P. michoacana; of the latter, the length and diameter of the first order lateral roots were also found in P. michoacana in relation to the other two species. The largest diameter of the roots of the first, second and third order was confirmed in P. michoacana and P. pseudostrobus compared to that of P. martinezii. These results of root morphology make evident the variation in carbon distribution among pine species and provide further evidence that the greater survival of P. michoacana could be related to the length and diameter of the lateral roots, followed by P. pseudostrobus and finally P. martinezii.

Conclusions

There is variation in the carbon distribution at inter-species level and coincides with the percentage of survival. This relationship was not consistent with variables such as dry weight of aerial part and root, proportion aerial part and root, volume and root length, but other variables such as root specific length (LER), total length of lateral roots, length and Root diameter by root order are parameters with greater accuracy.