SciELO - Scientific Electronic Library Online

 
vol.38 número4Efecto de biofertilizantes microbianos en las características agronómicas de la planta y calidad del fruto del chile xcat´ik (Capsicum annuum L.)Corrección de la sintomatología “oreja de ratón” en nogal pecanero con aplicaciones foliares de níquel índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados

Revista

Articulo

Indicadores

Links relacionados

  • No hay artículos similaresSimilares en SciELO

Compartir


Terra Latinoamericana

versión On-line ISSN 2395-8030versión impresa ISSN 0187-5779

Terra Latinoam vol.38 no.4 Chapingo oct./dic. 2020  Epub 12-Feb-2021

https://doi.org/10.28940/terra.v38i4.582 

Artículo científico

Relationship between El Niño Southern Oscillation and Mexico’s orange yield anomalies

Relación entre la Oscilación del Sur El Niño y anomalías de rendimiento de naranja en México

Fidel Blanco-Macías1 
http://orcid.org/0000-0001-7212-6380

Rafael Magallanes-Quintanar2 
http://orcid.org/0000-0002-2331-3275

Miguel Márquez-Madrid1 

Julián Cerano-Paredes3 
http://orcid.org/0000-0002-1528-5139

Martín Martínez-Salvador4 
http://orcid.org/0000-0002-2679-5070

Ricardo David Valdez-Cepeda1  5   
http://orcid.org/0000-0002-6990-3502

1 Universidad Autónoma Chapingo, Centro Regional Universitario Centro-Norte. Calle Cruz del Sur No. 100, Col. Constelación. Apdo. Postal 196, El Orito. 98085 Zacatecas, Zacatecas, México.

2 Universidad Autónoma de Zacatecas, Unidad Académica de Ingeniería Eléctrica. Ave. Ramón López Velarde 801. 98064 Zacatecas, Zacatecas, México.

3 Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias, CENID-RASPA. Margen Derecha Canal Sacramento km 6.5. 35140 Gómez Palacio, Durango, México.

4 Universidad Autónoma de Chihuahua, Facultad de Zootecnia y Ecología. Periférico Francisco R. Almada km 1. 31453 Chihuahua, Chihuahua, México.

5 Universidad Autónoma de Zacatecas, Unidad Académica de Matemáticas. Calzada Solidaridad s/n. 98064 Zacatecas, Zacatecas, México.


Summary:

El Niño-Southern Oscillation (ENSO) effects can be measured as its impacts on crop yields. Nonetheless, those effects on main Mexican crops have been scarcely studied. In this work, the main aim was to identify correlations between the Mexican lemon (Citrus limonia L. Osbeck) or orange (Citrus sinensis L.) yearly mean yield anomalies from 1980 to 2015 and El Niño-Southern Oscillation (ENSO) by involving the extended multivariate ENSO Index (MEI.ext) or the Oceanic Niño Index (ONI). Results indicate that ENSO and lemon annual mean yield anomalies were no correlated. On the other hand, ENSO extreme events in their phase El Niño have leading positively the Mexico’s orange yearly mean yield anomalies, especially from July to November.

Index words: MEI.ext; ONI; correlation; cross-correlation; squared wavelet coherence

Resumen:

Los efectos de la Oscilación del Sur El Niño pueden ser medidos como sus impactos sobre los rendimientos de los cultivos. Sin embargo, esos efectos sobre los cultivos principales en México han sido estudiados escasamente. En este trabajo, el objetivo principal fue identificar las correlaciones entre las anomalías de rendimiento medio anual desde 1980 hasta 2015 de los cultivos limón (Citrus limonia L. Osbeck) o naranja (Citrus sinensis L.) en México y la Oscilación del Sur El Niño (OSEN) al involucrar el Índice OSEN Multivariado Ampliado (MEI.ext) o el Índice Niño Oceánico (ONI). Los resultados indican que el OSEN y las anomalías de rendimiento medio anual de limón estuvieron no correlacionados. Por el contrario, los eventos extremos del OSEN en su fase El Niño han tenido efectos positivos sobre las anomalías de rendimiento medio anual de naranja, especialmente de julio a noviembre.

Palabras clave: MEI.ext; ONI; correlación; correlación cruzada; coherencia de ondeletas

Introduction

El Niño-Southern Oscillation (ENSO) consists of temperature increases and decreases of the sea-surface water in the Tropical (Equatorial) Pacific Ocean and may occur in two phases known as El Niño and La Niña. Both, El Niño, and La Niña are atmospheric phenomena characterized by temperature anomalies in tropical areas of the Pacific Ocean and changing wind and precipitation patterns in tropical and mid-latitude regions (Almeida Silva et al., 2020). Temperature increases are linked to El Niño events as defined by weak winds, whereas temperature decreases are related to La Niña events characterized by intense winds. Those extreme events may have potential impacts on agricultural systems through different ways as briefly described as follows.

It is widely recognized that precipitation, temperature, air humidity, and winds present large intra and interannual variability linked with ENSO. Then, relationships among ENSO, precipitation and temperature may provide relevant causal connections with agricultural production. This can be explained because ENSO extreme events may intensify dangerous climatic conditions. These circumstances could be, for instance, causes of warm and humid conditions, which may result in the spread of crop diseases (Iglesias and Rosenzweig, 2007). In addition, it could be possible crop yield increases or decreases depending on the prevailing phase of ENSO.

In summary, effects of ENSO events could be measured as their’ immediately impact on crop yields. For instance, corn yields of the Southeastern United States of America changed among ENSO phases (El Niño -warm-, Neutral and La Niña -cold-) from 1971 to 2013 (Mourtzinis et al., 2016); noteworthy, greater corn yields were linked with El Niño phase. The highest historical Brazilian coffee yields were generally correlated with the occurrence of El Niño (Almeida Silva et al., 2020). Nonetheless, ENSO effects on Mexico’s main crops have been scarcely studied. Then, the aim of this work is to identify significant correlations between ENSO indexes (bimonthly extended Multivariate ENSO Index -MEI.ext- or three-monthly Oceanic Niño Index -ONI-) and Mexican lemon (Citrus limonia (L.) Osbeck) or orange (Citrus sinensis L.) annual yield anomalies from 1980 to 2015.

Materials and Methods

Mexico’s lemon and orange annual mean yields from 1980 to 2015 were downloaded from http://infosiap.siap.gob.mx/gobmx/datosAbiertos.php (SIAP, 2017). Then, long-term trends were estimated for each time series by means of linear regression analyses using MicroSoft® Excel®, Version 14.7.1 for Mac (MicroSoft Corp., 2011), and residuals were then standardized (yearly yield standardized anomalies) as pointed out by Valdez-Cepeda et al. (2007). It is supported on the idea of removing effects that could be attributed to several technical advances, including the use of chemical fertilizers, mechanization, pesticides, and higher yielding cultivars (Valdez-Cepeda et al., 1998), among other issues.

Data of the extended multivariate ENSO Index (MEI.ext) and the Oceanic Niño Index (ONI) were downloaded from https://psl.noaa.gov/data/correlation/meiv2.data (ESRL, 2017) and https://psl.noaa.gov/data/correlation/oni.data (CPCIT, 2017), respectively. We involved those ENSO indexes because Wolter and Timlin (2011) recommended using MEI.ext for ENSO impacts research, whereas ONI has been used to demonstrate its correlation with meteorological events such as fire activity in the eastern Amazon as demonstrated by Chen et al. (2011).

By this way, Pearson correlation coefficients (r) were estimated by considering the time series of each bi-monthly ENSO Index (MEI.ext) or three-monthly ENSO Index (ONI) and the lemon or orange yearly yield standardized anomalies to estimate their linear correlations (r’s). The r’s were estimated using MicroSoft® Excel®, Version 14.7.1 for Mac (MicroSoft Corp., 2011). All pairs of variables with significant r coefficients were used to estimate cross correlation coefficients (rc’s) as pointed out by Li and Chen (2014). The rc’s were estimated using the Minitab version 16.2.4.0 (Minitab Inc., 2013). In fact, the rc coefficient has been used to evidence teleconnections between ENSO and climatic factors; for instance, a teleconnection between ENSO and minimum temperature (Li and Chen, 2014). In addition, squared wavelet coherence plots for the strongly correlated pairs of variables (significant r’s) were developed by means of the wavelet coherence analysis as proposed by Grinsted et al. (2004). This approach was carried out with R (R Core Team, 2013). Such a technique can be used to calculate squared wavelet coherence between time series, which can be thought of as the local correlation between the time series in time frequency domain; it may allow finding locally phase locked behavior (Grinsted et al., 2004).

Results and Discussion

Correlations (r’s) between lemon yearly mean yield standardized anomalies and the MEI.ext or ONI expressions were not significant. On the other hand, orange yearly mean yield standardized anomalies were significantly correlated with the MEI.ext in its bi-monthly July-August (r = 0.382, P = 0.038) and August-September (r = 0.399, P = 0.029) expressions, and with the ONI in its three-monthly July-August-September (r = 0.381, P = 0.038), August-September-October (r = 0.404, P = 0.027) and September-October-November (r = 0.373, P = 0.042) expressions. Then, those estimated r values suggest that the ENSO effect on the Mexico’s orange yearly mean yield standardized anomalies is in an important positive way from July to November, although other factors may also have had effects.

A careful analysis of the orange yearly mean yield anomalies, July-August and August-September MEI’s, and July-August-September, August-September-October and September-October-November ONI’s behaviors (Figure 1) indicates that moderately intense (1987-1988), and very intense (1997-1998) El Niño events probably influenced in positive ways on the orange annual mean yield during those years. On the other hand, there appears that La Niña extreme events (1988-1989, 2007-2008 y 2010-2011) did not have significant effects on the orange yearly mean yield.

Figure 1: Mexico’s lemon and orange yearly mean yield standardized anomalies, July-August and August-September extended multivariate El Niño-Southern Oscillation Indexes (MEI.ext), and July-August-September, August-September-October, and September-October-November Oceanic Niño Indexes (ONI) time series. 

Our results are very interesting because orange harvest in Mexico is performed during two seasons: August-October (Early) and November-April (Late). So, possible effects of ENSO extreme events on orange trees may tend being favorable from developing fruits to mature fruits stages depending on the harvest season. In other words, the evidenced positive effects of El Niño on the orange yearly mean yield standardized anomalies may depend on the region or locality considering the harvest season of the case.

Therefore, a doubt about whether or not the evidenced ENSO extreme event effects remain more than 12 months starting since they started. Thus, cross correlation coefficients were estimated by involving pairs of variables significantly correlated. For instance, Figure 2 shows July-August-September ONI correlates with the Mexico’s orange annual mean yield standardized anomalies since the ENSO extreme event starts (Lag 0, rc= 0.4, P ≤ 0.05). Similar results were estimated for the remaining strongly correlated pairs of variables (data not shown). Then, those results indicate that the positive relationship may be important within a year, i.e. ENSO and the orange annual mean yield standardized anomalies can be strongly tele connected. Then, Mexican orange orchard managers/owners should have in mind this possible strong relationship to take advantage on and to avoid possible insects, weeds, and pathogens proliferation. This recommendation could be important because most pest species may be favored by warm and humid atmospheric conditions (Iglesias and Rosenzweig, 2007).

Figure 2: Cross-correlations between July-August-September Oceanic Niño Index (ONI) and Mexico’s orange annual mean yield standardized anomalies from 1980 to 2015. 

July-August-September ONI and the Mexico’s orange annual mean yield standardized anomalies time series were involved in a wavelet coherence analysis (Figure 3). Results suggest their correlation is strong (P ≤ 0.05) from 1990 to 2004 at 2- to 6-yr periodicities. Notably, both phenomena were in-phase (horizontal arrows toward right) from 1990 to 2002 at a mean periodicity of 3.5 yr, and July-August-September ONI was leading the Mexico’s orange annual mean yield standardized anomalies (arrows pointing to the right-down) from 1992 to 2004 at mean periodicity of 5 yr and from 1996 to 2002 at mean periodicity of 2.5 yr. In addition, a significant correlation between both variables almost in anti-phase (arrows pointing to the left-down) is markedly noted from 2007 to 2009 at a 3-yr mean periodicity; in other words, the period 2007 to 2009 may be characterized as a short-term with dominant years when July-August-September ONI and orange yearly mean yield standardized anomalies shown opposite behaviors. It deserves be mentioned that during such a short-term took place one La Niña extreme event (2007-2008), which was followed by another (2010-2011, see Figure 1). This means La Niña effect on orange yields in Mexico may also be an important issue; thus, Mexican orchard managers/owners must have in mind such an influence could be negative.

Figure 3: Squared wavelet coherence between July-August-September Oceanic Niño Index (ONI) and Mexico’s orange annual mean yield standardized anomalies time series from 1980 to 2015. The 5% significance level against red noise is shown as a thick black contour. The thick grey contour designates the cone of influence, where edge effects might distort the picture is shown as a lighter shade. The significant red section on the left shows in-phase behavior (arrows toward right), and the significant red section on the right shows anti-phase behavior (arrows toward left). 

In general, our results are compelling evidence about El Niño positive effects on orange annual mean yields and agree with that from Mourtzinis et al. (2016) and Almeida Silva et al. (2020). In fact, Mourtzinis et al. (2016) consigned that corn yield of the Southeastern United States of America changed among ENSO phases from 1971 to 2013, and the greater yields were strongly linked with El Niño phase. Almeida Silva et al. (2020) found that the highest historical Brazilian coffee yields were generally correlated with the occurrence of El Niño. Then, our results support the idea about ENSO effects on crop yields at various regions around the world; this could be linked to ENSO influences seasonal precipitation in several places (Roy and Kripalani, 2019) and other atmospheric conditions (Roy et al., 2018).

Our novel and preliminary findings suggest that there remain unknown ENSO impacts on orange yields from main orange growing areas/regions throughout the country, and the underlying involved climatic mechanisms. Such a knowledge can be useful to develop strategies for orange varieties adaptation and orchard management addressed to Mexico’s regions (and localities) by means of quantification of ENSO effects on climatic conditions (e.g. temperature, maximum and minimum extreme temperatures, total and seasonal rainfall, and air humidity) or ENSO effects during specific time linked to orange trees’ phenological stages. That kind of strategies could also be useful in the cases of other perennial crops and species with growing seasons lower than a year.

Conclusions

MEI.ext and ONI were no correlated with Mexico’s lemon annual mean yield anomalies. On the other hand, July-August and August-September extended Multivariate El Niño-Southern Oscillation Indexes, and July-August-September, August-September-October and September-October-November Oceanic Niño Indexes were associated with Mexico’s orange annual mean yield anomalies when both phenomena were in-phase, especially from July to November during the 1990-2004 period. The main effect may correspond to an occurred very intense (1997-1998) El Niño event. Such an effect should be linked to the greater orange yields registered during that period (1990 to 2004). In addition, the period 2007 to 2009 may be characterized as a short-term with dominant years when ENSO and orange yearly mean yield standardized anomalies shown opposite behaviors. During such a short-term took place one La Niña extreme event (2007-2008) followed by another (2010-2011). This means La Niña effect on orange yields in Mexico may be an important issue. Future research works should dedicate efforts to further comprehension on possible relationships between ENSO indexes and Mexico’s orange yields at local or regional levels to provide strong elements to decision makers, i.e. orchard managers and owners.

Ethics Statement

Not applicable.

Consent for Publication

Not applicable.

Data Availability

Mexico’s lemon and orange annual mean yields from 1980 to 2015 are available at infosiap.siap.gob.mx/gobmx/datosAbiertos.php. The extended multivariate ENSO Index (MEI.ext) and the Oceanic Niño Index (ONI) data are available at https://psl.noaa.gov/data/correlation/meiv2.data and https://psl.noaa.gov/data/correlation/oni.data, respectively.

Competing Interests

No conflict of interest declared by authors.

Funding

Not applicable.

Authors Contributions

Conceived the study: Ricardo David Valdez-Cepeda, Fidel Blanco-Macías, Miguel Márquez-Madrid, Julián Cerano-Paredes. Analyzed the data: Ricardo David Valdez-Cepeda, Julián Cerano-Paredes, Rafael Magallanes-Quintanar. Wrote the manuscript: Ricardo David Valdez-Cepeda, Fidel Blanco-Macías, Martín Martínez-Salvador. All authors reviewed the manuscript.

Acknowledgments

Authors would like gratefully acknowledge helpful suggestions from the anonymous reviewers to improve the manuscript.

Literature Cited

Almeida Silva, K., G. de Souza Rolim, T. T. Borges Valeriano, and J. R. da Silva Cabral de Moraes. 2020. Influence of El Niño and La Niña on coffee yield in the main coffee-producing regions of Brazil. Theor. Appl. Clim. 139: 1019-1029. doi: https://doi.org/10.1007/s00704-019-03039-9. [ Links ]

Chen, Y., J. T. Randerson, D. C. Morton, R. S. DeFries, G. J. Collatz, P. S. Kasibhatla, L. Giglio, J. Yufang, E. Miriam, and E. Marlier. 2011. Forecasting fire season severity in South America using sea surface temperature anomalies. Science 334: 787-791. doi: https://doi.org/10.1126/science.1209472. [ Links ]

CPCIT (Climate Prediction Center Internet Team). 2017. Cold & warm episodes by season. National Oceanic & Atmospheric Administration. National Weather Service. Climate Prediction Center. Silver Spring, MD, USA. https://origin.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ONI_v5.php . (Consultation: May 20, 2017). [ Links ]

ESRL (Earth System Research Laboratories). 2017. MEI Index. National Oceanic and Atmospheric Administration. Earth System Research Laboratory. Physical Sciences Division. Boulder, CO, USA. https://psl.noaa.gov/research/ . http://www.esrl.noaa.gov/psd/enso/mei/table.html (Consultation: May 20, 2017). [ Links ]

Grinsted, A., J. C. Moore, and S. Jevrejeva. 2004. Application of the cross wavelet transform and wavelet coherence to geophysical time series. Nonlin. Proc. Geophys. 11: 561-566. doi: https://doi.org/10.5194/npg-11-561-2004. [ Links ]

Iglesias, A. and C. Rosenzweig. 2007. Climate and pest outbreaks. Chapter 53. pp. 87-89. In: D. Pimentel. (ed.). Encyclopedia of pest management. CRC Press. Boca Raton, FL, USA. ISBN-13: 978-1-4200-5361-6. [ Links ]

Li, Q. and J. Chen. 2014. Teleconnection between ENSO and climate in South China. Stoch. Environ. Res. Risk Assess. 28: 927-941. doi: https://doi.org/10.1007/s00477-013-0793-z. [ Links ]

MicroSoft Corporation. 2011. MicroSoft® Excel®, Version 14.7.1 for Mac. Santa Rosa, CA, USA. [ Links ]

Minitab Incorporation. 2013. Minitab version 16.2.4.0 [Computer software ]. State College, PA, USA. [ Links ]

Mourtzinis, S., B. V. Ortiz, and D. Damianidis. 2016. Climate change and ENSO effects on Southeastern US climate patterns and maize yield. Sci. Reports. 6: 1-7. doi: https://doi.org/10.1038/srep29777. [ Links ]

R Core Team. 2013. R: A language and environment for statistical computing. R Foundation for Statistical Computing. Vienna, Austria. http://www.R-project.org/ . (Consultation: May 30, 2017). [ Links ]

Roy, I. and R. H. Kripalani. 2019. The role of natural factors (part 1): addressing on mechanism of different types of ENSO, related teleconnections, and solar influence. Theor. Appl. Clim. 137: 469-480. doi: https://doi.org/10.1007/s00704-018-2597-z. [ Links ]

Roy, I., A. S. Gagnon, and D. Siingh. 2018. Evaluating ENSO teleconnections using observations and CMIP5 models. Theor. Appl. Clim. 136: 1085-1098. doi https://doi.org/10.1007/s00704-018-2536-z. [ Links ]

SIAP (Servicio de Información Agroalimentaria y Pesquera). 2017. Datos abiertos. Secretaría de Agricultura, Ganadería, Desarrollo Rural, Pesca y Alimentación (México). Servicio de Información Agroalimentaria y Pesquera, Servicio de Información Agroalimentaria y Pesquera, http://infosiap.siap.gob.mx/gobmx/datosAbiertos.php (Consultation: May 20, 2017). [ Links ]

Valdez-Cepeda, R. D. and E. Olivares-Sáenz. 1998. Fractal analysis of Mexico’s annual mean yields of maize, bean, wheat and rice. Field Crops Res. 59: 53-62. doi: https://doi.org/10.1016/S0378-4290(98)00108-7. [ Links ]

Valdez-Cepeda, R. D., O. Delgadillo-Ruiz, R. Magallanes-Quintanar, G. Miramontes de León, J. L. García-Hernández, A. Enciso-Muñoz, and B. Mendoza. 2007. Scale-invariance of normalized yearly mean grain yield anomaly series. Adv. Complex Syst. 103: 395-412. doi: https://doi.org/10.1142/S0219525907001161. [ Links ]

Wolter, K. and M. S. Timlin. 2011. El Niño/Southern Oscillation behaviour since 1871 as diagnosed in an extended multivariate ENSO index (MEI.ext). Int. J. Clim. 31: 1074-1087. doi: https://doi.org/10.1002/joc.2336. [ Links ]

Received: July 02, 2019; Accepted: August 05, 2020

Corresponding author (vacrida@gmail.com)

Creative Commons License This is an open-access article distributed under the terms of the Creative Commons Attribution License