Matrices of the population
The studied natural population of umbu trees has an average height of 5.31 m, a smaller crown diameter of 9.02 m, a larger crown diameter of 9.83 m, and a stem diameter of 30.83 cm (Fig. 3).
The matrices showed heights ranging from 4.38 to 6.45 m. Crown diameter measurements varied from 6.42 to 11.25 m for the smaller diameter and from 6.93 to 12.35 m for the larger diameter (Table 1).
Table 1
Geographical location and dendrometric information for 14 matrices of a natural population of umbuzeiros (Spondias tuberosa Arr. Câm.) located in Nossa Senhora da Glória, Sergipe: stem diameter (Deq), total height (H), major crown diameter (DMA), and minor crown diameter (DME).
Matriz | Geographical coordinates | Dendrometric information |
Latitude | Longitude | Deq (cm) | Ht (m) | DMA (m) | DME (m) |
01 | -10.224242° | -37.632607° | 53.26 | 6.25 | 12.35 | 11.25 |
02 | -10.224189° | -37.631187° | 35.88 | 6.45 | 11.10 | 9.70 |
03 | -10.224315° | -37.630318° | 18.24 | 4.38 | 6.93 | 6.42 |
04 | -10.224880° | -37.630009° | 17.83 | 5.50 | 11.10 | 10.18 |
05 | -10.224423° | -37.633086° | 21.16 | 4.67 | 7.40 | 8.20 |
06 | -10.226070° | -37.630203° | 36.06 | 5.50 | 10.00 | 9.50 |
07 | -10.226729° | -37.628901° | 32.63 | 5.00 | 8.80 | 8.20 |
08 | -10.226886° | -37.628639° | 40.40 | 6.15 | 12.20 | 10.75 |
09 | -10.226334° | -37.628160° | 28.30 | 4.83 | 12.26 | 9.00 |
10 | -10.225915° | -37.628010° | 21.96 | 4.61 | 9.76 | 9.40 |
11 | -10.225155° | -37.628593° | 25.40 | 4.55 | 8.20 | 8.20 |
12 | -01.022525° | -37.634045° | 28.50 | 5.40 | 9.86 | 8.96 |
13 | -10.226189° | -37.634709° | 28.33 | 5.08 | 7.50 | 6.80 |
14 | -10.226056° | -37.634904° | 43.60 | 5.90 | 10.10 | 9.70 |
Minimum | - | - | 17.83 | 4.38 | 6.93 | 6.42 |
Maximum | - | - | 53.26 | 6.45 | 12.35 | 11.25 |
Mean | - | - | 30.83 | 5.31 | 9.83 | 9.02 |
CV (%) | - | - | 33.20 | 12.85 | 18.87 | 15.17 |
The dendrometric measurements exhibit significant variability among the matrices. The Coefficient of Variation (CV%) indicates the relative dispersion of the data. The stem diameter has a CV of 33.20%, indicating considerable variability. The total height has a CV of 12.85%, indicating moderate variability. The crown diameters have CVs of 18.87% and 15.17%, respectively.
The values found for these variables fall within the ranges reported in the literature for individuals from Bahia, Pernambuco, Piauí, Minas Gerais, Ceará, Paraíba, and Rio Grande do Norte (Kill et al. 2016; Santos 1997; Santos et al. 1999).
The stem diameter was the variable that showed the greatest variation among matrices (17.83 to 53.26 cm; CV 33.20%). Previous studies on umbuzeiro used the base diameter measurement for the stem, making it impossible to compare with the results of this study (Santos 1997; Santos et al. 1999).
Based on dendrometric characteristics and the size of the fruits from each matrix, the presence of genetic dissimilarity is evident, estimated by the average Euclidean distance. The dissimilarity analysis resulted in the formation of three distinct groups in the population: group 1 (matrices 02, 04, 06, 09, 10, 12, and 14), group 2 (matrices 01 and 08), and group 3 (matrices 03, 05, 07, 11, and 13), as illustrated in Fig. 4. Dissimilarity between matrices ranged from 2.99 to 13.69% (range = 33.70%; CV = 55.86%). This analysis highlights phenotypic variability and suggests the existence of genetic diversity in the studied population.
The stem diameter, with the greatest variation among matrices, stands out as a critical variable for dendrometric assessments. The lack of direct comparisons in the literature emphasizes the uniqueness and importance of this measurement.
In this population, matrices 05 and 11 stand out as the most similar to each other in terms of the evaluated characteristics, while matrices 01 and 03 appear as the most divergent for dendrometric and fruit biometry characteristics, compared to the other trees.
Fruit Biometrics
The fruits from the matrices exhibited a range of mean values from 2.47 to 3.51 cm in length and from 2.25 to 2.96 cm for width. Matrix 10 showed mean values for fruit length (3.51 cm) and width (2.96 cm) above the population's average. For the commercialization of fresh fruits, selecting the larger fruits would be advantageous, considering that in this type of market, fruit size along with coloration is attractive to consumers. Consumers often associate these phenotypic characteristics with ripe and more flavorful fruits (Nunes et al. 2021).
In general, the umbu fruits from this natural population had an average length of 2.99 cm and a width of 2.57 cm (Fig. 5). These values are higher than the averages of fruits studied by Uchôa et al. (2020) sold in supermarkets in Teresina, Piauí.
The smallest values for fruit length and width were 1.77 and 1.60 cm, respectively. The largest values were 4.09 and 3.47 cm for length and width, respectively. Similar values are reported in the literature for fruits from Bahia, Paraíba, Piauí, and Rio Grande do Norte (Table 2).
Table 2
Ranges of length and width measurements for umbu fruits (Spondias tuberosa Arr. Câm.) from Bahia (BA), Paraíba (PB), and Rio Grande do Norte (RN).
Reference | Season | Procedence | Length (cm) | Width (cm) |
State | Municipality | Minimum | Maximum | | Minimum | Maximum |
Costa et al. (2004) | March/April 1998 | PB | Juazeirinho | 3.32 | 3.78 | | 2.61 | 2.88 |
Costa et al. (2015) | 2011 2012 | PB; RN | PB: Soledade, Juazeirinho, Campina Grande, Serra Branca, Picuí, Boqueirão, Caturité; RN: Currais Novos, Carnaúba dos Dantas | 2.53 | 4.35 | | 2.47 | 4.31 |
Dutra et al. (2017) | March 2016 | BA | Anagé, Belo Campo, Brumado, Caraíbas, Macarani | 3.39 | 4.15 | | 3.13 | 3.71 |
Menezes et al. (2017) | - | PB | São José de Espinharas | 2.89 | 3.35 | | 2.66 | 2.78 |
Pereira et al. (2021) | March/May 2020 | PB | São Vicente do Seridó, Queimadas, Boqueirão, Cabaceiras | 2.60 | 6.00 | | 2.30 | 6.00 |
Sergipe | April 2021; March 2022; 2023 | SE | Nossa Senhora da Glória | 2.47 | 3.51 | | 2.25 | 2.96 |
The umbu fruits are of the drupe type, ranging from rounded, ovoid to oblong (Neves and Carvalho 2005). The ratio between length and width of the fruits, close to 1, indicates that morphologically they are rounded or oval, making them more suitable for the pulp industry, as existing machines are suitable for processing fruits with these shapes (Costa et al. 2015).
The analysis of fruit circularity in this population shows that the fruits predominantly have a circular shape (Fig. 6A). Campos et al. (2018) evaluated fruit growth and stated that it is always accompanied by a variation in shape. The author classified fruits in different ripening stages, stating that fully green fruits have a less circular shape, similar to Fig. 6C, while mature fruits have a more circular shape (Fig. 6B). In addition to the shape, the fruit surface can be smooth when ripe or have four to five small protrusions in its distal portion when fully green (Fig. 6C).
The analyzed fruits from the population showed average skin color composition of 62% yellow, 23% orange, 9% olive, and 5% black (coloration present in small spots on the skins) (Fig. 7). However, according to Lima and Castricine (2019), there are genotypes that may exhibit areas of purplish coloration, generally starting from the peduncle and reaching the middle region of the fruit. In the population analyzed in Sergipe, this coloration was not observed.
The distribution of colors varied according to the ripening stage of the fruits, ranging from predominant green (olive) to orange. Santos et al. (2020) classified umbu fruits based on visual selection into ripening stages: completely green, beginning of pigmentation, predominance of yellowish, completely yellow, and yellow orange. Subsequently, the authors measured pH and soluble solids content. The pH remained stable across different stages, but the soluble solids content varied, with lower levels in green umbus, contrasting with higher values in yellow fruits.
Some fruits appeared entirely orange (100%), while others were nearly entirely yellow (99%). Fruits with a predominance of green also had yellow and/or orange in their color composition. This color combination characterizes fruits at the harvest point, referred to by local communities as "swollen" (Campos et al. 2018). Later, these fruits will progress in ripening, displaying a different color pattern, as they are climacteric fruits (Neves and Carvalho 2005), reaching maturity even after harvest (Batista 2015). According to Campos et al. (2018), fruits with a predominance of yellow and those entirely yellow are considered ripe.
Fruits from extractive systems like umbu should be harvested during a period that allows better preservation of quality for consumption, and color is recognized as an indicator of this ideal point (Lima and Castricini 2019). However, due to high genetic variability, color, and the harvest point (physiological maturity) are not always directly related.
However, color variation is directly related to fruit ripening. Therefore, colorimetry has a direct correlation with soluble solids content, which can aid in the selection of new plants. Additionally, the color of the peel and pulp is one of the most valued attributes by consumers and, as a result, it can include or exclude plants, according to Ferreira et al. (2017).
Therefore, advances in umbu exploitation depend on knowledge and technologies that provide fruits with a quality standard, considering visual and taste aspects (Lima and Castricini 2019).
Biometria dos endocarpos
While fruits are important commercially, endocarps are relevant for seedling production, as propagating the umbu tree by seeds increases the species' variability (Oliveira et al. 2018). However, it has the drawback of taking time for the plants to develop and start producing. Therefore, Fonseca et al. (2019) recommend using seeds for rootstock production, as seed-derived seedlings have an advantage in forming the xilopodium, increasing survival rates in the field (Batista 2015). This way, the characteristics of the mother plant and reproductive precocity are maintained (Oliveira et al. 2018).
Each seed is enclosed in a rigid and lignified endocarp (Fig. 8A) (Fonseca et al. 2019), constituting the dispersal unit of the species, commonly called a stone (Leite et al. 2021) or pyrene (Batista 2015).
In the population of this study, being a natural population, endocarps from the same matrix show significant variation in size (Fig. 8B). This variation among individuals is expected as propagation occurred through sexual reproduction. A considerable variability in endocarp size was also observed by Dutra et al. (2017) in samples for five municipalities in Bahia.
The overall average for the endocarps was 1.83 cm in length and 1.17 cm in width (Fig. 8C). The highest averages are for endocarps from matrix 10, with 2.20 cm in length and 1.53 cm in width, while the lowest averages are from matrix 08, with 1.42 cm in length and 1.00 cm in width, respectively.
Comparing with the measurements of endocarps from Bahia, Minas Gerais, and Rio Grande do Norte (Table 3), it is noticeable that, although they fall within the ranges, the averages in Sergipe are closer to the lower values of the range, both in length and width.
Table 3
Length and width of endocarps from umbu trees (Spondias tuberosa Arr. Câm.) originating from Bahia (BA) and Minas Gerais (MG).
Reference | Procedence | Length (cm) | Width (cm) |
Dutra et al. (2017) | BA | 1.87–2.77 | 1.14–1.35 |
Nobre et al. (2017) | MG | 2.26 | 1.66 |
Rodrigues et al. (2023) | BA | 1.32–2.11 | 0.98–1.44 |
Sergipe | SE | 1.42–2.20 | 1.00–1.53 |
Indeed, considering that larger endocarps may contain larger seeds with more stored reserves, they could be recommended for seedling production, even though small size is a desired characteristic for fruit commercialization (Oliveira et al. 2018).
Correlation of endocarp and seedling sizes
The endocarps used in the germination tests, which resulted in normal seedlings, had areas ranging from 0.91 to 2.86 cm². Based on the data range (1.95 cm²), they were grouped into three size classes (Fig. 9).
The seedlings obtained from endocarps of different sizes were analyzed through principal component analysis (PCA) and their fresh mass (g) and total size (cm) (main variables) were correlated with the size of the endocarps. The results are graphically represented in a PCA-biplot in Fig. 10, where the first dimension (Dim1) explains 52.7% of the total variance, and the second dimension (Dim2) explains 42.8%. In the graphical representation, the geometric points correspond to the seedlings, while the colors indicate the size class of the originating endocarp (Oliveira et al. 2018).
It can be observed that endocarps classified as large tend to result in seedlings with significantly greater fresh mass and total size, as indicated by their position in the upper right part of the graph. Medium-sized endocarps, on the other hand, produce seedlings with variable characteristics in terms of fresh mass and total size. Small-sized endocarps, on the other hand, tend to produce smaller seedlings for both analyzed variables. This observation suggests that larger endocarps may have larger seeds with more stored reserves, which can be utilized in the germination process (Oliveira et al. 2018).
Germination and Storage
The endocarps with different storage periods subjected to the germination test-initiated root protrusion from the sixth day after setting up the experiment. From the thirteenth day onwards, the emergence of the first normal seedlings was observed (Fig. 11A). On the tenth day, the first abnormal seedlings (Fig. 11B) and deteriorated seeds (Fig. 11C) became apparent.
The normal seedlings had an average total length of 11.30 cm (3.79 cm hypocotyl length and 7.51 cm root length) and a fresh weight of 0.32 g. Araújo et al. (2001) reported seedling emergence between 10 and 35 days, with a germination rate of 22.80% for seeds from freshly harvested fruits, while Souza et al. (2022) stated that germination can generally occur within 90 days, with a rate of 30%. The average germination rate in this study, i.e., of endocarps that produced normal seedlings, was 19%.
Considering the low germination performance of freshly harvested umbuzeiro seeds, studies indicate that endocarp storage can improve germination rate and speed. In an analysis of seeds stored for 12 and 24 months, Araújo et al. (2001) found that the longer storage period provided higher germination percentages, representing a practical method to overcome dormancy (Marcos-Filho 2005).
However, in this study, no significant differences were observed among endocarps stored for 5-, 17-, and 28-months regarding germination percentage, germination speed index, and average germination speed. Additionally, no significant differences were identified for the total length and fresh weight of seedlings obtained from endocarps stored for different periods (Table 4).
It is worth noting that the seeds used in Araújo et al.'s (2001) study were also stored at 10ºC; however, the material used came exclusively from a single matrix, and it is known that there is a possibility of variation in seed viability and vigor among matrices.
Table 4
Means of the variables percentage of germination (G), germination speed index (GSI), average germination speed (AGS), percentage of abnormal seedlings (AS), percentage of deteriorated seeds (DS), total length of normal seedlings (TLS), and fresh mass of normal seedlings (FMS) derived from endocarps of a natural population of umbu trees (Spondias tuberosa Arr. Câm) in Sergipe stored for 05 (T1), 17 (T2), and 28 (T3) months.
Storage (months) | G (%) | GSI | AGS | AS (%) | DS (%) | TLS (cm) | FMS (g) |
5 | 19.00 | 1.19 | 0.13 | 47.00* | 16.00* | 11.48 | 0.44 |
17 | 10.00 | 0.58 | 0.18 | 16.00* | 73.00* | 10.11 | 0.26 |
28 | 28.00 | 1.81 | 0.25 | 31.00* | 41.00* | 12.30 | 0.27 |
Mean | 19.00 | 1.19 | 0.19 | 31.33 | 43.33 | 11.30 | 0.32 |
Standard Deviation | 9.00 | 0.62 | 0.06 | 15.50 | 28.57 | 1.11 | 0.10 |
Coefficient of Variation (%) | 47.37 | 51.77 | 31.53 | 49.48 | 65.93 | 9.80 | 31.27 |
Significant at the 1% level of probability by the F-test.
Additionally, the endocarps in this study were subjected to a constant temperature of 30 ºC, a value close to the average soil temperature in the Caatinga of Pernambuco, which is 31.02 ºC (Dantas et al. 2020). Although species in the Caatinga are known for their tolerance to high temperatures and water scarcity conditions, the germination capacity to withstand extreme soil temperatures and water deficits still needs to be quantified (Dantas et al. 2020).
For the percentage of abnormal seedlings and deteriorated seeds, there was a significant difference in relation to the storage periods. As the percentage of abnormal seedlings decreases, there is a corresponding increase in the percentage of deteriorated seeds (Fig. 12). This analysis suggests that even if the seeds do not deteriorate, they will not give rise to normal seedlings.
It is possible that some seeds, even without showing external signs of deterioration, were internally compromised or unable to generate normal seedlings due to changes induced by the storage period or conditions, or other intrinsic factors related to the parent trees, affecting seed viability, such as genetic factors.