The species C. canephora originated from tropical rainforests and has its reproduction majorly done by outcrossing, due to its mechanism of gametophytic self-incompatibility, which prevents self-pollination and limits breeding between plants of similar genetic heritage (Devreux et al., 1959). This characteristic causes natural populations of this species to present high heterozygosity and genetic variability for several agronomic traits (Van Der Vossen, 1985; Carvalho et al., 1991). To help create plantations with more standardized plants to enhance the cultivation practices, the breeding programs developed clonal cultivars, exploring the advantages of the asexual propagation of a set amount of matrix genotypes (compatible among themselves for pollination). This is to create more homogeneous plantations regarding some agronomic aspects, such as higher uniformity of canopy architecture, ripening time, grain size, among others desirable characteristics (Bragança et al., 2001; Fonseca et al., 2008).

Among the forms of asexual propagation suitable for use in plants of C. canephora, the most commonly adopted in commercial nurseries is using cuttings of orthotropic stems to produce new clonal plantlets, this technique is viable in large scale and presents rooting percentage between 95 and 100% (Paulino et al., 1985). The most recent alternative for producing plantlets by cuttings is the use of plastic tubes in nurseries; however, it noticeably increases the production costs per plantlet, due to, among other factors, the requirement of using a more expensive substrate. Therefore, alternatives to decrease the production costs or to increase the quality of coffee plantlets produced in plastic tubes are important research goals.

Coffee husk is an interesting alternative among the materials that could be used to decrease the total amount of required substrate to produce plantlets. This material is a residue produced during the processing of coffee fruits (therefore already available in the coffee producing regions) and it is an excellent source of organic matter, being especially rich in nitrogen and potassium (Dzung et al., 2013; Zoca et al., 2014). However, the proportion in which coffee husk could be use in the mixture to partially replace the commercial substrate is still an incognita, as well as the possible effects that the use of this residue could have over the growth and quality of the plantlets.

In this context, the objective of this experiment was to investigate the influence of using coffee husk as partial component  of  the  substrate  to  produce  plantlets   of C.canephora; verifying its possible effects over the plant growth, plantlet quality and gas exchange rates.

Experimental design

The experiment was conducted under controlled conditions, in multiplication nursery specialized in producing clonal plantlets of C. canephora Pierre ex Froehner (Conilon coffee), located in the northern state of Espírito Santo, in Southeast Region of Brazil. The experiment followed a 3×6 factorial scheme, in a completely randomized design, studying three genotypes of Conilon coffee and six proportions of coffee husk in the composition of the substrate for plantlets production. Four repetitions were used and the experimental plots were composed by 16 plantlets grown in a 4×4 grid, with evaluations of the four central seedlings protected by borders in all adjacent spaces of the tray.

Selection and multiplication of genotypes

The three genotypes selected to be used in the experiment are components of the clonal cultivar "Vitória Incaper 8142" (National Cultivar Register: #20471) (Brasil, 2006). The grouping of 13 highly productive genotypes composes this cultivar, and, among those, the genotypes referred as 2V, 5V and 12V were selected for this study due to their known contrasting characteristics. Mature stems were obtained from adult matrix plants, from each genotype, grown in clonal garden conduced with bending of orthotropic stems to stimulate sprouting. The plants were standardized on the subject of age, nutritional and phytosanitary aspects. The branches were cut from the middle section of the stems, discarding both ends (apex and base). Cuttings were made from the collected stems, sectioning them in parts of nearly 4 cm of length, using straight cutting on the base and bevel cutting on the apex, and leaving a pair of leaves per cutting with nearly one third of their original area. The cuttings were made following the current recommendation for asexual propagation for plants of Conilon coffee (Ferrão et al., 2012; Verdin Filho, 2014).

Substrate preparation

The prepared cuttings were inserted in plastic tubes of 280 mL of volume, filled with prepared substrate. The substrate was prepared with the standard material used commercially for multiplication of coffee plantlets. To compose the treatments, different proportions of the commercial substrate were replaced by coffee husk to fill the plastic tubes, at the levels of 0, 20, 40, 60, 80 and 100% of replacement, respectively. The plantlets were cultivated in nursery (Figure 1) and their nutrition, irrigation and pest management were made in accordance with current recommendations for plantlets production of Conilon coffee (Ferrão et al., 2012; Prezotti et al., 2007).

After these evaluations, the plantlets were cut and each plant compartment (roots, stems and leaves) was separated and placed in paper bags, which were taken to laboratory oven with forced air circulation at 60°C, until constant weight. After drying, the biomass of each organ was stablished in electronic scale (0.001 g of precision). The sum of biomass of all organs was used to calculate the total dry matter (DM). The ratios between biomass accumulated in each organ and the total biomass per plantlet were used to calculate the leaf (LMR), root (RMR) and stem (SMR) mass ratios. The mass ratio between above ground (leaves + stems) and underground (roots) organs were used to calculate the proportion of biomass between shoot and roots (S:R). The ratio between the leaf area and the total biomass of the plantlet were used to calculate the leaf area ratio (LAR).

In addition to the plant height and biomass, the stem diameter was measured with digital caliper (precision: 0.01 mm) in order to estimate the quality of plantlets, which was calculated by the method proposed by Dickson et al. (1960), using the Equation 1:

Where: DQ: Dickson’s quality index; TDM: total dry matter (g); PH: plant height (cm); SD: stem diameter (mm); DMAP: dry matter of aerial part (g); DMRS: dry matter of root system (g).

Data analyses

The data were subjected to analysis of variance and, according to the results for each variable, the interactions were unfolded or the factors studied separately, using the Scott-Knott test for the study the means of the three genotypes and regression analysis for the proportions of coffee husk in the substrate. All tests were performed at 5% probability, using the SISVAR software (Ferreira, 2011).

The total accumulation of biomass (in the form of dry matter), the leaf area ratio, the net carbon assimilation and the quality index of the plantlets were influenced by both factors. Independent effects from the variation originated from the differences among genotypes and among proportions of coffee husk in the substrate, without the occurrence of interaction between these factors was observed.

The comparison among the response of the genotypes is presented in Figure 2, where a faster growth of the plants from the genotype 12V is observed, which resulted in plantlets with higher accumulation of biomass (Figure 2A) and higher quality index (Figure 2D). The genotype 2V seems to present slower initial growth, showing plantlets of lower biomass production (Figure 2A) and lower quality index (Figure 2D) in the end of the same period of time than the other genotypes. The genotype 5V presented intermediate growth between the other two genotypes. The leaf area ratio (Figure 2B) and the photosynthetic rates (Figure 2C) were homogeneous among the three genotypes.

The isolated effect of the proportions of coffee husk in the substrate is presented in Figure 3. The total accumulation of biomass (Figure 3A) presented fit to a 2nd degree linear regression model with maximum point at 38% of replacement of substrate with coffee husk. Similarly, the leaf area ratio (Figure 3B) increased with the use of coffee husk up to the level of 49% of replacement, decreasing after this point. The overall quality of the plantlets (Figure 3D) followed a similar behavior, presenting fit to a 2nd degree linear model with maximum point at 38%. The photosynthetic rate (Figure 3C), however, decreased linearly with the use of coffee husk in the substrate, which caused limitations up to 24% over the net carbon assimilation achieved by the plantlets grown without use of coffee husk.  For  the remaining variables, there was significant interaction between the effects of genotypes and levels of coffee husk in the substrate; therefore, the unfolding of the interactions was performed. The comparison among genotypes in each level of coffee husk used to replace the commercial substrate is presented on Table 1. For height, the genotype 2V presented the taller plantlets without the use of coffee husk in the substrate, showing that this genotype maybe metabolically invested in the early vertical gain, resulting in taller and thinner stems. For the levels of 20% and 40% of coffee husk in the substrate, the genotype 2V also presented taller plantlets, but the means did not differ from the genotype 12V. Above this level of use, the effects of the coffee husk seem to limit the differentiation among genotypes, resulting in similar means (Table 1).

The leaf area only differed among genotypes for the levels of coffee husk between 40 and 80%. The genotypes 12V and 5V presented larger leaf area for the levels of 40 and 60%, respectively, while the genotype 12V alone presented a higher mean of leaf area for the level of 80%; showing that this genotype invests early in the development of leaves (Table 1). The ratio of biomass between roots and aerial organs seems to be homogeneous among genotypes regardless of the use of coffee husk. The only exception happened with the proportion of 80% of the husk in the substrate. For this isolated condition, the genotype 12V allocated higher amount of its biomass towards the roots than the others did (Table 1). There were significant differences among genotypes for the mass ratio allocate towards each  organ  (leaves,  roots and stems); however, a similarity of patterns can be observed from the genotypes regardless of the use of coffee husk. Overall, the genotype 2V, compared to the others, seems to have an early allocation of biomass towards the leaves in detriment of stems (Table 1).

The unfolding of the effect of the use of coffee husk for each studied genotype is presented in the regression analyzes of Figure 4. Even if the magnitude of the coefficients is different, a similar pattern of response can be noticed among the genotypes. For the growth in height (Figure 4A) and leafiness (Figure 4B), the genotypes 12V and 5V present significant fit to linear regression models of 2nd degree, with maximum points at the levels of 46% (12V) and 43% (5V) of replacement of substrate with coffee husk for plant height, and at 46% (12V) and 49% (5V) for leaf area. The genotype 5V had both the plantlet height and leafiness decreasing linearly with the use of coffee husk (Figure 4A and B).

The mass ratios were not significantly influenced by the proportion of coffee husk in the substrate, each genotype keeping its own pattern of biomass allocation unchanged. Therefore, these variables did not present adjustment to linear regression models (Figure 4C, D, E and F). In relation to the other parameters of gas exchange, there was significant interaction between the factors for stomatal conductance, transpiration rate and for intrinsic water use efficiency. The factors did not cause significant effects over the intercellular COâ‚‚ concentration. The comparison among genotypes in each level of coffee husk used to replace the commercial substrate is presented on Table 2.

The water use efficiency presented difference among genotypes for most levels of coffee husk used in the substrate, only not differing at the levels of 40 and 80%. The genotype 12V presented higher efficiency than the others when grown without use of coffee husk did. Overall, for the plantlets grown using coffee husk in the substrate, this genotype together with the genotype 5V presented the higher water use efficiencies (Table 2). This fact may be related to the investment in stem growth and leafiness, which the genotypes 12V and 5V presented in the same conditions, which may have promoted the development of a more robust transport system to sustain the carbon metabolism of the larger leaves (Table 1). The unfolding of the effect of the use of coffee husk over the remaining gas exchange parameters for each studied genotype is presented in the regression analyses of Figure 5. For stomatal conductance, it was observed that only the behavior of the genotype 12V presented adjustment to linear region model of 2nd degree, with maximum point at 50% of coffee husk use. The genotypes 2V and 5V presented linear decrease in the stomatal conductance with the use of coffee husk, with a sharper decrease being observed for the genotype 2V, as observed by the higher angular coefficient from its regression model (Figure 5A).

The difference in responses among the genotypes is due to high phenotypic and genotypic variability of the species. Other studies with the same genotypes have showed that beside the existence of diversity of growth patterns among them, there is also different patterns of response to environmental changes related to the soil conditions and fertility (Martins et al., 2013; Contarato et al., 2014; Colodetti et al., 2014). This must be taken in consideration when planning for the multiplication of the plantlets. 

The coffee husk is a residue that may be used in mixtures in the substrate to promote plantlet growth and quality; however, this by-product can cause detrimental effects on the net photosynthetic rate, growth and quality of the plantlets if used in high proportions. Largely, the results suggest beneficial effects of the replacement of substrate with coffee husk up to the level of one third of the total volume used in the plastic tubes, with most negative effects showing above the level of 38% of substitution. There are different patterns of response to the use of coffee husk, therefore the before mentioned effects may show with different intensities depending on the genotype, which must be taken into consideration by further researches to help define safety levels and a possible recommendation to use coffee husk in the substrate.

The authors have not declared any conflict of interests.

The authors are grateful to the Consórcio Pesquisa Café and Empresa Brasileira de Pesquisa Agropecuária (Embrapa) for their financial support. They are also appreciative of the Centro de Ciências Agrárias e Engenharias (CCAE) of the Universidade Federal do Espírito Santo (UFES), for providing access to the necessary facilities and tools.