Adaptability and stability of maize ear length jala race in five environments
1. INIFAP-CIRNO, Campo Experimental Valle de Culiacán, Carr. Culiacán-El Dorado Km 16.5 C.P. 80000, Culiacán, Sinaloa., Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias, INIFAP, CIRNO,
Abstract
In Mexico, and in the world, maize (Zea mays L.) is grown under different environmental conditions that influence different expressions of genotypes, a phenomenon known as genotype-by-environment interaction (GEI). In breeding programs, it is important to define what outstanding genotypes are in terms of performance, adaptability and stability in multi-environments, to be able to recommend their commercial use. The objective of this work was to study the adaptability and stability of ear length in 14 representative populations of maize Jala race by means of AMMI and SREG methods. Populations were planted in 2012 in five locations, four of them in the State of Nayarit and the other one in the Colegio de Postgraduados (CP), Campus Montecillo, Montecillo, Texcoco, State of Mexico. Localities of Nayarit were established at the beginning of the rainy season and the one in CP was planted on May tenth, where irrigation was applied until the establishment of the rainy season. A randomized complete block design was used with three replications. The studied trait was the ear length. AMMI and SREG statistical analyzes, as well as biplot graphs, were carried out with SAS software. P1, P3 and P4 populations were those that presented a low population x locality interaction, while P5, UAN-2009C, and 13XT and 8XT varietal hybrids presented high values of interaction. The locality that best discriminated the environments was San José de Mojarras.
Received: 2018 March 23; Accepted: 2018 July 25
Keywords: Key words: Maize Jala race, AMMI, SREG, ear length.
Introduction
In Mexico and around the world, according to Nzuve et al. (2013) cited by Sánchez et al. (2016), maize (Zea mays L.) is cultivated in a wide range of environmental conditions associated with soil moisture state, solar radiation, temperature, soil types and production systems. These conditions cause different expressions of genotypes, a phenomenon known as genotype-by-environment interaction (GEI) (McDermott & Coe, 2012). Balzarini et al. (2005) mention that interaction drives to two situations; a) stability, referring to the ability of genotypes to express themselves consistently and it can be in both directions: high or low yield performance and b) specific adaptability, described as the outstanding expression of a genotype in a determined range of environments or localities.
In plant breeding programs, it is important to define which are the outstanding genotypes, in terms of performance, adaptability and stability, assessing these properties in multi-environments to be able to recommend their commercial use (Crossa et al., 2006; Gordón-Mendoza et al., 2006). Adaptation in a broad sense refers to the best performance related to a genotype in most of the test environments, while adaptation in a specific sense shows the genotype with the most suitable performance in a test environment (Fuentes & Queme, 2005).
A challenge for plant breeding programs is to obtain genotypes with higher yields, but performance potential is frequently masked up by GEI (Lozano et al., 2015). Gordón et al. (2006) mention that GEI occurs when genotypes respond differently to environmental variants. Crossa et al. (2006) mention that studying GEI in plant breeding is very important, since it represents the result of the response of each genotype to environmental variations. The plant breeder has to face these problems to obtain a genotype closer to the ideal one.
Tests of genotypes in multi-environments are important for plant breeding as indicated by Crossa et al. (2006), because genotypes are evaluated in different environmental conditions, their response is compared, their general stability and adaptability are assessed, GEI is studied, and the best genotypes in specific environments and across environments are selected for comparison in experimental trials or recommended for their commercial use.
There are various models to interpret the response of genotypes across environments and to study and interpret GEI in agricultural experiments, among them lineal, bilineal and lineal-bilineal models. Models of fix lineal-bilineal effects, as SREG (Crossa & Cornelius, 1997) and of Additive Main effect & Multiplicative Interaction (AMMI) (Gauch, 1988; Gauch & Zobel, 1997) are used to study patterns of genotypic response across environments.
AMMI model is defined as the Additive Main effect and Multiplicative Interaction model (Gauch, 2006; Glaz & Kang, 2008), it essentially consists in combining technics of analysis of variance and Principal Component Analysis (PCA) in a single model, here the analysis of variance allows to study the main effects of genotypes and environments, while GEI is addressed in a multivariate way by means of PCA, where a reparametrization of the regression model is performed to improve interpretation of the interaction (Zobel et al., 1988).
Interpretation of results obtained from the AMMI analysis is based on a graphic representation called GGE biplot, making the identification and visual interpretation of genotypes and assessed environments easier, as well as the exploration of behavioral patterns attributed to GEI effects (Yant et al., 2000; McDermott & Coe, 2012). GGE biplot graphs are made using the two first principal components (PC1 and PC2). Thus, the genotype that is in the vertex is the one that best responds in assessed environments, since the two first principal components explain the highest portion of the variation in the GEI (Yant et al., 2001).
From the AMMI model, the effect of genotypes, combined with the effect of the interaction, (G + G x E), can be estimated, so that a model, called site regression (SREG), is generated. This model is recommended when environmental effects are the main source of variation, and it presents other extra advantages. Indeed, SREG model, which include G+GE in the bilineal term, provides a graphic analysis of easy interpretation of genotypes performance plus a GEI effect’, called GGE biplot (Yan et al., 2000). The graph allows visualizing genotypes and environments clusters (mega-environments or macroenvironments) with similar response patterns, as well as identifying the most representative and discriminatory environments (Yan & Rajcan 2002).
GGE biplot is constructed from the first two principal components (PC) of the SREG model. The first component, when is found to be highly correlated to the main effect of the genotype, represents the proportion of the yield due only to genotype traits. The second component represents the portion of yield due to GEI. Genotypes close to each other in GGE biplot present similar response pattern across environments. Environments close to each other, given by the acute angle between their vectors, indicate a positive environmental association, that is, similar response patterns in the relative performance of a cluster of genotypes. An absence of association between environments is given by a right angle between vectors and a negative association by an obtuse angle (Yan and Rajcan 2002). Therefore, this study aimed to investigate the adaptability and stability of ear length of 14 populations of maize Jala race by AMMI and SRG methods.
Materials and Methods
Fourteen populations of maize, representative of the Jala race or its germplasm, were assessed. Five populations were sampled from farmers of the Jala region in Nayarit during 2012, which were referred as P1, P2, P3, P4 and P5; seven populations were composites of individual selection (Aguilar & Carballo, 2007) UAN2008, UAN-2009A, UAN-2009B, UAN-2009C, UAN-2010, UAN-2011, and Montecillo-2007; and two populations were varietal hybrids 13 XT and 8 XT with 50% of Jala germplasm, which were used as checks.
Populations were planted in 2012 in five localities (Table 1), three of them (L1, L2 and L3) were established in plots property of cooperating farmers from the state of Nayarit. L4 locality was established in the facilities of Agriculture Academic Unit of the Autonomous University of Nayarit, in Xalisco, Nayarit, and L5 locality in the Postgraduate College, Montecillo Campus, Montecillo, Texcoco, State of Mexico.
Edaphoclimatic characteristics and geographic location of selected sites for the evaluation of 14 populations of maize Jala race.
Location | Name | Soil texture | pH | AR (mm) | Altitude (m) | AAT (°C) | N L | WL |
---|---|---|---|---|---|---|---|---|
L1 | San José de Mojarras | Clay | 5.8 | 1113 | 912 | 24.3 | 21°25’ | 104°36’ |
L2 | Ixtlán del Río | Loam | 6.5 | 859.8 | 1038 | 23 | 21° 02’ | 104°33’ |
L3 | Jala | Sand | 5.6 | 837.4 | 1016 | 23.2 | 21°05’ | 104°31 |
L4 | Xalisco | Sandy Loam | 4.7 | 1232.4 | 984 | 23 | 21°22’ | 104°24 |
L5 | Montecillo | Clay | 8.4 | 655.6 | 2250 | 16.3 | 19°30 | 98°52’ |
^{TFN1}AR: Average Rainfall; AAT: Average Annual Temperature; pH: soil pH; NL: North Latitude; WL: West Longitude
The trials in L1, L2, L3 and L4 localities coincided with the beginning of the rainy season; in L5 locality maize was planted on May 10^{th}, where irrigation was applied until the establishment of the rainy season. In each locality, assessment was performed in a randomized complete block design with three replications. In all cases, the experimental unit was two rows of 5 m long and 0.80 m between rows, and two seeds were sowed every 0.35 m. Fertilization and cultivation activities were performed according to traditional practices of each locality. The 100N-60P-30K formula was used to fertilize plants; an application of 50% of nitrogen and all phosphorus and potassium was performed at sowing and the rest of nitrogen was applied later during cultivation.
From each plot, five plants with complete competency (cc) were randomly selected, to which ear length trait was measured: ear were dried at room temperature and length was measured when they reached constant moisture (LMZ, in cm).
The analysis according to AMMI method is based on a lineal-bilineal statistical method (Crossa & Cornelius, 2000), in which the main effects of genotypes and environments, considered as lineal terms, are explained by means of a conventional analysis of variance; the bilineal component (not additive) is attributed to GEI and is analyzed by means of Principal Component Analysis (PCA). If the two first principal components explain an important portion of the variance of the GEI matrix (60%) (Gauch & Zobel, 1988), a graphic representation (biplot) of the variability of the observations can be performed, where environments and genotypes are considered (Kempton, 1984; Crossa, 1990).
For the analysis, AMMI programming methods were used, as described by Vargas & Crossa (2000), using the following mathematic model:
<mml:msub>
<mml:mrow>
<mml:mo>Y</mml:mo>
</mml:mrow><mml:mrow>
</mml:msub><mml:mo>i</mml:mo>
<mml:mo>j</mml:mo>
</mml:mrow><mml:mo>=</mml:mo>
<mml:mi> </mml:mi>
<mml:mo>µ</mml:mo>
<mml:mi> </mml:mi>
<mml:mo>+</mml:mo>
<mml:mi> </mml:mi>
<mml:msub>
<mml:mrow>
<mml:mo>g</mml:mo>
</mml:mrow><mml:mrow>
</mml:msub><mml:mo>i</mml:mo>
</mml:mrow><mml:mi> </mml:mi>
<mml:mo>+</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mo>e</mml:mo>
</mml:mrow><mml:mrow>
</mml:msub><mml:mo>j</mml:mo>
</mml:mrow><mml:mi> </mml:mi>
<mml:mo>+</mml:mo>
<mml:mi> </mml:mi>
<mml:mi> </mml:mi>
<mml:msub>
<mml:mrow>
<mml:mo>λ</mml:mo>
</mml:mrow><mml:mrow>
</mml:msub><mml:mo>k</mml:mo>
</mml:mrow><mml:mi> </mml:mi>
<mml:msub>
<mml:mrow>
<mml:mo>α</mml:mo>
</mml:mrow><mml:mrow>
</mml:msub><mml:mo>i</mml:mo>
<mml:mo>k</mml:mo>
</mml:mrow><mml:mi> </mml:mi>
<mml:msub>
<mml:mrow>
<mml:mo>γ</mml:mo>
</mml:mrow><mml:mrow>
</mml:msub><mml:mo>j</mml:mo>
<mml:mo>k</mml:mo>
</mml:mrow><mml:mi> </mml:mi>
<mml:mo>+</mml:mo>
<mml:mi> </mml:mi>
<mml:msub>
<mml:mrow>
<mml:mo>E</mml:mo>
</mml:mrow><mml:mrow>
</mml:msub><mml:mo>i</mml:mo>
<mml:mo>j</mml:mo>
</mml:mrow> |
Where: Y_{ij} = yield performance of the i^{th} genotype in the j^{th} environment; µ = general average; g_{i} = effect of the i^{th} genotype; e_{j} = effect of the j^{th} environment; λ_{k} = squared root of the characteristic vector of the k^{th} axe of the PCA; α_{ik} = qualification of the PCA for the k^{th} axe of the i^{th} genotype; γ_{jk} = qualification of the PCA for the k^{th} axe of the j^{th} environment; E_{ij} = value of error.
The site regression model (SREG) is based on a model similar to the AMMI model, but lineal terms of genotypes are not considered individually, adding themselves to the multiplicative term of GEI. SREG model is useful for environments clustering with no opposite interaction, that is, ordering without change of genotypes in environments that conform a cluster (Cornelius &Crossa, 1999; Yan et al., 2000). Moreover, it allows the simultaneous representation of genotypes and environments variability, based on Principal Component Analysis (Yan et al., 2000). The SREG site regression model is the following: Yij-Ȳi= λ1Ɛi1ƞj1+ λ2Ɛi2ƞj2+ eij; where: Yij = average performance observed for genotype i in the environment j; Ȳi = average of genotypes in the environment j; λ1 = intrinsic value of principal component 1 (PC1); λ2 = intrinsic value of principal component 2 (PC2); Ɛi1 = score of the genotype i on PC1; Ɛi2 = score of genotype j on PC2; ƞj1 = score of the environment i on PC1; ƞj2 = score of environment j on PC2; eij = residual.
AMMI and SREG statistical analysis, as well as biplot graphs were performed with SAS® software (SAS Institute Inc., 2002), according to procedures established by Vargas & Crossa (2000).
Results and discussion
AMMI combined analysis of variance (Table 2) detected highly significant differences among localities and populations, while the interaction of both factors resulted not significant. A coefficient of variation of 8.13 was obtained, and it is considered as an acceptable value regarding the condition of this experiment. Environmental conditions and their effects on populations were observed to be different in all the test environments. The absence of significance of GEI for ear length, which is a highly quantitative trait, and therefore very influenced by environmental conditions, indicates that statistically there was no differential response in ear length of genotypes among the different test environments (Canales et al., 2016).
AMMI analysis of variance for ear length of 14 populations of maize Jala race assessed in five localities during 2012.
S.V. | df. | SS | MS |
---|---|---|---|
Locations (l) | 4 | 298.81 | 74.70^{**} |
Populations (p) | 13 | 221.34 | 17.02^{**} |
Interactions (pxl) | 52 | 179.6 | 3.45 |
PC1 | 16 | 98.53 | 6.15 |
PC2 | 14 | 41.39 | 2.95 |
Error | 138 | 342.75 | 2.48 |
Total Corrected | 209 | 1056.3 | |
CV (%) | 8.13 | ||
Mean | 19.36 |
^{TFN2}S.V: source of variation; df: degrees of freedom; **: significant; SS: Sum of squares; MS: Mean Square; CV: Coefficient of Variation; PC1 and PC2: Principal Components 1 and 2.
Besides, in the quantity of total sum of squares, in the AMMI analysis, the effect of locations contributed to 28% while the effects of population and of locations x population interaction represented 21 and 17%, respectively. The above-mentioned indicated that the effects of locations contributed in a higher proportion to the variation in the expression of ear length, in comparison with factors of population and of population x location interaction. These results have been obtained in previous research, where environments and GEI factors were reported to be superior to the effects of genotypes (Alejos et al., 2006; Palemón et al., 2012).
Populations that presented a higher ear length were UAN-2011, P5 and UAN-2008 in all the test localities. Locations with the highest expression of ear length were L2 (Ixtlán del Rio) and L4 (Xalisco). Populations that obtained the lowest absolute values of PC1, that is, which less interacted with the environment, were P1 with 0.20 and P8 (UAN-2009A) with 0.35, all with values close to zero, which can be considered as the most stable across environments (Medina et al., 2002; Alejos et al., 2006; Palemón et al., 2012). Locations with the best performance regarding absolute values of PC1 were L2 (Ixtlán del Rio) with -0.61 and L3 (Jala) with 0.72 (Table 3).
Means of ear length in the different populations and environments and PC values of the 14 assessed populations in five environments during 2012 period.
Populations | Environments | Means | PC1 | ||||
---|---|---|---|---|---|---|---|
L1 | L2 | L3 | L4 | L5 | |||
P1 | 18.5a | 19.8ab | 16.7a | 18.6bd | 17.1a | 18.1ce | 0.20 |
P2 | 17.9a | 20.4ab | 16.4a | 18.7bd | 15.8a | 17.8de | -0.24 |
P3 | 20.7a | 20.7ab | 17.6a | 20.8ad | 18.5a | 19.6ad | -0.10 |
P4 | 21.1a | 20.6ab | 17.1a | 19.9ad | 18.2a | 19.4ad | -0.14 |
P5 | 21.1a | 22.9a | 18.7a | 22.6a | 17.7a | 20.6a | -0.82 |
Montecillo 2007 | 21.5a | 20.2ab | 18.8a | 19.6ad | 16.7a | 19.2ad | -0.26 |
UAN-2008 | 22.0a | 21.1ab | 17.3a | 21.2ac | 18.9a | 20.1ab | -0.32 |
UAN-2009A | 18.8a | 22.3ab | 19.1a | 20.6ad | 18.9a | 19.9ac | 0.35 |
UAN-2009B | 20.0a | 21.3ab | 18.4a | 21.6ab | 17.8a | 19.8ac | -0.30 |
UAN-2009C | 21.4a | 21.8ab | 17.4a | 21.3ac | 17.6a | 19.9ac | -0.70 |
UAN-2010 | 20.4a | 21.4ab | 16.9a | 22.0ab | 17.7a | 19.7ad | -0.67 |
UAN-2011 | 21.9a | 19.9ab | 20.8a | 21.8ab | 19.8a | 20.8a | 0.43 |
13 xt | 18.2a | 19.2ab | 17.6a | 17.6cd | 19.5a | 18.4be | 1.27 |
8 xt | 16.0a | 17.9b | 15.9a | 17.4d | 18.6a | 17.1e | 1.32 |
Means | 19.9a | 20.7a | 17.7b | 20.2a | 18.0b | ||
PC1 | -0.88 | -0.61 | 0.72 | -0.98 | 1.75 |
^{TFN3}PC1; Principal Component 1. Means with the same letter in column were not statistically different (Tukey, P ≤ 0.05).
Figure 1 presents the interpretation of the extent which populations interacted with the environment. P12 (UAN2011), P5 and P7 (UAN-2008) populations presented ear lengths superior to the general mean, while P14 (8XT) and P2 populations presented ear lengths lower than the mean (Figure 1). Populations with low values of PC1 that interacted less with the environment were P3 and P4, while P5, P10 (UAN-2009C), P11 (UAN2010), P13 (13XT) and P14 (8XT) populations presented absolute values superior to 0.60 on PC1 and therefore it was inferred that they contributed to a greater extent to the interaction (Crossa 1990; Medina et al., 2002). L3 (Jala), L2 (Ixtlán del Rio) and L1 (San José de Mojarras) localities were identified as those with the best performance by their PC1 values, as they were closer to zero or for presenting lower variation between them.
[Figure ID: f1] Figure 1.
Graphic representation of PC1 according to the avarege ear length of 14 populations evaluated in five locations.
In AMMI analysis, the first principal component explained 54% of the sum of squares and the second component explained 23% of the total sum of squares, therefore, both principal components described 77% of the effect of GEI. Yan et al. (2000) mention that environments whose angles were lower than 90° cluster similar genotypes. In these results two clusters of environments were observed; in the first environment were found L1 (San José de Mojarras), L2 (Ixtlán del Rio) and L4 (Xalisco) localities, in the second environment were found L3 (Jala) and L5 (Montecillos) localities. Based on the major length of the vectors of each environment, localities that better distinguished populations were; L1 (San José de Mojarras), L2 (Ixtlán del Rio) and L5 (Montecillo), according to criteria applied by Kempton (1984) and Yan et al. (2000). Such criteria indicate that the length of vectors gives indication on variance magnitude, that is why vectors of major length distinguished genotypes in a better way.
Three populations showed a trend close to zero, and, in a narrower sense, populations (P3, P4 and P1), were the most stable in the localities where they were assessed (Figure 2).
[Figure ID: f2] Figure 2.
Biplot AMMI for 14 populations. The points (P) represent populations and vector (L) the localities.
The interpretation of the SREG biplot model is similar to the one of the AMMI biplot, the two first components explained 66% of the variation for component 1 and 16% for component 2, which are represented in ‘biplot’ of Figure 3, exhibiting response patterns of 14 populations of maize Jala race assessed in five localities. Crossa (1990) mentions that genotypes and environments with high coordinates in component 1 direction, considered in absolute value, contributed to a greater extent to GEI, while genotypes and environments with values close to zero in component 1 direction have a few participation on this effect.
[Figure ID: f3] Figure 3.
SREG biplot for 14 populations of maize Jala race. Points (P) represent populations and vectors (L) represent localities.
In GGE biplot, each population was represented by a point or a defined marker when the coefficients of each population were graphically represented in PC1 (X-axis) against the respective coefficients in PC2 (Y-axis). Environments were presented as vectors originating from the coordinate (0,0) of the biplot and extended until the corresponding marker. According to Yan et al. (2000), when the points are drawn in a graph, representing scoring of both principal axis, a polygon is formed with cultivars that remain in the extern part of the figure (in this case, P12 (UAN-2011), P13 (13xt), P14 (8XT), P2 and P5 populations). Theses populations are those that bring the most to the interaction, that is, they are those of highest and lowest length. Perpendicular lines to each one of the polygon sides and that cross by the biplot origin divided polygon in five sectors. Localities remained clustered in only two of these sectors. The genotype corresponding to a vertex of the polygon included into a sector has the best performance in the environments that are inside of this sector (Ibañez et al., 2006; Yan et al. 2000). P12 vertex population (UAN-2011) had the highest mean of ear length (20.8 cm) in L1 (San José de Mojarras), L3 (Jala) and L5 (Montecillo) localities that were inside of one of the sectors. P5 vertex population had the best performance (22.8 cm) in L2 (Ixtlán del Rio) and L4 (Xalisco) localities, included inside of the other sector (Figure 3). P2 and P14 (8XT) populations, corresponding to vertexes of the polygon that remained far away from the assessed localities, were those that presented the lowest means of ear length across all the localities (Figure 3). Populations, whose markers stayed located inside of the polygon, were those of lower response to environments of expression and those located closer to biplot origin were even more stable.
According to Yan & Rajcan (2002), ideal genotypes must have a high scoring for PC1 (yield ability) and a low scoring for PC2 (high stability). P7 (UAN2008) population presented a high potential of yield performance (high PC1) and a predictable or stable performance (PC2 close to zero) according to Figure 3 and may be considered as the ideal population for its general adaptation in all the localities.
According to Yan et al. (2001), ideal cultivars must have a high scoring for PC1 (high mean for yield) and a scoring for PC2 close to zero (higher stability). Similarly, ideal test environments must have a high scoring for PC1 (more differentiation for cultivars) and scoring for PC2 close to zero (more representative for a mean of environments). In this case, the most representative locality is L1 (San José de Mojarras).
Conclusions
P1, P3 and P4 populations were those that presented a low population x locality interaction, while P5, UAN-209C, and the varietal hybrids 13XT and 8XT presented high values of interaction. The locality that best differentiated the environments was San José de Mojarras.
The GGE-SREG biplot model, P7 (UAN-2008) was identified as the population with the highest capacity of adaptation in the test localities.
In general, it was observed that the population that presented the lowest interaction was P3, and the one that most contributed to the interaction was P5. The highest variation was attributed to the effects of locality.
^{fn1}Cite this paper: López-Guzmán, J. A., Aguilar-Castillo, J. A., GarcíaZavala, J. J., Lobato-Ortiz, R. (2018). Adaptability and stability of maize ear length jala race in five environments. Revista Bio Ciencias 5(nesp), e472. doi https://doi.org/10.15741/revbio.05.nesp.e472
1. | |
2. | |
3. | |
4. | |
5. | |
6. | |
7. | |
8. | |
9. | |
10. | |
11. | |
12. | |
13. | |
14. | |
15. | |
16. | |
17. | |
18. | |
19. | |
20. | |
21. | |
22. | |
23. | |
24. | |
25. | |
26. | |
27. | |
28. | |
29. |
Enlaces refback
- No hay ningún enlace refback.
Revista Bio Ciencias, Año 13, vol. 9, Enero 2022. Sistema de Publicación Continua editada por la Universidad Autónoma de Nayarit. Ciudad de la Cultura “Amado Nervo”, Col. Centro, C.P.: 63000, Tepic, Nayarit, México. Teléfono: (01) 311 211 8800, ext. 8922. E-mail: revistabiociencias@gmail.com, revistabiociencias@yahoo.com.mx, http://revistabiociencias.uan.mx. Editor responsable: Dr. Manuel Iván Girón Pérez. No. de Reserva de derechos al uso exclusivo 04-2010-101509412600-203, ISSN 2007-3380, ambos otorgados por el Instituto Nacional de Derechos de Autor. Responsable de la última actualización de este número Dr. Manuel Iván Girón Pérez. Secretaria de Investigación y Posgrado, edificio Centro Multidisciplinario de Investigación Científica (CEMIC) 03 de la Universidad Autónoma de Nayarit. La opinión expresada en los artículos firmados es responsabilidad del autor. Se autoriza la reproducción total o parcial de los contenidos e imágenes, siempre y cuando se cite la fuente y no sea con fines de lucro.
Revista Bio Ciencias por Universidad Autónoma de Nayarit se encuentra bajo una licencia de Creative Commons Reconocimiento-NoComercial-SinObraDerivada 4.0 Internacional
Fecha de última actualización 13 de Enero de 2022
licencia de Creative Commons Reconocimiento-NoComercial-SinObraDerivada 4.0 Internacional