Selection of Genipa americana L. seed trees by genetic divergence in fruits, seeds, and seedlings

doi:10.21203/rs.3.rs-2132665/v1

Download PDF

Research Article

Selection of Genipa americana L. seed trees by genetic divergence in fruits, seeds, and seedlings

https://doi.org/10.21203/rs.3.rs-2132665/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 26 Nov, 2023

Read the published version in Genetic Resources and Crop Evolution →

Version 1

posted

You are reading this latest preprint version

Analyses with phenotypic characters are more practical to implement in forest nurseries, considering that genetics should be examined in forest restoration projects. Genipa americana L. is an essential species in producing seedlings for this purpose. This study aimed to select G. americana seed trees, for seed collection, in an area of Atlantic Forest, through the genetic divergence study in fruits, seeds, and seedlings. The study was conducted in the Saltinho Biological Reserve in Tamandaré - PE, Brazil. The ripe fruits of 11 G. americana seed trees were collected. The physical and physicochemical characterization of the fruits, biometric characterization of the seeds, and seed emergence analysis were performed. For the genetic analyses, the model used was the 81 of the Selegen REML/BLUP software and the cluster analysis made by the UPGMA method. The seed trees selection index was also created. The coefficients of determination of the seed trees’ effects were low because there is considerable interaction between the sites in which each one is inserted. It is possible to state that all variables contribute to explaining the genetic dissimilarity between the seed trees. Seed trees 1, 4, and 6 had the best average ranks for the variables collected concerning genotypes. The Saltinho Biological Reserve is a good source of genetic and reproductive material for the species G. americana for forest restoration purposes.

forest seeds

selection index

image analysis

genotypic value

cluster analysis

genipap

Brazil has already lost 87% of its initial forest cover (Fundação Sos Mata Atlântica and INPE 2020), with a 30% increase in the Atlantic Forest deforestation rate between 2018–2019 (Fundação Sos Mata Atlântica 2020). Thus, in addition to the urgent preservation of the country's forest remnants, incentives are needed to halt the decline of Brazilian ecosystems.

Mistakenly, often only the tree planting stage is considered when restoring a degraded site, omitting the multidisciplinary character of the project. Financial and labor efforts are in vain, since treating restoration in such a way culminates in the project failure (Chazdon and Brancalion 2019; Holl and Brancalion 2020).

The reproductive material genetic composition used in the projects must be considered for the success of forest restoration (Thomas et al. 2014). The study of genetic divergence among potential seed trees is fundamental. Knowledge of the genetic origin of propagation material and ex situ maintenance of the species complement conservation efforts, serving as a parameter for in situ management (Sebbenn 2002; Potter et al. 2017). The seed trees selection through biometric evaluations in fruits and seeds, fruit chemistry, and seed germination is commonly performed. They represent the parent trees genetic variability and avoid the chance of inbreeding in future projects that use them as a source of reproductive material (Sebbenn 2002; Vieira and Gusmão 2008; Correia et al. 2019).

Analyses in seeds and fruits need to be done to determine which phenotypic traits actually represent greater material genetic control and explain the genetic convergences that may arise (Madrera et al. 2020). The works of Peng et al. (2020), with Punica granatum L., and Hssaini et al. (2020), with Ficus carica L., clearly demonstrate the importance of this study and how genetic information can influence essential species properties, such as fruit flavor, seed production per fruit, and germination speed, among others.

An important Brazilian species for reforestation projects that require a larger pool of genetic information is Genipa americana L., popularly known as genipap. The highlight of this species for restoration is its use as food and shelter for local fauna (Ragusa-Netto 2015; Cunha et al. 2020), facilitating nuclei of diversity with the arrival of new species brought by fauna to its surroundings. Also, G. americana can be implemented in areas with restrictions, susceptible to flooding, and with hydromorphic soils, because the species is resistant to these conditions (Brandani et al. 2018; Pires et al. 2018).

Recent studies involving the genetics of Genipa americana L. are conducted mainly by extracting DNA using either dominant or co-dominant molecular markers (Rabbani et al. 2012; Manoel et al. 2014; Silva et al. 2014; Manoel et al. 2015a; Silva et al. 2015; Ruzza et al. 2018; Melo et al. 2021). Still, when phenotypic characteristics are evaluated, they are placed in the context of understanding their phenotypic plasticity in the face of adverse conditions (Lima et al. 2010; Brandani et al. 2018; Pires et al. 2018; Souza et al. 2021), without extracting and analyzing the genotypic values of these populations.

The general purpose of this work is to subsidize forest restoration projects in Brazil with G. americana, selecting seed trees based on the species genetic divergence in fruits, seeds, and seedlings. Thus, the addressed hypothesis is that the G. americana population in the Saltinho Biological Reserve is a good source of reproductive material for the species.

2.1 Sampling area

The study was conducted in the Saltinho Biological Reserve, with 548 hectares, located in Tamandaré - Pernambuco. The region's climate is humid tropical (As'), with average annual precipitation of 1,500 mm, an average annual temperature of 25°C, and an altitude between 40 and 120 m (CPRH 2001; IBAMA 2003). The relief is irregular, varying from 5 to 154 m, the region's soil is classified as Red-Yellow Latosol, and the reserve is located within the Watershed of the Saltinho Stream (IBAMA 2003). The vegetation is characterized as Dense Ombrophylous Lowland Forest (Veloso et al. 1991).

The Saltinho Management Plan indicated the occurrence of at least 325 plant species belonging to 80 botanical families and 205 genera (CPRH 2011). The same document also states that the Reserve is an important source of plant material for the restoration of areas in the region, highlighting its importance for the Zona da Mata of Pernambuco.

2.2 Data collection

The material collection was carried out in April 2021, through a sampling of G. americana in the Reserve, which was done by walking within the forest, and collecting data from all trees of the species found. The data on circumference at breast height (CBH = 1.30 m), total height, and collection of available fruit were obtained (Solís-Guillén et al. 2017), as well as the geographical coordinates with the aid of a Garmin GPS 64s, totaling 11 seed trees.

The study was conducted in the Forest Seed Analysis Laboratory (LASF) of the Forest Science Department (DCFL), of the Federal University of Rural Pernambuco (UFRPE), where they were separated and identified in batches by seed trees.

2.3 Physical and physicochemical characterization of the fruits

After processing the fruits, and discarding the green and overripe ones, they were stored for 16 days in a refrigerator (± 4 ºC), with the number of fruits varying from 8 to 20 fruits per seed tree, depending on availability. The physical analyses were performed at the LASF, and the physicochemical analyses at the Food Physicochemical Analysis Laboratory of the Domestic Sciences Department, at UFRPE.

The evaluations performed followed the methodologies of Instituto Adolfo Lutz (IAL) (2008); Braga Filho et al. (2014); Associação Oficial de Químicos Analíticos (AOAC) (2016); and Carvalho et al. (2017).

2.4 Biometric characterization of the seeds

After seed processing, pictures of 6 to 25 seeds were obtained from each fruit, randomly selected from a total of 200 fruits, fitting seeds from five fruits in each picture, all identified by seed tree. For the photos, an iPhone 6s celular phone camera and a light-controlled photo studio were used to avoid shadow interference in the images as much as possible.

In the software RStudio (R CORE TEAM 2020), where the photo processing and data extraction were performed, the packages ExpImage (Azevedo 2021) and EBImage (Oles et al. 2020) were used. The variables obtained were area (cm²), perimeter, medium radius, the standard deviation of the radius, minimum radius, maximum radius, principal axis, and eccentricity, all in cm (Fig. 1).

2.5 Analysis of seed emergence

The same seeds used for the biometric characterization were sown in a nursery, each pot representing one fruit with all respective seeds. The pots were randomly distributed, each corresponding to fruit from a seed tree, for a total of 200 pots. The substrate comprised 60% soil, 25% rice husk, and 15% organic compost in 3 L pots.

The emergence process was monitored daily from day one until the emergence rate stabilized, for a total period of 110 days. The seeds that were considered to have emerged were those that emitted cotyledons.

2.6 Statistical analysis of genetic parameters, divergence and genetic selection

The genotypic values and genetic parameters were obtained using the mixed models' methodology for the diversity studies on fruits, seeds, and seedling emergence (Resende 2002). They were submitted to a completely randomized design, with testing of unrelated genotypes, with the repetitions varying from 8 to 20 fruits per tree, and 6 to a maximum of 25 seeds per fruit, depending on the availability of fruits and seeds suitable for analysis.

The model used was 81 from Selegen REML/BLUP software (Resende 2016), as follows:

$$y=Xu+Zg+Wp+e$$

where: y is the data vector, u is the general mean effect (fixed), g is the genotypic effects vector (random), p is the progeny effects vector (random), and e is the errors or residuals vector (random). The capital letters represent the incidence matrices for these effects. X, Z, and W = incidence matrices for u, g, and p, respectively.

Pearson's genetic correlation was obtained with the genotypic values (µ + g), and Pearson's genetic correlation with p < 0.01 and p < 0.05 was considered significant.

We also performed the Maximum Likelihood Ratio Test and Deviance Analysis, obtained as described by Resende (2007), using the model with and without the respective effects. In addition, cluster analysis was performed using the UPGMA (Unweighted Pair Group Method with Arithmetic Mean) method, using the Mahalanobis genetic distance as a measure of dissimilarity (Mahalanobis 1936; Lima 2006) and Principal Component Analysis (PCA) (Kassambra 2017). The variables obtained in the study for all analyses are shown in Table 1:

Table 1

List of variables for characterization of fruits and seeds of *Genipa americana* L. from the Saltinho Biological Reserve - PE, Brazil.
Variable	Acronym
Seed area (cm)	AreaS
Seed perimeter (cm)	PerS
Medium seed radius (cm)	RadMedS
Minimum seed radius (cm)	RadMinS
Maximum seed radius (cm)	RadMaxS
Seed radius standard deviation (cm)	RadSdS
Principal seed axis (cm)	AxPrinS
Seed eccentricity (cm)	EccenS
Fruit weight (g)	WeightF
Fruit longitudinal and transversal diameter (mm)	DiLongF; DiTransF
Peel and pulp weight (g)	PeelPulpF
Peel and pulp yield (%)	YieldF
Peel and pulp thickness (mm)	ThickF
Number of seeds per fruit	NumbSeedF
Fruit soluble solids (°Brix)	SolSol
Fruit titratable acidity (%)	ATT
Fruit hydrogen potential	pH
Percentage of emergence (%)	E%
Emergence index speed	ISE
Medium emergency time	TME
Emergence medium speed	SME

The software Selegen REML/BLUP was used to obtain variance components, genotypic values, genetic correlations, and genetic grouping analysis (Resende 2016). Genetic principal component analysis (PCA) and graphical representation of the results were performed with a RStudio software (R CORE TEAM 2021) and aided by the packages ggplot2 (Wickham 2016), corrplot (Wei and Simko 2017), and factoextra (Kassambara and Mundt 2020).

2.7 Seed tree selection index

The seed tree selection index was created with the seed trees genotypic values in the variables chosen for ranking, and processed in the Selegen REML/BLUP software (Resende 2016). The average rank index (Resende 2016) adapted from Mulamba and Mock (1978) was performed.

The variables that were not significant in the components of genetic variance, both in total genotypic effects (χ²h²) and in matrix effects (χ²c²), were discarded from the selection index: DiLongF, PeelPulpF, and TME.

In addition, based on the correlation matrix, the variables PerS, RadMedS, AxPrinS, DiTransF, SeedMucF, and ISE were also dropped from the index because they were more than 80% correlated with the others. Thus, 11 variables remained to compose the index: AreaS, EccenS, WeightF, YieldF, ThickF, NumbSeedF, SolSol, pH, ATT, E% and SME. The classification of this ranking was given with the best seed trees having the highest value for all variables, except for YieldF, ThickF, and ATT, which were considered the lowest values.

3.1 Variables analyzed in G. americana fruits and seeds

Regarding seed biometric variables, seed tree 6 obtained higher values in six variables (AreaS, PerS, RadMedS, AxPrinS, RadMinS, and RadMaxS). As for the fruit morphometric variables, seed tree 1 obtained higher values in three variables (DiTransF, WeightF, and SeedMucF). The averages of the fruit chemical variables and the emergence indices did not show any seed tree with higher values in more than one variable (Table 2).

Table 2

Mean of biometric variables of seeds and fruits, chemical variables of fruits, and emergence indexes of *Genipa americana* L. seedlings from the Saltinho Biological Reserve, in Tamandaré - PE, Brazil.
Seed tree	AreaS	PerS	RadMedS	AxPrinS	EccenS	RadSdS	RadMinS	RadMaxS
1	0.410	2.259	0.363	0.859	0.658	0.047	0.282	0.456
2	0.529	2.548	0.413	0.958	0.620	0.049	0.327	0.509
3	0.411	2.239	0.364	0.853	0.635	0.045	0.287	0.450
4	0.541	2.607	0.419	0.978	0.634	0.051	0.330	0.523
5	0.476	2.370	0.391	0.889	0.578	0.039	0.323	0.468
6	0.627	2.886	0.451	1.054	0.631	0.056	0.352	0.562
7	0.478	2.458	0.392	0.918	0.636	0.049	0.306	0.488
8	0.573	2.725	0.430	1.026	0.668	0.058	0.329	0.542
9	0.569	2.657	0.428	0.985	0.605	0.048	0.343	0.526
10	0.525	2.558	0.412	0.968	0.647	0.051	0.325	0.513
11	0.478	2.494	0.392	0.905	0.614	0.045	0.313	0.483
Seed tree	DiTransF	DiLongF	WeightF	PeelPulpF	SeedMucF	YieldF	ThickF	NumbSeedF
1	95.656	72.212	243.350	144.714	91.767	0.601	12.579	169.182
2	75.319	68.274	167.689	122.865	31.054	0.737	12.849	128.286
3	76.000	67.585	171.176	135.209	32.203	0.795	13.261	94.538
4	88.518	71.663	210.791	135.364	49.618	0.681	13.052	222.600
5	79.753	70.033	203.754	141.889	42.456	0.674	13.235	112.600
6	74.231	70.725	186.585	140.970	44.060	0.728	11.347	109.900
7	87.529	70.237	202.559	154.791	45.328	0.759	17.486	126.800
8	88.047	70.608	224.846	161.216	50.368	0.720	12.503	172.100
9	78.561	71.556	210.902	153.027	53.039	0.730	12.604	79.400
10	72.044	69.160	181.825	154.891	57.253	0.755	16.306	132.875
11	88.557	74.433	219.560	159.005	59.595	0.727	15.034	104.500
Seed tree	SolSol	pH	ATT	E%	ISE	TME	SME
1	10.047	3.757	1.273	60.053	0.496	31.961	0.029
2	11.587	3.471	2.264	40.839	0.342	40.181	0.027
3	12.621	3.458	1.706	53.335	0.448	39.033	0.028
4	9.613	3.493	1.731	54.413	0.437	40.460	0.027
5	10.313	3.429	2.047	25.033	0.258	43.295	0.025
6	8.753	3.627	1.799	60.849	0.443	39.625	0.027
7	10.687	3.505	1.368	54.188	0.416	38.519	0.029
8	11.707	3.306	2.096	33.718	0.215	44.379	0.024
9	11.238	3.430	1.288	31.816	0.271	39.093	0.027
10	9.663	3.493	1.457	53.764	0.401	40.051	0.027
11	8.325	3.505	1.457	47.600	0.389	32.480	0.032
AreaS: seed area; PerS: seed perimeter; RadMedS: average seed radius; RadSdS: seed radius standard deviation; RadMinS: minimum seed radius; RadMaxS: maximum seed radius; AxPrinS: main seed axis; ExcenS: seed eccentricity; DiTransF: transverse fruit diameter; DiLongF: longitudinal fruit diameter; WeightF: total fruit weight; PeelPulpF: peel and pulp weight; SeedMucF: fruit seed and mucilage weight; YieldF: fruit shell and pulp yield; ThickF: fruit shell and pulp thickness; NumbSeedF: amount of seeds per fruit; SolSol: soluble solids; pH: hydrogen potential; ATT: total titratable acidity; E%: percentage of emergence; ISE: emergence speed index; TME: medium time of emergence; SME: medium speed of emergence.

3.2 Genetic parameters in G. americana fruits, seeds, and seedlings

Table 3 shows the variance components of all variables. All characters had significance for total effects (χ²h²). They had genetic differences, except for DiLongF, PeelPulpF, and TME. The traits with genetic differences among the matrices (x²c²) were AreaS, PerS, RadMedS, RadSdS, RadMinS, RadMaxS, AxPrinS, and SolSol.

Resende (2015) classifies the heritability into: low (< 0.15), intermediate (0.15 to 0.50), and high (≥ 0.50). The variable with the highest adjusted individual heritability (h²aj) for matrix effect was the number of seeds per fruit (NumbSeedF), with 0.45, as well as the individual heritability in the broad sense (h²g). This variable, considering the total genotypic effects, is classified as intermediate.

Table 3

Components of variance of seed and fruit biometric variables, fruit chemical variables, and seedling emergence rates of *Genipa americana* L., from the Saltinho Biological Reserve, in Tamandaré - PE, Brazil.
	AreaS	PerS	RadMedS	RadSdS	RadMinS	RadMaxS	AxPrinS	EccenS
σ²g	0.005	0.038	0.001	0.000	0.000	0.001	0.004	0.001
σ²parc	0.001	0.008	0.000	0.000	0.000	0.000	0.001	0.000
σ²e	0.011	0.102	0.001	0.000	0.001	0.004	0.015	0.018
σ²f	0.017	0.149	0.003	0.000	0.003	0.006	0.020	0.018
h²g	0.265 ± 0.021	0.256 ± 0.020	0.267 ± 0.021	0.067 ± 0.010	0.182 ± 0.017	0.214 ± 0.019	0.216 ± 0.019	0.032 ± 0.007
h²aj	0.283	0.271	0.285	0.067	0.191	0.221	0.224	0.033
c²parc	0.065	0.055	0.063	0.006	0.042	0.033	0.035	0.001
CVgi%	13.270	7.738	6.804	10.888	6.703	7.126	6.885	3.888
CVe%	7.797	4.400	3.945	8.727	4.235	3.871	3.765	4.301
Accuracy	0.950	0.950	0.950	0.937	0.948	0.949	0.949	0.921
Average	0.508	2.521	0.404	0.049	0.319	0.501	0.943	0.630
χ²h²	1367.62*	1300.62*	1380.25*	282.64*	843.13*	1030.84*	1044.25*	128.79*
χ²c²	162.96*	159.97*	166.38*	9.10*	93.04*	97.38*	100,32*	1.07^ns
	DiTransF	DiLongF	WeightF	PeelPulpF	SeedMucF	YieldF	ThickF	NumbSeedF
σ²g	52.991	0.952	413.394	78.233	243.298	0.001	2.848	1537.708
σ²parc	0.631	2.363	17.735	25.449	2.693	0.000	0.032	11.756
σ²e	130.815	75.126	3853.627	1675.734	402.951	0.007	8.465	1855.843
σ²f	184.437	78.441	4284.755	1779.416	648.942	0.010	11.346	3405.307
h²g	0.287 ± 0,094	0.012 ± 0.019	0.096 ± 0.055	0.044 ± 0.038	0.375 ± 0.110	0.257 ± 0.091	0.251 ± 0.090	0.451 ± 0.176
h²aj	0.288	0.013	0.097	0.045	0.376	0.265	0.252	0.453
c²parc	0.003	0.030	0.004	0.014	0.004	0.031	0.003	0.003
CVgi%	8.845	1.383	10.047	6.070	30.748	6.927	12.376	29.929
CVe%	9.874	8.954	21.791	20.165	28.167	8.495	15.145	23.396
Accuracy	0.904	0.450	0.800	0.671	0.916	0.894	0.890	0.902
Average	82.303	70.571	202.367	145.708	50.729	0.718	13.636	131.024
χ²h²	57.62*	0.41^ns	13.45*	3.82^ns	93.46*	51.68*	42.23*	41.29*
χ²c²	0.02^ns	0.87^ns	0.01^ns	0.19^ns	0.03^ns	0.87^ns	0.01^ns	0.00^ns
	SolSol	pH	ATT	E%	ISE	TME	SME
σ²g	1.879	0.013	0.102	127.435	0.006	6.580	0.000
σ²parc	0.570	0.000	0.000	1.837	0.000	0.274	0.000
σ²e	3.004	0.028	0.157	686.622	0.058	126.763	0.000
σ²f	5.452	0.041	0.259	815.895	0.065	133.617	0.000
h²g	0.345 ± 0.134	0.324 ± 0.123	0.392 ± 0.143	0.156 ± 0.079	0.097 ± 0.065	0.049 ± 0.046	0.066 ± 0.053
h²aj	0.385	0.325	0.393	0.157	0.097	0.049	0.066
c²parc	0.104	0.004	0.001	0.001	0.001	0.001	0.001
CVgi%	13.111	3.274	18.866	23.497	20.875	6.584	6.744
CVe%	18.081	4.726	23.495	54.613	63.664	28.928	25.364
Accuracy	0.899	0.888	0.903	0.832	0.759	0.644	0.696
Average	10.455	3.519	1.690	48.044	0.380	38.962	0.028
χ²h²	38.05*	37.92*	49.88*	17.09*	7.31*	2.59^ns	4.16*
χ²c²	5.50*	0.01^ns	0.00^ns	0.00^ns	0.00^ns	0.00^ns	0.00^ns
AreaS: seed area; PerS: seed perimeter; RadMedS: average seed radius; RadSdS: seed radius standard deviation; RadMinS: minimum seed radius; RadMaxS: maximum seed radius; AxPrinS: main seed axis; EccenS: seed eccentricity; DiTransF: transverse fruit diameter; DiLongF: longitudinal fruit diameter; WeightF: total fruit weight; PeelPulpF: fruit shell and pulp weight; SeedMucF: fruit seed and mucilage weight; YieldF: fruit shell and pulp yield; ThickF: fruit shell and pulp thickness; NumbSeedF: amount of seeds per fruit; SolSol: soluble solids; pH: hydrogen potential; ATT: total titratable acidity; E%: percentage of emergence; ISE: emergence speed index; TME: medium time of emergence; SME: medium speed of emergence; σ²g: genotypic variance; σ²parc: genotypic variance between matrices; σ²e: residual variance; σ²f: individual phenotypic variance; h²g: individual heritability in the broad sense, that is, of the total genotypic effects; h²aj: genotypic heritability adjusted for matrix effects; c²parc: coefficient of determination of matrix effects; CVgi%: genotypic variation coefficient; CVe%: residual variation coefficient; χ²h²: chi-square obtained from total genotypic effects; χ²c² : chi-square obtained from matrix effects; *: significant at 5%; ^ns: not significant at 5%.

The variables that had intermediate individual heritability were: seed biometrics - area (AreaS), perimeter (PerS), medium radius (RadMedS), minimum radius (RadMinS), maximum radius (RadMaxS), principal axis (AxPrinS); fruit biometrics - transverse diameter (DiTransF), weight of seeds with mucilage (SeedMucF), peel and pulp yield (YieldF), peel and pulp thickness (ThickF), number of seeds per fruit (NumbSeedF); fruit chemistry - soluble solids (SolSol), hydrogen potential (pH), total titratable acidity (ATT); emergence indices - emergence percentage (E%).

On the other hand, those with low individual heritability were: seed biometrics – seed eccentricity (EccenS), standard deviation of radius (RadSdS); fruit biometrics - longitudinal diameter (DiLongF), total fruit weight (WeightF), peel and pulp weight (PeelPulpF); emergence indices - emergence speed index (ISE), mean time to emergence (TME) and mean speed to emergence (SME).

The high accuracy values observed indicate that the experiment was adequate and reliable for evaluating the traits (Resende and Duarte 2007), and extracting their genotypic values, except for the variables DiLongF, PeelPulpF, TME, and SME, which had accuracy below 70%. The variables with the lowest c²parc were RadSdS, EccenS, DiTransF, WeightF, SeedMucF, ThickF, NumbSeedF, pH, ATT, E%, ISE, TME, and SME, with c²parc values lower than 0.01.

According to Sebbenn et al. (1998), a genetic variation coefficient (CVgi%) higher than 7.0% is considered high. Variables AreaS, PerS, RadSdS, RadMaxS, DiTransF, WeightF, SeedMucF, ThickF, NumbSeedF, SolSol, ATT, E% and ISE have high CVgi%. The highest values were 30.7%, 29.9% and 23.4% for the variables SeedMucF, NumbSeedF, and E%, respectively.

3.3 Multivariate analysis in fruits, seeds, and seedlings of G. americana

In the correlation matrix of the variables, the correlation of ATT was low and negative with pH (-0.65), demonstrating the low influence of ATT on pH (Fig. 2). YieldF, on the other hand, had a low and negative correlation with the variables DiTransF, WeightF, SeedMucF, and NumbSeedF (-0.64, -0.72, -0.73, and − 0.58, respectively), indicating that the yield of the skin and pulp of G. americana fruit is inversely proportional to these four fruit-related variables.

In the cluster analysis using the UPGMA method, it was observed that the presence of four groups of seed trees with genetic similarity from the genotypic values indicated high genetic variability, as follows: group 1: trees 11, 5, and 7; group 2: trees 1 and 3; group 3: tree 6; and group 4: trees 8, 9, 2, 4 and 10 (Fig. 3).

3.4 Selection index of G. americana seed trees

Seed trees 1, 4, and 6 obtained the best average ranks concerning the genotypes, respectively (Table 4). This means that they are the best genitors for fruit and seeds' biometric, morphological, and chemical variables. These three individuals have an average distance of 585.0 m between them, and all are present within the fragment. The worst performing seed trees were 5, 2, and 3, respectively. These three individuals have an average distance of 1,003.6 m between them.

Table 4

Sorting of the 11 seed trees of *Genipa americana* L. from the Saltinho Biological Reserve - PE, Brazil.
Seed tree	AreaS	EccenS	WeightF	YieldF	ThickF	NumbSeedF	SolSol	pH	ATT	E%	SME	Ranking
1	11	2	1	1	3	3	6	1	1	1	2	1
4	4	6	5	2	6	1	9	7	7	3	8	2
6	1	7	9	5	1	8	10	3	8	2	5	3
8	3	1	2	4	2	2	4	11	10	9	11	4
7	8	4	7	10	11	5	5	5	2	4	3	5
9	2	10	4	7	4	11	2	2	3	10	9	6
11	7	9	3	6	9	9	11	4	4	7	1	7
10	6	3	8	9	10	4	8	6	5	6	7	8
3	10	5	10	11	7	10	1	9	6	5	4	9
2	5	8	11	8	5	6	3	8	11	8	6	10
5	9	11	6	3	8	7	7	10	9	11	10	11
AreaS: seed area; EccenS: seed eccentricity; WeightF: fruit total weight; YieldF: peel and pulp yield; ThickF: peel and pulp thickness; NumbSeedF: number of seeds per fruit; SolSol: soluble solids; pH: hydrogen potential; ATT: total titratable acidity; E%: emergence percentage; SME: emergence average speed.

Seed tree 1 has better fruit-related characters such as weight, pH, ATT, and E%. Retrieving the values of the correlations of the variables in Figure 6, fruit weight (WeightF) is highly correlated (0.84) with seed weight with mucilage (SeedMucF), just as E% is highly correlated (0.93) with ISE. However, pH and ATT have no significant correlation, with at least 70% positive or negative correlation, with any other character.

Variables related to fruits and seeds can vary between plants of the same species, reproductive years, or even within the same plant (Lovatel et al. 2021). Seed size is one factor influencing seedling emergence through the quantity of reserves present in the seed (Pádua et al. 2010). Furthermore, fruit biometric variables also help in the plant breeding of several species, quantifying phenotypic divergence through quantitative and qualitative descriptors (Abud et al. 2018), and it is also possible to extract genotypic values from these attributes. Therefore, the genetic variability within and between tree populations can be characterized by fruit and seed biometry, and can also be related to environmental factors, helping conservation and genetic improvement programs of species (Silveira et al. 2019).

Genotypic variances (σ²g) smaller than the residual variances (σ²e), in all variables, result in low or intermediate levels of heritability. In other words, the effect of the environment (residual variance) was predominant in the genotypic effect of the progenies. Heritability is the parameter that determines phenotypic traits that well represent the genetic gain of progenies. The amount of phenotypic plasticity is a consequence of the genotype, and it is a reliable guide to use in genetic selection (Moraes et al. 2015; Silva et al. 2018).

The coefficients of determination of the matrix effects (c²parc) were low because there is a significant interaction between the sites in which each one is inserted (area, soil, pollen source, etc.), resulting in the high genetic variability of the progenies (Santos 2021). This result is expected because, in allogamous species, it is common for genetic diversity to be greater within than between populations, or in this case, seed trees (Golunski et al. 2015; Siqueira et al. 2021).

Seed biometry is an efficient criterion to verify genetic divergence among individuals (Lovatel et al. 2021). It should also be noted that the heritability coefficient, in the strict sense, reflects how much of the phenotypic variation, observed in a quantitative character, is heritable due to the additive gene effects transferred from parental individuals to their descendants (Costa et al. 2010). Due to lack of information on family structure, in this work, the heritability coefficient in the strict sense could not be calculated.

The collection of reproductive material from seed trees with better emergence rates and fruits with attractive characteristics, helps in the genetic improvement of forest restoration projects. The genetic quality of the sources of material is essential for the establishment, rapid growth, and sustainability of tree populations (Jansson et al. 2016; Mazhula et al. 2021). Kavaliauskas et al. (2021) recommends that seed collection evaluation should depend on attributes such as the number of viable and reproductive individuals, CBH, and height. They also highlight the importance of selecting material sources based on phenotypic characteristics and genetic diversity criteria.

Also, Kijowska-Oberc et al. (2020) brings up the issue of proper conservation of seed stocks in gene banks since phenotypic plasticity plays an essential role in plant adaptation to environmental changes. Reserves with greater genetic diversity, with populations grown in different locations, have a better chance of surviving under new climatic conditions. Given the importance of genotype selection and the rapid development of phenotyping technology, it is becoming increasingly important to know how additional phenotypic traits can improve the prediction accuracy for a target trait (Arouisse et al. 2021).

However, since forest restoration projects rarely consider the genetic diversity associated with seed quality for their development, there is a lack of market demand for native species that meet these needs. Thus, expanding knowledge about the diversity of high-quality native seed species is critical (Atkinson et al. 2021).

The G. americana population of the Saltinho Biological Reserve is a good source of reproductive material for the species since seed trees 1, 4, and 6, respectively, were genetically satisfactory to be designated for this purpose. However, it should be noted that it is not recommended to select only three seed trees for a restoration or conservation project with allogamous and dioecious species. Considering the assumptions of Nunes (2021), based on a mean of 441 seeds and a variance of 1,458, this work has an effective population size of 12 and a potential inbreeding for the restoration project of 4.16%. For this to occur, it is necessary, for the restoration project, for seedlings to be formed from all these seeds, implanted, and conducted, without any mortality.

Melo et al. (2021) found a low effective population size for G. americana of 2.69 and 3.27 in a natural population (NP) of Sergipe and progeny test (PT), respectively, due to the high outcrossing rate between related trees. The number of alleles for both populations ranged from 8 to 23, with 96 alleles and a mean of 13.7. Of these, 19 alleles were deprived of the NP population and 15 from the PT population. Therefore, the number of seed trees needed for the collection was 56 in NP and 46 in PT. From this, the authors recommend that the same number of male and female trees should be maintained in each progeny to maximize the effective population size of the parental population of G. americana for offspring production.

Manoel et al. (2015b) obtained a mean effective population size of 2.5 and a mean number of alleles of 23.5 in a population of G. americana in São Paulo. In this work, it was recommended to collect seeds from at least 60 seed trees. Ancestry must be reduced and effective sample variance increased for ex situ conservation and forest restoration. Collecting seeds from multiple trees helps achieve these parameters.

Through the analysis of genetic parameters, it was possible to define a more significant genetic divergence between progenies than between seed trees, an expected result in allogamous species. In addition, there was a considerable influence of the environment on phenotypic attributes, indicating that each individual's location within the reserve is vital in determining the phenotypic plasticity of the species.

Therefore, through the seed tree selection index, the individuals that will serve as the best sources of genetic material in the Reserve are trees 1, 4, and 6, respectively. They are in different groups in the cluster analysis by the UPGMA method, always remembering that it is necessary to consider the effective population size and potential inbreeding. Considering their genetic divergence, these seed trees obtained a better ranking in most variables. Finally, the Saltinho Biological Reserve is a good source of genetic and reproductive material for the species G. americana for forest restoration purposes.

Species should be collected in more areas, applying the same selection methodology as in this paper. Thus, the effective population size would be increased, inbreeding decreased, and a wide genetic divergence of fruit, seed, and seedling emergence characters would be achieved. A better understanding of the species’ evolutionary processes will provide a greater success rate in restoration projects.

Funding: The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Competing Interests: The authors have no relevant financial or non-financial interests to disclose.

Author Contributions: Raquel Elvira Cola, Ricardo Gallo, and Maria da Penha Moreira Gonçalves contributed to the study conception and design. Material preparation, data collection and analysis were performed by Raquel Elvira Cola, Lidiana Nayara Ralph, and Larissa Santiago Ritt Ordonho. The first draft of the manuscript was written by Raquel Elvira Cola and Ricardo Gallo, Maria da Penha Moreira Gonçalves, and Maria Inês Sucupira Maciel commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Data Availability: The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request

Acknowledgments

To Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) and to the Forest Seeds Analysis Laboratory and the Laboratory of Physical-Chemical Food Analysis at UFRPE.

Abud HF, Araujo RF, Pinto CMF, Araujo EF, Araujo AV, Santos JA (2018) Rev Bras Agropecuária Sustentável 8:29–39CARATERIZAÇÃO MORFOMÉTRICA DOS FRUTOS DE PIMENTAS MALAGUETA E BIQUINHO
Agência Estadual de Meio Ambiente (CPRH) (2011) Área de Proteção Ambiental de Guadalupe - ENCARTE, vol 3. Análise da Unidade de Conservação. Recife
Agência Estadual de Meio Ambiente (CPRH) (2001) Diagnóstico socioambiental:Litoral sul de Pernambuco. Recife
AOAC (2016) Official methods of analysis of AOAC international
Arouisse B, Theeuwen TPJM, van Eeuwijk FA, Kruijer W (2021) Improving Genomic Prediction Using High-Dimensional Secondary Phenotypes. Front Genet 12:715. https://doi.org/10.3389/FGENE.2021.667358/BIBTEX
Atkinson RJ, Thomas E, Roscioli F, Cornelius JP, Zamora-Cristales R, Franco Chuaire M, Alcázar C, Mesén F, Lopez H, Ipinza R, Donoso PJ, Gallo L, Nieto V, Ugarte J, Sáenz-Romero C, Fremout T, Jalonen R, Gaisberger H, Vinceti B, Valette M, Bosshard E, Ekué M, Wiederkehr Guerra G, Kettle C (2021) Seeding Resilient Restoration: An Indicator System for the Analysis of Tree Seed Systems. Diversity 13:367. https://doi.org/10.3390/D13080367/S1
Azevedo AM (2021) ExpImage-package: ExpImage: Experimental Image Analysis Tools in ExpImage. Tool For Analysis of Images in Experiments
Braga Filho JR, Naves RV, Chaves LJ, Pires LL, Mazon LT (2014) Caracterização física e físico-química de frutos de araticum (Annona crassiflora Mart. Biosci J 30:16–24
Brandani JZ, Junglos MS, Santiago EF, Scalon SPQ, Mussury RM (2018) Are seeds of Genipa americana L. (Rubiaceae) tolerance to water submersion? Floresta e Ambient 25:20170764. https://doi.org/10.1590/2179-8087.170764
Carvalho AV, Chaves RPF, Alves RM (2017) Caracterização física e físico-química de frutos em matrizes de cajazeira no Estado do Pará. Belém. Embrapa Amazônia Oriental
Chazdon R, Brancalion P (2019) Restoring forests as a means to many ends. Science (80-.). 364:24–25. https://doi.org/10.1126/SCIENCE.AAX9539
Correia LAS, Felix FC, Araújo FS, Ferrari CS, Pacheco MV (2019) Morphometric descriptors and physiological seed quality for selecting Aspidosperma pyrifolium Mart. matrix trees. Rev Caatinga 32:751–759. https://doi.org/10.1590/1983-21252019V32N319RC
Costa RB, Resende MDV, Gonçalves PS, Roa RAR, Feitosa KCO (2010) Predição de parâmetros e valores genéticos para caracteres de crescimento e produção de látex em progênies de seringueira. Bragantia 69:49–56. https://doi.org/10.1590/S0006-87052010000100007
Cunha FLR, Bernhard R, Vogt RC(2020) Diet of an Assemblage of Four Species of Turtles (Podocnemis) in the Rio Uatumã, Amazonas, Brazil. https://doi.org/10.1643/CE-18-117 108, 103–115. https://doi.org/10.1643/CE-18-117
Fundação SOS Mata Atlântica, 2020. Relatório anual 2020
Fundação SOSM, Atlântica, Instituto Nacional de Pesquisas Espaciais, 2020. Atlas dos remanescentes florestais da Mata Atlântica: Período 2018–2019. São Paulo
Golunski CM, Miotto SPS, Junior CV, Corazza T, Mielniczki-Pereira AA, Mossi AJ, Budke JC, Cansian RL (2015) Diversity and genetic structure in Ocotea odorifera (Vell.) Rohwer (Lauraceae) from southern Brazil. Perspectiva 39:41–52
Holl KD, Brancalion PHS (2020) Tree planting is not a simple solution. Science (80-.). 368:580–581. https://doi.org/10.1126/SCIENCE.ABA8232/SUPPL_FILE. /ABA8232_HOLL_SM.PDF
Hssaini L, Hanine H, Razouk R, Ennahli S, Mekaoui A, Guirrou I, Charafi J (2020) Diversity Screening of Fig (Ficus Carica L.) Germplasm through Integration of Morpho-agronomic and Biochemical Traits. Int J Fruit Sci 20:939–958. https://doi.org/10.1080/15538362.2019.1700871
IAL (2008) Métodos físico-químicos para análise de alimentos. Brasília
IBAMA (2003) Resumo executivo do plano de manejo da Reserva Biológica de Saltinho. Brasília
Jansson G, Hansen JK, Haapanen M, Kvaalen H, Steffenrem A (2016) The genetic and economic gains from forest tree breeding programmes in Scandinavia and Finland. Scand J For Res 32:273–286. https://doi.org/10.1080/02827581.2016.1242770
Kassambara A, Mundt F(2020) Factoextra:Extract and Visualize the Results of Multivariate Data Analyses
Kassambra A(2017) Practical Guide To Principal Component Methods in R, 1st ed
Kavaliauskas D, Šeho M, Baier R, Fussi B (2021) Genetic variability to assist in the delineation of provenance regions and selection of seed stands and gene conservation units of wild service tree (Sorbus torminalis (L.) Crantz) in southern Germany. Eur J For Res 140:551–565. https://doi.org/10.1007/S10342-020-01352-X
Kijowska-Oberc J, Staszak AM, Kamiński J, Ratajczak E(2020) Adaptation of Forest Trees to Rapidly Changing Climate. Forests 11, 123. https://doi.org/10.3390/F11020123
Lima EL(2006) Álgebra linear
Lima MAO, Mielke MS, Lavinsky AO, França S, Almeida A-AF, Gomes FP (2010) Crescimento e plasticidade fenotípica de três espécies arbóreas com uso potencial em sistemas agroflorestais Growth and phenotypic plasticity of three woody species with potential use in agroforestry systems. Sci For 38:527–534
Lovatel QC, Navroski MC, Gerber TR, de Oliveira LM, Pereira M, de Silveira O MF (2021) Genetic variability in juvenile characters of progenies of Apuleia leiocarpa. Floresta 51:547–556. https://doi.org/10.5380/rf.v51
Madrera RR, Negrillo AC, Valles BS, Fernández JJF (2020) Characterization of extractable phenolic profile of common bean seeds (Phaseolus vulgaris L.) in a Spanish diversity panel. Food Res Int 138:109713. https://doi.org/10.1016/j.foodres.2020.109713
Mahalanobis PC (1936) On the generalized distance in statistics. in: NATIONAL INSTITUTE OF SCIENCE OF INDIA, pp 49–55
Manoel RO, Freitas MLM, Barreto MA, Moraes MLT, Souza AP, Sebbenn AM (2014) Development and characterization of 32 microsatellite loci in Genipa americana (Rubiaceae). Appl Plant Sci 2:1300084. https://doi.org/10.3732/apps.1300084
Manoel RO, Freitas MLM, Júnior EF, Alves PF, Moraes MLT, Sebbenn AM (2015a) Individual, fruit, and annual variation in correlated mating in a Genipa americana population. Silvae Genet 64:108–116. https://doi.org/10.1515/sg-2015-0010
Manoel RO, Freitas MLM, Tambarussi EV, Cambuim J, Moraes MLT, Sebbenn AM (2015b) Mendelian inheritance, genetic linkage, and genotypic disequilibrium at microsatellite loci in Genipa americana L. (Rubiaceae). Genet Mol Res 14:8161–8169. https://doi.org/10.4238/2015.July.27.4
Mazhula O, Fuchylo Y, Hayda Y, Kharytonov M, Matkovcka S (2021) Progeny testing of Pinus sylvestris L. of seed orchard in different environmental conditions. Ecol Quest 33:1–14. https://doi.org/10.12775/EQ.2022.002
Melo MFV, Sebbenn AM, Rossini BC, Muniz AVCS, Rodrigues CJ, Marino CL, Moraes MLT (2021) Estimating genetic diversity, mating system and pollen dispersal to inform ex situ conservation of the tree Genipa americana L. Plant Genet Resour 19:9–19. https://doi.org/10.1017/S1479262121000022
Moraes CB, Carvalho EV, Zimback L, Luz OSL, Pieroni GB, Mori ES, Leal TCAB (2015) Variabilidade genética em progênies de meios-irmãos de eucaliptos para tolerância ao frio. Rev Árvore 39:1047–1054. https://doi.org/10.1590/0100-67622015000600007
Oles A, Pau G, Smith M, Sklyar O, Huber W(2020) EBImage: Image processing and analysis toolbox for R version 4.32.0 from Bioconductor
Pádua GP, Zito RK, Arantes NE, Neto JBF (2010) Influência do tamanho da semente na qualidade fisiológica e na produtividade da cultura da soja. Rev Bras Sementes 32:9–16. https://doi.org/10.1590/S0101-31222010000300001
Peng Y, Wang G, Cao F, Fu FF (2020) Collection and evaluation of thirty-seven pomegranate germplasm resources. Appl Biol Chem 63:15. https://doi.org/10.1186/s13765-020-00497-y
Pires HRA, Franco AC, Piedade MTF, Scudeller VV, Kruijt B, Ferreira CS (2018) Flood tolerance in two tree species that inhabit both the Amazonian floodplain and the dry Cerrado savanna of Brazil. AoB Plants 10:ply065. https://doi.org/10.1093/aobpla/ply065
Potter KM, Jetton RM, Bower A, Jacobs DF, Man G, Hipkins VD, Westwood M (2017) Banking on the future: progress, challenges and opportunities for the genetic conservation of forest trees. New For 48:153–180. https://doi.org/10.1007/S11056-017-9582-8
Core Team R(2020) R: A language and environment for statistical computing
Rabbani ARC, Silva-Mann R, Ferreira RA (2012) Variabilidade genética de Genipa americana L. pertencente ao baixo curso do Rio São Francisco. Rev Arvore 36:401–409. https://doi.org/10.1590/S0100-67622012000300002
Ragusa-Netto J (2015) Ecologia alimentar do aracuã-do-pantanal (Ortalis canicollis) em uma floresta ripária no Pantanal Sul. Brazilian J Biol 75:49–57. https://doi.org/10.1590/1519-6984.07113
Resende MD (2015) Genética Quantitativa e de Populações, 1st edn. Suprema, Viçosa
Resende MD (2007) Matemática e estatística na análise de experimentos e no melhoramento genético. Embrapa Florestas, Colombo
Resende MDV(2002) Genética biométrica e estatística no melhoramento de plantas perenes.Embrapa Florestas975
Resende MDV, Duarte JB (2007) Precisão e controle de qualidade em experimentos de avaliação de cultivares. Pesqui Agropecuária Trop 37:182–194
Resende MDV (2016) Software SENelegen-REML/BLUP: A useful tool for plant breeding. Crop Breed Appl Biotechnol 16:330–339
Ruzza DAC, Rossi AAB, Bispo RB, Tiago AV, Cochev JS, Rossi FS, Fernandes JM (2018) The genetic diversity and population structure of Genipa Americana (Rubiaceae) in Northern Mato Grosso. Brazil Genet Mol Res 17:gmr18017. https://doi.org/10.4238/gmr18017
Santos PCS (2021) Divergência genética em progênies de Mimosa caesalpiniifolia Benth. via análise de imagens e estresse salino em sementes. UFRPE, Recife
Sebbenn AM (2002) Número de árvores matrizes e conceitos genéticos na coleta de sementes para reflorestamentos com espécies nativas. Rev do Inst Florest 14:115–132
Sebbenn AM, Siqueira ACM, de F, Kageyama, Machado PY JAR(1998) Parâmetros genéticos na conservação da cabreúva - Myroxylon peruiferum L.F. Allemão.Sci. For.31–38
Silva AV, Silva AVC, Muniz EN, Almeida CS, Vitoria MF, Melo MFV, Ledo AS, Rabbanni ARC(2015) Genetic diversity and sex identification in Genipa americana L.Trop. Subtrop. Agroecosystems18
Silva AVC, Freire KCS, Lédo A, da S, Rabbani ARC (2014) Diversity and genetic structure of jenipapo (Genipa americana L.) Brazilian accessions. Sci Agric 71:387–393. https://doi.org/10.1590/0103-9016-2014-0038
da Silva LGC, Moreira JFL, Holanda HBB, Rocha ELB, Dias PC (2018) Evaluation of carnauba progenies and estimates of genetic parameters in the juvenile phase. Rev Caatinga 31:917–925. https://doi.org/10.1590/1983-21252018V31N414RC
Silveira GF, Aparecida A, Rossi B, Oliveira UA, Varella TL, Rossi FS, Cavallari AA, Galvanin EAS (2019) Zappi Nativa 7:138–144. https://doi.org/10.31413/NATIVA.V7I2.6554. Análise biométrica de frutos e sementes de Passiflora cristalina Vanderplank &
Siqueira MVBM, Bajay MM, Grando C, Campos JB, Toledo JAM, Domingues GT, Macrini C, Tambarussi EV, Brancalion PHS, Rodrigues RR, Pinheiro JB, Zucchi MI (2021) Genetic diversity of reintroduced tree populations of Casearia sylvestris in Atlantic forest restoration sites. For Ecol Manage 502:119703. https://doi.org/10.1016/J.FORECO.2021.119703
Solís-Guillén I, Chaires-Pacheco M, Juárez-Gómez J, O’Connor-Sánchez A, Peña-Ramírez YJ (2017) Development of an Ideotype-Based Selection Tool for Native Tropical Tree Breeding by Smallholder Planters in Mexico’s Maya Forest. Small-scale For 16:521–534. https://doi.org/10.1007/s11842-017-9368-z
Souza RR, Oliveira Paiva PD, Souza AR, Silva RR, Silva DPC, Reis MV, Paiva R (2021) Morpho-anatomical changes and antioxidant enzyme activity during the acclimatization of Genipa americana. Acta Physiol Plant 43:1–10. https://doi.org/10.1007/S11738-021-03263-9
Thomas E, Jalonen R, Loo J, Boshier D, Gallo L, Cavers S, Bordács S, Smith P, Bozzano M (2014) Genetic considerations in ecosystem restoration using native tree species. For Ecol Manage 333:66–75. https://doi.org/10.1016/j.foreco.2014.07.015
Veloso HP, Rangel Filho AL, Lima JCA (1991) Classificação da vegetação brasileira, adaptada a um sistema universal. Rio de Janeiro
Vieira FA, Gusmão E (2008) Biometria, armazenamento de sementes e emergência de plântulas de Talisia esculenta Radlk. (Sapindaceae). Ciência e Agrotecnologia 32:1073–1079. https://doi.org/10.1590/S1413-70542008000400006
Wei T, Simko V (2017) R package “corrplot”. Visualization of a Correlation Matrix
Wickham H(2016) ggplot2: Elegant Graphics for Data Analysis

No competing interests reported.

Download PDF

Journal Publication

published 26 Nov, 2023

Read the published version in Genetic Resources and Crop Evolution →

Version 1

posted

You are reading this latest preprint version

Selection of Genipa americana L. seed trees by genetic divergence in fruits, seeds, and seedlings

Status:

Journal Publication

Version 1

Abstract

Figures

1. Introduction

2. Material And Methods

2.1 Sampling area

2.2 Data collection

2.3 Physical and physicochemical characterization of the fruits

2.4 Biometric characterization of the seeds

2.5 Analysis of seed emergence

2.6 Statistical analysis of genetic parameters, divergence and genetic selection

2.7 Seed tree selection index

3. Results

3.1 Variables analyzed in G. americana fruits and seeds

3.2 Genetic parameters in G. americana fruits, seeds, and seedlings

3.3 Multivariate analysis in fruits, seeds, and seedlings of G. americana

3.4 Selection index of G. americana seed trees

4. Discussion

5. Conclusions

Statements And Declarations

References

Additional Declarations

Status:

Journal Publication

Version 1