Vegetative propagation of Manilkara bidentata (A.DC.) A.Chev. using mini-tunnels in the Peruvian Amazon region

Aim of study: Manilkara bidentata (A.DC.) A.Chev. (‘quinilla’) is a Neotropical tree highly threatened by extensive agricultural practi-ces and climate change, resulting in a substantial reduction of natural stands and seed availability. Commercially, the propagation through seeds of this species is severely impeded by a low germination rate. Vegetative propagation could be an alternative tool for overcoming this limitation. This study aims to evaluate the vegetative propagation of M. bidentata by rooted cuttings using mini-tunnels in the Amazon. Area of study: National University of San Martín, Tarapoto, Region of San Martín, Peru. Forested areas at three localities in the Picota province and two localities in the Bellavista province, San Martín Region, were visited for the selection of plus trees and collection of epicormic shoots from stumps. Materials and methods: the study was performed on leaf areas of 0, 50, and 100% with indole-3-butyric acid (IBA) treatments of 0, 3000, and 6000 ppm, for 9 treatments combinations. The experiment consisted of 3 mini-tunnels, with 3 growth trays established on each mini-tunnel, and 16 cuttings established on each tray (and per treatment), for a total of 144 cuttings. Main results: after 55 days, the best scores in rooting rates (75%), number of roots (3.88), root length (3.26 cm), and budding percentage (94%) were obtained for the combination of 50% leaf area left with 3000 ppm of IBA. Research highlights: we propose the technique of mini-tunnels as a tool for the cloning, rescue, and germplam conservation of M. bidentata . and 100%) in the vegetative propagation of M. bidentata rooted cuttings using plastic mini-tunnels.


Introduction
Manilkara bidentata (A.DC.) A.Chev. ('quinilla' in Spanish) is a forest tree species belonging to the family Sapotaceae used in the Peruvian Amazon for its wood. Besides climate change, this species is highly threatened by extensive agricultural and grassland establishment, leading to its genetic erosion through the substantial reduction of natural stands and seed availability. M. bidentata is a shade-tolerant perennial and dioecious species with epigeal, limited, and irregular germination (Reynel et al., 2003), and a low number of aggregated seedlings in the understory growing near the maternal tree (Uriarte et al., 2005). Thus, seedling emergence is more dependent on seeds availability, space, and resources under the maternal tree canopy, than on the seeds' germination capacity. Still, seeds may present some type of dormancy associated to the hardness of the seed coat growing at low light levels.
There is an urgent need to plant and protect M. bidentata in the Peruvian Amazon, but so far this has been done mainly using seeds and heterogeneous and low-quality germplasm. Unlike seed propagation, clonal propagation allows selecting bigger, more vigorous genotypes. Vegetative propagation emerges as an alternative tool to overcome dispersal limitations and massively replicate selected (high-quality) genotypes in M. bidentata (Vallejos-Torres et al., 2014;Solis et al., 2019), hopefully allowing the rescue of this economically important species. Auxin application can be further improved with the use of mini-tunnels, technique so far not proven in M. bidentata, and which can allow the efficient transformation of hundreds of cuttings into plants using relatively few space and time (Vallejos-Torres et al., 2014, 2020. In Peru, rooting of cuttings in this type of mini-tunnels has proven very successful in a number of tropical species as Calycophyllum spruceanum (Vallejos-Torres et al., 2014), Coffea arabica (Vallejos-Torres et al., 2020), Plukenetia volubilis (Solis et al., 2019;Ruiz-Solsol & Mesén 2010) Hevea brasiliensis (Vallejos-Torres et al., 2021), and Simarouba amara (Soudre et al., 2010).
It is well established for each species that there is an optimal leaf area for rooting, as it reaches a maximum when cuttings/seedlings are photosynthetically active, leading to root development and producing elongation assimilates. This has been proven in several tropical trees (Díaz, 1991). Auxins and co-factors are generated in young leaves and cuttings, then transported to the stem base were they induce rooting (Moraes et al., 2014). Despite this, in some Amazonian trees the relationship between leaf area and rooting is not so clear (Díaz, 1991). A combination of an unequal hormonal distribution in the plant material and a need for reducing its transpiration (Wu et al., 2015) could result in this unclear relationship, which needs further and specific exploration in M. bidentata. The objective of this study was to determine the effects of IBA dose (0, 3000, and 6000 ppm) and leaf area (0%, 50%, and 100%) in the vegetative propagation of M. bidentata rooted cuttings using plastic mini-tunnels.

Study site and preparation of plant material
This study was conducted in the nursery of the Plant-Tissue Culture Laboratory, National University of San Martín, Tarapoto, San Martín Region, Peru, in July-August, 2019 (06°35'28''S, 76°18'47''W; altitude: 330 m.a.s.l.; mean annual temperature: 24.6°C; mean monthly precipitation: 63.1 mm; 45% average humidity) (El Porvernir weather station, 2019). Forested areas at the localities of Winge, Huañipo, and Leoncio Prado in the Picota province (6º58'22.7899"S, 76º15'40.6191"W; 276 m.a.s.l.), and at the localities of San Rafael and Carhuapoma in the Bellavista province (6º59'06.4255"S, 76º32'55.6043"; 286 m.a.s.l.) (San Martín Region) were visited for the selection of plus trees, and for the collection of 10-15 cm (in length) epicormic shoots from stumps that were cut and/or burned by farmers. 20 shoots were collected by each ortet; 10 ortets were collected on each locality between 6:00 am to 8:00 am to avoid hydric stress due to cold transportation (ice and freeze gel at the bottom followed by layers of wet paper and the shoots).
The 10-15 cm hardwood epicormic shoots had an average diameter of 6.5 mm and were trimmed carefully just above a node (straight cut) in 6-8 cm long terminal shoot cuttings, with leaf are left intact (reduction 0%), halved (reduction 50%), or totally removed (reduction 100%). Cuttings were placed in containers with water to avoid physiological stress, soaked in a sodium hypochlorite solution diluted in sterile water (2%) for 5 minutes, and then disinfected in a fungicide solution (thiophanate methyl 2 g/L) for 3 minutes. The auxin used was the indole-3-butyric acid (IBA) (96%, Merck) -dissolved in absolute ethanol as a solvent, in three concentrations: 0 ppm, 3000 ppm, and 6000 ppm. IBA application was precisely directed to the cuttings base with a 10 µl micropipette, ensuring that all cuttings received the same solution amount (10 µl), allowing an exact control of the IBA amount and concentration applied on each cutting. Alcohol was evaporated with a cold air stream for 30 s directed to each cutting base.

Rooting substrate and mini-tunnels environment
Medium-size sand, sieved, and washed with running water (rinsed 4 times) and sodium hypochlorite (diluted in sterile water at 2%), was used as rooting substrate. The sand was sun-dried and sterilized in an autoclave (131ºC x 15 lb pressure, for 2 hours). After applying the 10 µl IBA dose to the cuttings base, they were placed in growth trays (52 cm X 28 cm; 180 g) with the sterilized sand at an approximate depth of 2 cm. The rooting was carried out in a 48 m 2 greenhouse with 6 mini-tunnels, each with 3 nebulizers to prevent desiccation and hydric stress. Mini-tunnels operate with an automated irrigation/cooling system that maintains relative humidity above 75%, and a temperature of 28°C -35°C. An automated irrigation system controlled by a timer allowed 1-minute irrigation every 2 hours (from 7 am to 5 pm) with a frequency of 5 times/day, supplying a total of 21 L of water/day per mini-tunnel. Mini-tunnels were made of a galvanized metal frame with dimensions 1 m x 3 m x 0.60 m with 1 m-long support legs and covered with white transparent plastic which allows the passage of diffuse sunlight.
After 55 days, when four 3.5 cm roots per cutting were obtained in average (determined by randomly picking cuttings), the established trays in mini-tunnels were removed to extract the M. bidentata rooted cuttings. The experiment consisted of 3 mini-tunnels, with 3 growth trays established on each mini-tunnel, and 16 cuttings established on each tray, for a total of 144 cuttings. The rooted cuttings were acclimatized in a nursery with a shade system allowing 20% sunlight. Micro-spray irrigation was made 4 times per day (eight 200 ml sprinklers were 3 Vegetative propagation of quinilla by mini-tunnels. used, supplying a total of 6.4 L of water/day) to maintain the turgidity of the rooted cuttings, which achieved their hardening after 100 days.

Experimental design and data analyses
The experimental design adopted was a full factorial of the 2 experimental factors investigated, i.e. IBA concentrations (0, 3000, and 6000 ppm; each one in a separate mini-tunnel) and leaf surface (0, 50, and 100%), corresponding to 9 combinations (each one in a separate tray). 16 cuttings were used for each combination, for a total of 144 cuttings. Data were analyzed in SPSS by the Tukey test with a significance level of p <0.05 to determine the nature of the differences between treatments and combinations. ANOVAs were performed to examine the effects of IBA dose, leaf area, and their interaction in the different parameters measured. Prior to analysis, the number of roots was transformed to √(x + 1), and the percentages of rooting and budding were transformed to arcosin √% of rooting. Before and after the transformations, the homogeneity of variances and the normality of the residuals were checked using the Bartlett test and graphical checks, respectively.

Results and Discussion
The treatment combination of 3000 ppm of IBA with a 50% of leaf area had a greater number of roots (3.88 roots per cutting), rooting (75%), root length (3.26 cm), and budding (94%), compared to the other combinations (Table  1a). When doing ANOVAs, just budding was affected by IBA dose and leaf area (but not their interaction), while the other variables were not (Table 1b). As in other studies, we found that the auxin dose reached a plateau (saturation), after which rooting decreased, possibly due to the toxic effects of overdose. Possibly, with the above-mentioned combination, a greater photosynthetic, hormonal, and/or co-factor balance and movement of carbohydrate reserves in the cuttings during the rooting process were promoted, at the same time covering the nutritional needs during this process. Rooting is maximized when cuttings are most photosynthetically active (Newton et al., 1992;Mesén et al., 1997). For Plukenetia polyadenia and P. volubilis, Solis et al. (2017Solis et al. ( , 2019 has obtained akin results using similar combinations of leaf area (50 -75%) with IBA dose (2000 ppm). In Simarouba amara, a combination of 60 cm 2 of leaf area and a higher IBA dose (8000 ppm), did not lead to a higher rooting (64%) (Soudre et al., 2010).
Our results are also consistent with those of other Amazonian trees (Mesén, 1998). Calycophyllum spruceanum was successfully rooted (99%) using 2 leaves cuttings and 3000 ppm of IBA (Vallejos-Torres et al., 2014). Similar results were obtained for Plukenetia volubilis (Ruiz-Solsol & Mesén, 2010), Cabralea canjerana (Gimenes et al., 2015), Prunus africana (Tchoundjeu et al., 2002), Gmelina arborea (Díaz 1991), and Cheilocostus speciosus but using 500 mg/L of 1-naphthalenacetic acid (Ballesteros a & Álvarez, 2017). De Souza et al., (2014) reached 98-100% rooting of Prosopis alba cuttings under different IBA doses (3000 -7500 mg/L). These studies tend to show the positive effect of leaf area on the stimulation of the rooting process (Moraes et al., 2014) -in part related to photosynthetic activity during propagation, but there is not always a direct relationship between leaf area and rooting, especially in tropical species (Díaz, 1991). Thus, it is vitally important to know the optimal range of leaf area and its effect on forest species to achieve optimal rooting. Root induction is one of the most exploited physiological effects of auxins, driven by the stimulation of cell dedifferentiation. Auxins have been associated to an increase in carbohydrates and co-factors transport and DNA synthesis, resulting in higher cell division, and therefore higher rooting (Hartmann & Kester, 1983;Gaspar & Hofinger, 1988). This, in turn, strongly affects photosynthetic performance, spectral reflectance, water potential, cell tissue electrical conductivity, and chlorophyll content, all of which have great importance for plant survival (Wu et al., 2015). Just in a couple of studies the vegetative propagation of M. bidentata has been assessed. When auxins were not used and 2-4 cm cuttings were sown in a soil-sandsawdust mix, rooting of this species just reached 40% (Rodríguez & Avella, 2005). In contrast, Cervantes Owaki (2011) obtained an 83% rooting in M. bidentata cuttings (obtained from commercial seedlings) growing on a sand medium with 0.8% of indole-3-butyric acid (IBA) under sub-irrigation chambers. Although our maximum rooting was somewhat lower (75%), Cervantes Owaki (2011) cut living seedlings to obtain the cuttings, thus killing one plant to obtain other -while our study was based on epicormic shoots from stumps that were already cut and/ or burned by farmers. Given that the cutting and burning of M. bidentata is -unfortunately-a common practice (to expand paddocks and crops, and/or the wood to be used for fences and firewood), we think that it is better to take advantage of this situation. Another difference is the use of costly (Cachique et al., 2011) sub-irrigation chambers, which, unlike mini-tunnels, allow the propagation of only a few dozen cuttings. Thus, our study can be a first step into the massive production of advantageous genotypes of M. bidentata in a relatively inexpensive manner. Murillo et al. (2001) considers that 70% is the minimum rooting percentage required to justify a commercial clonal propagation program. In this study we optimally achieved (rooting of 75%) the vegetative propagation of Manilkara bidentata through the use of cuttings with a 50% of leaf area and 3000 ppm of IBA, successfully obtaining vigorous cuttings. This technique could allow the rescue and repopulation of this species in its natural habitat. The propagation by rooting through mini-tunnels can be used to conserve the germplasm of this species for future genetic improvement, rapidly multiplying superior genotypes, thus avoiding the loss of plant material due to the expansion of the agricultural frontier.