Estimates of genetic parameters and correlations for the breeding of jambu (Acmella oleracea)

ABSTRACT

Jambu is important nationally and internationally for its dietary and medicinal use, notably for espilanthol. Despite its versatility, breeding work is scarce. This study aimed to evaluate aerial traits and estimate genetic parameters in jambu accesses to support a breeding program. The experiment was conducted in a greenhouse at the Universidade Federal da Amazônia, Capitão Poço campus. The treatments included 172 genotype clones from seven municipalities in the Northeast mesoregion, obtained from a farmer’s market. Seedlings were cloned and transplanted after three weeks. The design was completely randomized with three replications and 20 cm × 20 cm spacing. At harvest, genotypes were manually removed and taken to the Campus Agricultural Engineering Laboratory to measure 17 aerial part traits. Variance components, genotypic values, and genetic parameters were estimated using REML/BLUP methodology. Genetic and phenotypic correlation matrices and selection gains of 75, 50, and 25% of the genotypes were calculated. Significant genetic variance and high heritability were observed for traits such as plant fresh mass, leaf fresh mass, and stem fresh mass, which are favorable for selection. Indirect selection can enhance leaf and capitula mass without significantly increasing stem diameter. Selecting 50% of the population yields good selection gains without compromising genetic variability.

Key words
Acmella oleracea ; phenotypic correlations; plant breeding

INTRODUCTION

Acmella oleracea (L.) R. K. Jansen, an annual species belonging to the Asteraceae family, is commonly found in tropical and subtropical regions around the world, especially in northern Brazil, where it is known as jambu (Lalthanpuii et al. 2020). In the Brazilian Amazon, the crop is widely used in food; it is a basic ingredient in several typical dishes such as pato no tucupi and tacacá (Brito and Silva 2022).

Jambu is an annual, autogamous, highly branched, semi-fleshy herbaceous species that measures 20 to 50 cm in height (Gusmão and Gusmão 2013). It has simple leaves and its inflorescences, which measure about 2.5 cm, are formed by 400 to 600 tiny yellow flowers, arranged in chapters (Martins et al. 2012).

Spilanthol, the main bioactive compound of the species, is also responsible for the jambu use in traditional medicine for the treatment of various disorders, which makes the crop of global interest (Neves et al. 2019). There is a great versatility of pharmacological uses, such as in the treatment of anemia, cancer, constipation, diuresis, fever, inflammation, liver abscess, peptic and gastric ulcers, toothache, joint rheumatism, dysentery, tuberculosis, among others (Dubey et al. 2013, Rondanelli et al. 2020, Stein et al. 2021).

Although it has great economic potential, both for food purposes and as an herbal product, studies aimed at greater productivity are still scarce, especially using genetic breeding techniques, an extremely relevant alternative since breeding for the environment produces results only to a certain extent point (Ramalho et al. 2012).

In the literature, only one jambu cultivar released by Embrapa Amazônia Oriental is known, the cultivar ‘Nazaré’, the result of seven cycles of individual plants selection for resistance to smut caused by the fungus Thecaphora spilanthes (Poltronieri et al. 2000). Therefore, the need for research aimed at genetic breeding, and the study of traits and conditions that favor the best crop yield stands out (Martins et al. 2012).

In a genetic breeding program, the identification and selection of superior genotypes are two of the most important steps in obtaining new cultivars (Ramalho et al. 2012). For this, in the initial stages, it is essential to know and characterize the genetic variability existing in the population, normally using morphological characteristics for this purpose (Domiciano et al. 2015). Since jambu is a leafy vegetable, genetic breeding of the species mainly emphasizes the plant aerial part. The most relevant characteristics include the increase in the number and size of leaves, the increase in branches and inflorescences, and the reduction in stem diameter.

The improvement of certain morphoagronomic characteristics depends on basic knowledge about the way the characters are inherited, the available genetic variability and estimates of genetic parameters. Therefore, obtaining genetic parameters estimates is essential to elucidate the genetic structure of the populations that will make up future breeding programs for the species (Gomes Jr. et al. 2014). Among the parameters, the following stands out: genetic and phenotypic variance, heritability, selective accuracy, and genetic and phenotypic correlation, which allow inferences to be made about the population genetic variability and what can be expected from the selection gain (Cruz et al. 2014). Despite their extreme importance, works on these themes are not available for jambu yet.

Motivated by the lack of elite material available, we hypothesized that there was variability among the jambu genotypes cultivated by producers in the region and the aerial parts traits had favorable genetic attributes for genotype selection, which can be used in crop breeding programs. In this context, this study aimed to evaluate aerial traits and estimate genetic parameters in jambu accesses to support a breeding program.

MATERIAL AND METHODS

The research was conducted in a plastic greenhouse measuring 16 m × 30 m belonging to the Universidade Federal Rural da Amazônia, campus Capitão Poço (1°44’04.80”S; 47°03’23.33”W).

The genotypes evaluated were obtained from 172 jambu plants, collected in areas cultivated by rural producers and in street markets in seven municipalities in the Northeastern mesoregion of the state of Pará: Capanema, Capitão Poço, Castanhal, Irituia, Ourém, Santa Maria, and São Miguel (Fig. 1). In each municipality, the plants were classified according to the different producers’ locations, based on the information provided by the traders at the time of purchase, resulting in 29 different sources. To conduct the experiment, the plants were cloned by cuttings, according to the recommendations of Santos and Gentil (2015). For each genotype, six cuttings of approximately 10 cm in length were used, taken from basal and apical branches of secondary stems, all with good phytosanitary characteristics.

The cuttings were properly identified, and the clones were planted in disposable cups with a capacity of 500 mL, filled with black soil, for rooting. To reduce water loss through evapotranspiration, the cuttings leaves were cut in half. In addition, all cuttings were buried vertically for at least 1/3 of their length.

After three weeks, the seedlings were transplanted to the final beds. Nineteen rows were needed, in a screened environment. Each row was 7 m long and 1 m wide. A foundation fertilizer corresponding to 8 L of barnyard manure per square meter was applied (Poltronieri et al. 2000). The experiment followed a completely randomized design with three replications. The experimental plot consisted of two of the six clones of each plant, with a spacing of 20 cm between plants and 20 cm between rows, according to the recommendations of Poltronieri et al. (2000).

Figure 1
Location map of the seven collection municipalities, Northeast Pará.

As maintenance fertilizer, foliar fertilizers were applied every 15 days, starting from transplanting. To supply NPK to the plants, the foliar fertilizer UBYFOL was used (nitrogen 15% P/V, phosphorus 10% P/V, and potassium 15% P/V), at a concentration of 80 mL of the product for 20 L of water, as indicated for vegetables. The micronutrients were supplied to the plants by MICROPLEX foliar fertilizer (magnesium 3.6% P/V, sulfur 9.3% P/V, boron 0.14% P/V, cobalt 0.14% P/V, molybdenum 0, 71% P/V, iron 1.14% P/V, copper 0.14% P/V, zinc 5.7% P/V, and manganese 2.9% P/V), in a concentration of 30 mL of the product in 20 L of water, as indicated on the product for vegetables.

During the jambu development, early flower buds were pruned, in addition to manual weeding, to remove weeds whenever necessary. During the rooting phase, watering was carried out using common watering cans, twice a day, the first in the early morning and the other in the late afternoon. In the second phase, when the plants were already in permanent beds, irrigation was carried out using microperforated santene-type hoses, activated twice a day for 2 hours.

At the harvest time, the plants from each plot were manually uprooted and immediately taken to the campus’ Irrigation Engineering Laboratory to measure 17 characteristics: fresh mass matter of the plants (FMP, g), the leaves (FML, g), the open capitulas (FMOC, g), the closed capitulas (FMCC, g), and the stems (FMS, g), all evaluated with a digital scale with a precision of 0.01 g; plant length (PL, cm) using a measuring tape; leaf blade length (LBL, cm) and leaf blade width (LBW, cm), in which four leaf blades per plant were measured with the aid of a 30-cm ruler; open capitula length (OCL, cm) and open capitula diameter (OCD, cm), in which four random capitulas were measured with the aid of a digital caliper; main stem diameter (MSD, mm) also using a digital caliper; number of leaves (NL), number of open capitulas (NOC), number of closed capitulas (NCC), and number of branches (NB); score for caterpillar attack (SCA), score scale from 1 to 5, attributed to damage to the leaves caused by caterpillar feeding, in which 1 represents no damage to the plant and 5 represents a very severe attack on the plant; score for the presence of galls (SPG) caused by Thecaphora spilanthes: scores from 1 to 5, in which 1 represents plants without the presence of galls and 5 represents plants with a large number of galls on the stems.

Before carrying out the statistical analyses, the data for each measured characteristic were tested for the analysis of variance (ANOVA) assumptions. The characteristics fresh mass matter of the plant, along with number of branches, number of leaves, number of open, and closed capitulas, needed transformation to meet the assumptions. The transformations were carried out using the Box-Cox transformation family (Box and Cox 1964), with the aid of the R software (R Core Team 2017), highlighting that the values indicated in the table representations were the real values, before of transformations.

The genetic parameters were obtained based on ANOVA for completely randomized design evaluated in a single location and a single harvest, using the matrix model (Eq. 1):

$y=Xr+Zg+e$ (1)

where: y: the data vector; r: the fixed effects vector of replication added to the general average; g: the random effect vector of genotypes; e: the errors or residuals; X and Z: the incidence matrices for r and g, respectively.

Random effects (best linear unbiased prediction–BLUP) were estimated by solving systems of Henderson’s equations (1975). The variance components (${\widehat{\sigma }}_g^2$: genetic variance, ${\widehat{\sigma }}_e^2$: environmental variance, and ${\widehat{\sigma }}_f^2$: phenotypic variance) were estimated using the residual or restricted maximum likelihood (REML) methodology (Patterson and Thompson 1971). The genetic parameters that comprise broad sense heritability, with respective standard error, and genetic and environmental coefficients of variation for each trait evaluated were estimated using model 83 of the SELEGEN software (Resende 2016).

The selective accuracy (${\widehat{r}}_{\widehat{g}g}$) was estimated by Eq. 2:

${\widehat{r}}_{\widehat{g}g}\sqrt{1-\left(PEV/{\sigma }_g^2\right.}$ (2)

where: PEV: the prediction error variance, obtained from the diagonal of the generalized inverse of the coefficient matrix of the mixed model equations (Resende 2016).

From the genetic and phenotypic variance components, the genetic and phenotypic correlation matrices were obtained. The correlation coefficients significance was obtained using the bootstrap method proposed by Efron (1979). These matrices, as well as the correlation coefficients significance, were obtained using the GENES software (Cruz 2013).

Direct and indirect gains were estimated with selection intensities of 75, 50, and 25% of the evaluated clones. The selection was applied to increase the fresh mass matter of the plants and leaves and to the number of open capitulas. As for the stem diameter, the selection applied was to reduce this trait. For each trait, the direct gain was obtained by averaging the BLUP estimates of the clones selected according to each selection intensity. The indirect responses, that is, the response of trait x when selection is carried out on trait y, were obtained by averaging the BLUP estimates of trait x when ranking the best genotypes for trait y, for each selection intensity applied.

RESULTS

Genetic parameters

There was high selective accuracy for the traits fresh mass matter of the plant, fresh mass matter of the leaves, and fresh mass matter of the stems, with values of 0.72, 0.73, and 0.72, respectively (Table 1). On the other hand, for the number of branches (0.40), number of leaves (0.47), and number of closed capitulas (0.42), the experimental precision was low, as well as for the score for caterpillar attack (0.49) and the score for the presence of galls (0.52). For the other traits, experimental precision was moderate.

Table 1
Estimates of genetic and phenotypic parameters for traits of the aerial part of jambu plants [Acmella oleracea (L.) R. K. Jansen]. Capitão Poço, PA, Brazil.

Traits	CVg/CVe	${\widehat{\sigma }}_{\mathrm{g}}^2$	${\widehat{\sigma }}_{\mathrm{e}}^2$	${\widehat{\sigma }}_{\mathrm{f}}^2$	${\widehat{\mathrm{h}}}_{\mathrm{a}}^2\text{ (\%) }$	${\widehat{\mathrm{r}}}_{\widehat{\mathrm{g}g}}^2\left(\mathrm{\%}\right)$	CVg (%)	CVe (%)	Average
FMP (g)	0.74	0.30**	0.55	0.85	35.42(± 9.08)	0.72	11.23	15.17	112.05
FML (g)	0.77	0.42**	0.37	0.66	36.93(± 9.28)	0.73	14.26	18.64	28.53
FMOC (g)	0.45	0.05*	0.24	0.28	16.59(± 6.23)	0.63	9.69	21.49	18.14
FMCC (g)	0.42	0.01 ns	0.02	0.03	15.13(± 6.00)	0.54	5.95	14.10	2.14
FMS (g)	0.74	0.60**	1.10	1.70	35.39(± 6.00)	0.72	14.33	19.36	58.63
PL (cm)	0.51	16.96**	65.71	82.66	20.51(± 6.92)	0.58	10.68	21.02	38.56
LBL (cm)	0.59	0.34**	0.99	1.33	25.66(± 7.72)	0.66	9.42	16.04	6.20
LBW (cm)	0.53	0.27**	0.97	1.24	21.96(± 7.72)	0.61	9.85	18.58	5.30
OCL (cm)	0.33	0.02 ns	0.19	0.21	9.98(± 4.82)	0.55	7.16	21.49	2.00
OCD (cm)	0.44	0.01*	0.03	0.03	15.95(± 6.10)	0.56	5.76	13.30	1.20
MSD (mm)	0.62	0.27**	0.69	0.95	27.86(± 8.06)	0.66	10.00	16.09	5.14
NB	0.31	0.00 ns	0.00	0.00	8.75(± 4.52)	0.40	0.79	2.55	10.76
NL	0.38	0.35 ns	2.36	2.71	12.87(± 5.48)	0.47	7.40	19.24	140.85
NOC	0.49	0.04**	0.17	0.21	19.56(± 6.75)	0.63	8.97	18.18	18.03
NCC	0.28	0.23 ns	2.98	3.21	7.25(± 4.11)	0.42	9.68	34.61	15.79
SCA	0.40	0.10 ns	0.63	0.74	13.93(± 5.70)	0.49	13.42	33.37	2.37
SPG	0.42	0.10*	0.55	0.65	15.27(± 5.97)	0.52	27.85	65.61	1.13

• ^ns
  non-significant by the likelihood ratio χ² test;
• ^*
  p < 0.05 by the likelihood ratio χ² test;
• ^**
  p < 0.01 by the likelihood ratio χ² test;
• ${\widehat{\mathrm{\sigma }}}_{\mathrm{g}}^2$: genotypic variance; ${\widehat{\mathrm{\sigma }}}_{\mathrm{e}}^2$: environmental variance; ${\widehat{\mathrm{\sigma }}}_{\mathrm{f}}^2$: phenotypic variance; ${\widehat{\mathrm{h}}}_{\mathrm{a}}^2$: broad sense heritability; $\widehat{\mathrm{r}}\underset{\mathrm{gg}}{2}$: selective accuracy; CV_g: coefficient of genetic variation; CV_e: coefficient of environmental variation; FMP: fresh mass matter of the plant; FML: fresh mass matter of the leaves; FMOC: fresh mass matter of the open capitulas; FMCC: fresh mass matter of the closed capitulas; FMS: fresh mass matter of the stems; PL: plant length; LBL: leaf blade length; LBW: leaf blade width; OCL: open capitula length; OCD: open capitula diameter; MSD: main stem diameter; NB: number of branches; NL: number of leaves; NOC: number of open capitulas; NCC: number of closed capitulas; SCA: score for caterpillar attack; SPG: score for the presence of galls.

When experimental precision is evaluated by the coefficient of environmental variation (CV_e), there is, in general, good precision for most traits, except the score for the presence of galls. Furthermore, most traits presented genetic variance significantly different from 0 by the maximum likelihood ratio test (Table 1).

The amplitudes of variances can best be observed by the coefficient of genetic variation (CV_g), which varies from 0 to 100%. The traits with the greatest genetic variance were, respectively, the score for the presence of galls and the fresh mass matter of the stems and leaves, which indicates that, among all the traits evaluated, these showed greater variability. However, there was no prevalence of genetic variation over environmental variation in all traits studied.

Heritability presented moderate values for the traits fresh mass matter of the plant (35.42%), fresh mass matter of the leaves (36.93%), and fresh mass matter of the stems (35.39%). However, it was low for the number of closed capitulas (7.25%) and number of branches (8.75%) (Table 1).

The CV_g/CV_e ratio, also called genetic variation index (IV_g), gives the proportion of genetic variance in relation to the residual error. Therefore, there is no influence of the population average. This ratio can be used as an index indicating the degree of progenies selection ease for each trait. Thus, when the estimated ratio is equal to or greater than unity, there is a very favorable situation for the selection process (Correa et al. 2012). However, no ratios greater than unity were observed.

Estimates of genetic and phenotypic correlations

Correlation estimates indicate good agreement of signs between phenotypic and genotypic correlations. For some traits, the phenotypic correlation was higher than the genetic one, as observed between the fresh mass matter of the leaves and the fresh mass matter of the open and closed capitulas (Table 2). The fresh mass matter of the plant showed strong and positive correlations with some interest traits, such as FMP, FMCC, and NL. The genotypic correlation between the fresh mass matter of the plant and the fresh mass matter of the open capitula was weak for phenotypic and null for genetics. Only the score for caterpillar attack and the score for the presence of galls did not correlate with any other characteristic (Table 2).

Table 2
Phenotypic (below the main diagonal) and genetic (above the main diagonal) correlations between the aerial part traits of jambu plants [Acmella oleracea (L.) R. K. Jansen]. Capitão Poço, PA, Brazil.

	FMP	FML	FMOC	FMCC	FMS	PL	MSD	LBL	LBW	NB	NL	NOC	NCC	SPG	SCA
FMP	1	0.93**	-0.13ns	0.39*	0.58**	0.58**	0.46**	0.46**	0.53**	-0.45*	0.70**	0.50**	0.51**	0.01ns	0.01ns
FML	0.93**	1	-0.11ns	0.41*	0.55**	0.50**	0.47**	0.56**	0.61**	-0.40**	0.68**	0.36*	0.52**	-0.01ns	0.05ns
FMOC	0.44*	0.33ns	1	0.11ns	0.01ns	-0.08ns	-0.10ns	-0.04ns	-0.03ns	0.07ns	-0.13ns	-0.14*	-0.06ns	0.04ns	-0.02ns
FMCC	0.43*	0.44*	0.28ns	1	0.53**	0.16ns	0.21ns	0.23ns	0.30ns	-0.24ns	0.31ns	0.14ns	0.39*	-0.11ns	-0.09ns
FMS	0.58**	0.55**	0.65**	0.56**	1	0.40*	0.30ns	0.30ns	0.33ns	-0.28ns	0.47**	0.40*	0.33ns	0.12ns	-0.01ns
PL	0.58**	0.50**	0.32ns	0.18ns	0.40*	1	0.29ns	0.35*	0.37*	-0.40*	0.56**	0.44*	0.42*	0.10ns	0.11ns
MSD	0.47**	0.47**	0.16ns	0.22ns	0.31ns	0.30ns	1	0.36*	0.40*	-0.27ns	0.43*	0.18ns	0.32ns	-0.09ns	-0.02ns
LBL	0.46**	0.56**	0.06ns	0.24ns	0.30ns	0.35*	0.36*	1	0.89**	-0.06ns	0.29ns	-0.04ns	0.37*	-0.12ns	0.08ns
LBW	0.53**	0.61**	0.13ns	0.31ns	0.33ns	0.38*	0.40*	0.89**	1	-0.11ns	0.37*	0.08ns	0.45*	-0.14ns	0.04ns
NB	-0.45*	-0.40*	-0.26ns	-0.26ns	-0.28ns	-0.41*	-0.28ns	-0.06ns	-0.11ns	1	-0.66**	-0.63**	-0.52**	-0.06ns	-0.01ns
NL	0.70**	0.68**	0.34*	0.33ns	0.47**	0.56**	0.44*	0.30ns	0.37*	-0.66**	1	0.67**	0.73**	0.10ns	0.05ns
NOC	0.50**	0.37*	0.52**	0.15ns	0.41*	0.44*	0.19ns	-0.04ns	0.08ns	-0.63**	0.67**	1	0.37*	0.15ns	0.01ns
NCC	0.51**	0.52**	0.19ns	0.41*	0.33ns	0.43*	0.33*	0.37*	0.45*	-0.53**	0.73**	0.38*	1	-0.00ns	0.02ns
SPG	0.01ns	-0.01ns	0.07ns	-0.10ns	0.12ns	0.10ns	-0.09ns	-0.12ns	-0.14ns	-0.06ns	0.10ns	0.15ns	0.00ns	1	0.24ns
SCA	0.01ns	0.05ns	-0.02ns	-0.08ns	-0.01ns	0.11ns	0.01ns	0.08ns	0.04ns	-0.02ns	0.06ns	0.02ns	0.03ns	0.24ns	1

• ^ns
  non-significant by the bootstrap method based on 5,000 simulations;
• ^*
  p < 0.05 by the bootstrap method based on 5,000 simulations;
• ^**
  p < 0.01 by the bootstrap method based on 5,000 simulations;
• FMP: fresh mass matter of the plant; FML: fresh mass matter of the leaves; FMOC: fresh mass matter of the open capitulas; FMCC: fresh mass matter of the closed capitulas; FMS: fresh mass matter of the stems; PL: plant length; LBL: leaf blade length; LBW: leaf blade width; OCL: open capitula length; OCD: open capitula diameter; MSD: main stem diameter; NB: number of branches; NL: number of leaves; NOC: number of open capitulas; NCC: number of closed capitulas; SCA: score for caterpillar attack; SPG: score for the presence of galls.

Selection gains

It was observed that the gains were greater for a selection intensity of 25% of the population. However, the gains with 50% of the selected population were not substantially smaller (Table 3).

Table 3
Direct* and indirect+ gains in percentage for the selection of jambu genotypes [Acmella oleracea (L.) R. K. Jansen] with greater FMP; FML; NOC; and smaller MSD for selection intensity of 25, 50 and 75% of the population. Capitão Poço, PA, Brazil.

Trait	FMP			MFF			NOC			MSD
	25%	50%	75%	25%	50%	75%	25%	50%	75%	25%	50%	75%
FMP	9.61*	6.41*	3.44*	8.75+	5.75+	3.24+	5.41+	3.74+	1.64+	-6.30+	-3.17+	-1.28+
MFF	11.40+	7.11+	3.97+	12.46*	7.78*	4.24*	4.62+	3.68+	1.50+	-7.74+	-4.10+	-1.66+
FMS	5.79+	5.03+	2.36+	6.54+	4.36+	2.63+	6.12+	3.20+	1.69+	-5.77+	-3.00+	-0.93+
FMOC	3.02+	2.03+	0.99+	1.78+	1.40+	1.01+	3.75+	2.13+	1.17+	-1.20+	-0.66+	-0.26+
FMCC	0.90+	1.13+	0.48+	1.31+	1.10+	0.59+	0.42+	0.33+	0.31+	-1.26+	-0.55+	-0.08+
PL	3.97+	2.94+	1.73+	3.52+	2.43+	1.54+	3.25+	2.21+	1.18+	-3.74+	-1.62+	-0.35+
MSD	3.62+	2.29+	1.26+	3.03+	2.51+	1.32+	1.78+	1.04+	0.33+	-8.12*	-5.14*	-2.79*
LBL	2.61+	1.60+	1.42+	4.25+	2.68+	1.38+	-0.82+	0.25+	-0.19+	-3.58+	-1.77+	-0.92+
LBW	3.12+	2.04+	1.54+	4.75+	2.92+	1.52+	-0.18+	0.67+	0.21+	-3.87+	-2.31+	-0.92+
OCL	1.46+	1.06+	0.61+	1.01+	0.71+	0.43+	2.20+	1.63+	1.03+	-0.72+	-0.06+	0.10+
OCD	2.19+	1.33+	0.63+	1.58+	1.14+	0.67+	1.23+	0.98+	0.60+	-0.81+	-0.31+	-0.26+
NB	-0.15+	-0.13+	-0.06+	-0.16+	-0.09+	-0.06+	-0.24+	-0.14+	-0.09+	0.17+	0.09+	0.02+
NL	2.84+	2.00+	1.08+	2.94+	1.79+	1.07+	2.62+	1.95+	1.03+	-2.64+	-1.33+	-0.41+
NOC	3.56+	2.38+	1.18+	2.35+	1.38+	1.03+	6.31*	4.06*	2.28*	-2.13+	-0.94+	-0.10+
NCC	1.83+	1.44+	0.96+	2.40+	1.60+	0.98+	1.15+	0.90+	0.60+	-2.35+	-1.16+	-0.25+
SPG	0.45+	0.11+	0.23+	-0.52+	0.90+	0.05+	4.12+	2.53+	0.78+	2.52+	2.38+	0.99+
SCA	-0.83+	-0.04+	0.20+	0.71+	-0.36+	-0.21+	-0.81+	0.37+	0.05+	-0.57+	0.37+	0.06+

• FMP: fresh mass matter of the plant; FML: fresh mass matter of the leaves; FMOC: fresh mass matter of the open capitulas; FMCC: fresh mass matter of the closed capitulas; FMS: fresh mass matter of the stems; PL: plant length; LBL: leaf blade length; LBW: leaf blade width; OCL: open capitula length; OCD: open capitula diameter; MSD: main stem diameter; NB: number of branches; NL: number of leaves; NOC: number of open capitulas; NCC: number of closed capitulas; SCA: score for caterpillar attack; SPG: score for the presence of galls.

DISCUSSION

Genetic parameters

One of the most relevant parameters for evaluating the experiment precision is selective accuracy, as it refers to the correlation between predicted genetic values and true genetic values of individuals (Pimentel et al. 2014). Resende and Duarte (2007) proposed a selective accuracy classification in terms of precision as very high (≥ 0.90), high (0.70 ≤ ≤ 0.89), moderate (0.50 ≤ 0.69), and low (≤ 0.49). However, selective accuracy has not been evaluated in sets of jambu genotype trials yet, and the real values of experimental precision based on this statistic are unknown for the crop.

The accuracy results for the traits fresh mass matter of the plant, fresh mass matter of the leaves, and fresh mass matter of the stems added to moderate heritability values shows high genetic variability, accuracy in identification and possibility of success in population selection, as observed by Pimentel et al. (2014) when estimating genetic parameters of populations and individuals from segregating wheat populations.

It is worth mentioning that there was no agreement between CV_e and ${\widehat{r}}_{\widehat{g}g}$ for some traits, such as the number of branches, in which CV_e indicated very high experimental precision, but for r_ĝg the precision was low. As the number of branches did not present genetic variance, the selective accuracy indicated low precision, which is more appropriate, given that the environmental variation was approximately ten times greater than the genetic variation for this trait. For Cargnelutti Filho and Storck (2009), the selective accuracy and heritability statistics are more appropriate than coefficient of variation to evaluate experimental precision.

The genetic variance results are promising, as this is a parameter of great importance for the breeder’s work progress, indicating greater possibilities of genetic gain with selection (Rocha et al. 2009). It is up to the breeder to define the genetic influence proportion on the environment in the trait of interest expression. When it comes to CV_g, the results for the traits SPG, FMP, and FML make it possible to select (Correa et al. 2012), especially in recurrent selection cycles (Borém and Miranda 2013).

Broad sense heritability is the proportion of genetic variance over the total phenotypic variance, that is, the heritable proportion of the total variability (Ramalho et al. 2012). For this parameter, it is possible to predict good gains for traits that presented moderate values, the opposite applies to the traits NCC and NB, which indicates less genetic control of these traits and highlights greater difficulties in their selection (Correa et al. 2012). It is worth highlighting, however, that the heritability value is not immutable and can be increased by introducing greater genetic variation into the population, as well as improving experimental conditions to reduce the contribution of variation due to the environment to the total phenotypic variation (Carias et al. 2016).

From an experimental point of view, greater control over crop conditions or even the establishment of more rigorous designs can benefit the more accurate estimation of variance components, such as genetic variance. Furthermore, the possibility of controlled crosses in the crop can introduce greater genetic variability into the population through new allelic combinations.

Even if the evaluated traits do not have IV_g greater than 1, progress can be made for traits that presented a ratio close to unity, such as the fresh mass matter of the leaves (CV_g/CV_e = 0.77), fresh mass matter of the plant (CV_g/CV_e = 0.74), and fresh mass matter of the stems (CV_g/CV_e = 0.74), in accordance with the interpretation of this value recommended by Vencovsky (1987).

Estimates of genetic and phenotypic correlations

Genetic and phenotypic correlations are also of great importance in breeding, as through them it is possible to draw more effective selection strategies, such as correlated responses to a low trait heritability or difficult to measure (Cruz et al. 2012). Phenotypic correlation is estimated based on phenotypic variances, which include genetic and environmental variation, whereas genetic correlations are estimated only based on genetic variance (Cruz et al. 2012). Positive correlation coefficient estimates indicate the tendency for one variable to increase when the other increases, negative correlations indicate the tendency for one variable to increase while the other decreases (Nogueira et al. 2012). It is also worth showing that the non-significance of the estimates obtained in genotypic correlations can be attributed to the absence of phenomena such as gene linkage or pleiotropy (Ribeiro et al. 2009).

The strong positive correlations of FMP with some interest traits indicate that the breeder simply needs to select the plants with the greatest mass, which will indirectly increase the FML, NL and NCC, which are more costly traits to evaluate, demanding more time and work. In turn, the null genotypic correlation between FMP and FMOC indicates that the breeder will have to separately evaluate the capitulas mass, by reason of the great commercial capitulas interest as they are the plant part with the highest spilanthol concentration (Chakraborty et al. 2010). As regards the trait SPG, the correlation may have been affected by the variance absence, since covariance and variance between traits are necessary to estimate it (Cruz et al. 2012).

In general, the various traits of jambu’s economic importance are positively correlated with each other, which makes it difficult to obtain cultivars with a high leaves mass and inflorescences, but with less pronounced stems. This latter is particularly important because, when the stem has a larger circumference, it can make chewing difficult, and in cooking both the leaves and the primary and secondary stems of the plant are widely used. However, breeders have statistical tools, such as selection indices, to circumvent or reduce the undesirable correlated responses impacts (Cruz et al. 2012).

Selection gains

One of the most impactful contributions of quantitative genetics to plant breeding was the possibility of predicting gains from selection. These gains can be affected by some factors, such as the selection intensity and the trait heritability under selection, which can increase the gains (Bernardo 2010). However, strong selection intensities lead to inbreeding and, consequently, the loss of variability in subsequent generations (Falconer and Mackay 1996).

To preserve variability, a selection intensity of 50% appears to be the most suitable, particularly in the breeding programs early cycles, to avoid diminishing the effective population size. Populations with small effective sizes are more susceptible to the loss of favorable alleles through genetic drift, leading to a decrease in the average of quantitative traits with economic importance (Falconer and Mackay 1996).

In autogamous species such as jambu, genetic drift can occur when, upon eliminating deleterious alleles by natural or artificial selection, non-deleterious alleles that are linked to them are also eliminated (Veasey et al. 2011). In this context, with the increase in homozygosity through autogamy, genetic variation within populations decreases and random changes in allele frequencies from one generation to another become more evident.

When selection was exclusively targeted at the fresh mass matter of the plant, employing a 50% selection intensity, the indirect gain estimate for fresh mass matter of the leaves closely matched what would be achieved with direct selection for fresh mass matter of the leaves (Table 3). This result supports the strong correlation between these traits. To enhance mass matter of the leaves, the breeder can simply select plants with the highest total mass, eliminating the need to detach the leaves. Additionally, the fresh mass matter of the open capitula yields the same indirect gain as selecting based on the number of open capitulas.

A drawback of exclusively selecting jambu genotypes based on the fresh mass matter of the plant is that it would lead to an increase in the mass and diameter of stems. Consequently, the new cultivars might have the disadvantage of featuring stems that are too thick for traditional culinary uses.

In conclusion, it’s worth noticing that there is a lack of studies providing estimates of genetic parameters, correlations, and gains with selection for morphological traits in jambu in the existing literature. Consequently, making a detailed comparison between the results obtained in this study and others is challenging—a difficulty also encountered by Domiciano et al. (2015) when estimating genetic parameters and diversity in macaúba progenies. Therefore, this work could potentially serve as a blueprint for future studies aimed at the jambu genetic breeding.

CONCLUSION

The results found in this study revealed that economically important characteristics of jambu, such as the mass of fresh matter of the plant and leaves, in addition to being positively correlated, are the most favorable to selection. Thus, selecting only for the mass of fresh matter of the plant is a viable strategy since it simultaneously increases the mass of leaves and inflorescences, although it also increases the mass of the stem. Therefore, it is recommended to use selection indexes to increase the mass of leaves and inflorescences, without increasing the diameter of the stems.

ACKNOWLEDGMENTS

To the Universidade Federal Rural da Amazônia and to the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil, that financed part of this study.

• How to cite: Albuquerque, D. P., Teixeira, D. H. L., Gonçalves Júnior, D. H., Oliveira, L. J. S., Silva Júnior, A. D. and Véras, G. J. S. (2025). Estimates of genetic parameters and correlations for the breeding of jambu (Acmella oleracea). Bragantia, 84, e20240136. https://doi.org/10.1590/1678-4499.20240136
• FUNDING
  Coordenação de Aperfeiçoamento de Pessoal de Nível Superior
  Finance Code 001

DATA AVAILABILITY STATEMENT

The datasets generated and/or analyzed during the current study are available from the corresponding author.

REFERENCES

• Bernardo, R. (2010). Breeding for quantitative traits in plants. 2. ed. Minnesota: Stemma Press.
• Borém, A. and Miranda, G. V. (2013). Melhoramento de plantas. 6. ed. Viçosa: Editora UFV.
• Box, G. E. P. and Cox, D. R. (1964). An analysis of transformations. Journal of the Royal Society: Series B (Methodological), 26, 211-243. https://doi.org/10.1111/j.2517-6161.1964.tb00553.x
  » https://doi.org/10.1111/j.2517-6161.1964.tb00553.x
• Brito, A. E. O. and Silva, C. S. M. (2022). Atividade antimicrobiana de extratos vegetais de especiarias do norte do Brasil. Research, Society and Development, 11, e52011226047. https://doi.org/10.33448/rsd-v11i2.26047
  » https://doi.org/10.33448/rsd-v11i2.26047
• Cargnelutti Filho, A. and Storck, L. (2009). Medidas do grau de precisão experimental em ensaios de competição de cultivares de milho. Pesquisa Agropecuária Brasileira, 44, 111-117. https://doi.org/10.1590/S0100-204X2009000200001
  » https://doi.org/10.1590/S0100-204X2009000200001
• Carias, C. M. O. M., Gravina, G. A., Ferrão, M. A. G., Fonseca, A. F. A., Ferrão, R. G., Vivas, M. and Viana, A. P. (2016). Predição de ganhos genéticos via modelos mistos em progênies de café conilon. Coffee Science, 11, 39-45. Available at: http://www.sbicafe.ufv.br:80/handle/123456789/8171 Acessed on: June 5, 2024.
  » http://www.sbicafe.ufv.br:80/handle/123456789/8171
• Chakraborty, A., Devi, B. R. K., Sanjebam, R., Khumbong, S. and Thokchom, I. S. (2010). Preliminary studies on local anesthetic and antipyretic activities of Spilanthes acmella Murr. in experimental animal models. Indian Journal of Pharmacology, 42, 277-279. https://doi.org/10.4103/0253-7613.70106
  » https://doi.org/10.4103/0253-7613.70106
• Correa, A. M., Ceccon, G., Correa, C. M. A. and Delben, D. S. (2012). Estimativas de parâmetros genéticos e correlações entre caracteres fenológicos e morfoagronômicos em feijão-caupi. Revista Ceres, 59, 88-94. https://doi.org/10.1590/s0034-737x2012000100013
  » https://doi.org/10.1590/s0034-737x2012000100013
• Cruz, C. D. (2013). GENES - a software package for analysis in experimental statistics and quantitative genetics. Acta Scientiarum, 35, 271-276. https://doi.org/10.4025/actasciagron.v35i3.21251
  » https://doi.org/10.4025/actasciagron.v35i3.21251
• Cruz, C. D., Carneiro, P. C. S. and Regazzi, A. J. (2014). Modelos biométricos aplicados ao melhoramento genético. 3. ed. Viçosa: Editora UFV. v. 2.
• Cruz, C. D., Regazzi, A. J. and Carneiro, P. C. S. (2012). Modelo biométrico aplicado ao melhoramento genético. 4. ed. Viçosa: Editora UFV. v. 1.
• Domiciano, G. P., Alves, A. A., Laviola, B. G. and Conceição, L. D. H. C. S. (2015). Parâmetros genéticos e diversidade em progênies de Macaúba com base em características morfológicas e fisiológicas. Ciência Rural, 45, 1599-1605. https://doi.org/10.1590/0103-8478cr20140909
  » https://doi.org/10.1590/0103-8478cr20140909
• Dubey, S., Maity, S., Singh, M., Saraf, S. A. and Saha, S. (2013). Phytochemistry, pharmacology and toxicology of Spilanthes acmella: a review. Advances in Pharmacological Sciences, 1-9. https://doi.org/10.1155/2013/423750
  » https://doi.org/10.1155/2013/423750
• Efron, B. (1979). Bootstrap methods: Another look at the jackknife. The Annals of Statistics, 7, 1‐26. https://doi.org/10.1214/aos/1176344552
  » https://doi.org/10.1214/aos/1176344552
• Falconer, D. S. and Mackay, T. F. C. (1996). Introduction to quantitative genetics. London: Longman Malaysia.
• Gomes Jr., R. A., Gurgel, F. L., Peixoto, L. A., Bhering, L. L., Cunha, R. N. V., Lopes, R., Pina, A. J. A. and Veiga, A. S. (2014). Evaluation of interspecific hybrids of palm oil reveals great genetic variability and potential selection gain. Industrial Crops and Products, 52, 512-518. https://doi.org/10.1016/j.indcrop.2013.10.036
  » https://doi.org/10.1016/j.indcrop.2013.10.036
• Gusmão, M. T. A. and Gusmão, S. A. L. (2013). Jambu da Amazônia: Acmella oleracea [(L.) R. K. Jansen]: Características gerais, cultivo convencional, orgânico e hidropônico. Belém: Editora UFRA.
• Henderson, C. R. (1975). Best linear unbiased estimation and prediction under a selection model. Biometrics, 31, 423-447. https://doi.org/10.2307/2529430
  » https://doi.org/10.2307/2529430
• Lalthanpuii, P. B., Zokimi, Z. and Lalchhandama, K. (2020). The toothache plant (Acmella oleracea) exhibits anthelmintic activity on both parasitic tapeworms and roundworms. Pharmacognosy Magazine, 16, 193-198. https://doi.org/10.4103/pm.pm_321_19
  » https://doi.org/10.4103/pm.pm_321_19
• Martins, C. P. S., Melo, M. T. P., Honório, I. C. G., D’Ávila, V. A. and Carvalho Júnior, W. G. O. (2012). Morphological and agronomic characterization of Jambu (Spilanthes oleracea L.) accessions under the conditions of North Minas Gerais State, Brazil. Revista Brasileira de Plantas Medicinais, 14, 410-413. https://doi.org/10.1590/S1516-05722012000200023
  » https://doi.org/10.1590/S1516-05722012000200023
• Neves, D. A., Schmiele, M., Pallone, J. A. L., Orlando, E. A., Risso, E. M., Cunha, E. C. E. and Godoy, H. T. (2019). Chemical and nutritional characterization of raw and hydrothermal processed jambu (Acmella oleracea (L.) RK Jansen). Food Research International, 116, 1144-1152. https://doi.org/10.1016/j.foodres.2018.09.060
  » https://doi.org/10.1016/j.foodres.2018.09.060
• Nogueira, A. P. O., Sediyama, T., Sousa, L. B., Hamawaki, O. T., Cruz, C. D., Pereira, D. G. and Matsuo, É. (2012). Análise de trilha e correlações entre caracteres em soja cultivada em duas épocas de semeadura. Bioscience Journal, 28, 877-888. Available at: https://seer.ufu.br/index.php/biosciencejournal/article/view/14576 Accessed on: Jun. 5, 2024.
  » https://seer.ufu.br/index.php/biosciencejournal/article/view/14576
• Patterson, H. D. and Thompson, R. (1971). Recovery of inter-block information when block sizes are unequal. Biometrika, 58, 545-554. https://doi.org/10.1093/biomet/58.3.545
  » https://doi.org/10.1093/biomet/58.3.545
• Pimentel, A. J. B., Guimarães, J. F. R., Souza, M. A., Resende, M. D. V., Moura, L. M. and Ribeiro, G. (2014). Estimação de parâmetros genéticos e predição de valor genético aditivo de trigo utilizando modelos mistos. Pesquisa Agropecuária Brasileira, 49, 882-890. https://doi.org/10.1590/S0100-204X2014001100007
  » https://doi.org/10.1590/S0100-204X2014001100007
• Poltronieri, M. C., Muller, N. R. M. and Poltronieri, L. S. (2000). Recomendações para a produção de jambú: cultivar Nazaré. Circular Técnica Embrapa Amazônia Oriental, 11. Available at: https://www.infoteca.cnptia.embrapa.br/bitstream/doc/382634/1/ORIENTALCirTec11.pdf Accessed on: June 5, 2024.
  » https://www.infoteca.cnptia.embrapa.br/bitstream/doc/382634/1/ORIENTALCirTec11.pdf
• R Core Team (2017). R: A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing. Available at: https://www.R-project.org/ Accessed on: Jan. 20, 2024.
  » https://www.R-project.org/
• Ramalho, M. A. P., Abreu, A. F. B., Santos, J. B. and Nunes, J. A. R. (2012). Aplicações da genética quantitativa no melhoramento de plantas autógamas. Lavras: Editora UFLA.
• Resende, M. D. V. (2016). Software Selegen-REML/BLUP: a useful tool for plant breeding. Crop Breeding and Applied Biotechnology, 16, 330-339. https://doi.org/10.1590/1984-70332016v16n4a49
  » https://doi.org/10.1590/1984-70332016v16n4a49
• Resende, M. D. V. and Duarte, J. B. (2007). Precisão e controle de qualidade em experimentos de avaliação de cultivares. Pesquisa Agropecuária Tropical, 37, 182-194. Available at: http://repositorio.bc.ufg.br/handle/ri/13214 Acessed on: Jun. 5, 2024.
  » http://repositorio.bc.ufg.br/handle/ri/13214
• Ribeiro, E. H., Pereira, M. G., Coelho, K. S. and Freitas Júnior, S. P. (2009). Estimativas de parâmetros genéticos e seleção de linhagens endogâmicas recombinantes de feijoeiro comum (Phaseolus vulgaris L.). Revista Ceres, 56, 580-590. Available at: https://ojs.ceres.ufv.br/ceres/article/view/3470 Acessed on: Jun. 5, 2024.
  » https://ojs.ceres.ufv.br/ceres/article/view/3470
• Rocha, M. M., Carvalho, K. J. M., Freire Filho, F. R., Lopes, A. C. A., Gomes, R. L. F. and Sousa, I. S. (2009). Controle genético do comprimento do pedúnculo em feijão-caupi. Pesquisa Agropecuária Brasileira, 44, 270-275. https://doi.org/10.1590/S0100-204X2009000300008
  » https://doi.org/10.1590/S0100-204X2009000300008
• Rondanelli, M., Fossari, F., Vecchio, V., Braschi, V., Riva, A., Allegrini, P., Petrangolini, G., Iannello, G., Anna, F. M., Peroni, G., Nichetti, M., Gasparri, C., Spadaccini, D., Vittoria, I., Sakina, M., Alalwan, T. and Perna, S. (2020). Acmella oleracea for pain management. Fitoterapia, 140, 104419. https://doi.org/10.1016/j.fitote.2019.104419
  » https://doi.org/10.1016/j.fitote.2019.104419
• Santos, E. R. and Gentil, D. F. O. (2015). Propagação de jambu por estaquia. Comunicata Scientiae, 6, 26-32. https://doi.org/10.14295/cs.v6i1.535
  » https://doi.org/10.14295/cs.v6i1.535
• Stein, R., Berger, M., Cecco, B. S., Mallmann, L. P., Terraciano, P. B., Driemeier, D., Rodrigues, E., Beys-da-Silva, W. O. and Konrath, E. L. (2021). Chymase inhibition: A key factor in the anti-inflammatory activity of ethanolic extracts and spilanthol isolated from Acmella oleracea Journal of Ethnopharmacology, 270, 113610. https://doi.org/10.1016/j.jep.2020.113610
  » https://doi.org/10.1016/j.jep.2020.113610
• Veasey, E. A., Piotto, F. A., Nascimento, W. F. D., Rodrigues, J. F., Mezette, T. F., Borges, A., Biguzzi, F. A, Santos, F. R. C., Sobierajski, G. R., Recchia, G. H and Mistro, J. C. (2011). Processos evolutivos e a origem das plantas cultivadas. Ciência Rural, 41, 1218-1228. https://doi.org/10.1590/S0103-84782011000700018
  » https://doi.org/10.1590/S0103-84782011000700018
• Vencovsky, R. (1987). Herança quantitativa. In E. Paterniani (ed.). Melhoramento e produção de milho no Brasil (p. 122-201). 2. ed. Fundação Cargil.

Edited by

• Section Editor:
  Freddy Mora https://orcid.org/0000-0002-1163-2899

Publication Dates

• Publication in this collection
  20 Jan 2025
• Date of issue
  2025

History

• Received
  21 June 2024
• Accepted
  15 Oct 2024