Inglés (pdf)
Articulo en XML
Referencias del artículo
Traducción automática
Enviar articulo por email
Citado por SciELO 
Similares en
SciELO 
Permalink
INTRODUCTION
The leaf is the exchange surface between the plant and the environment. It is the organ where the conversion of sunlight into biochemical energy occurs (Blanco and Folegatti, 2005; Pandey and Singh, 2011). The intensity of these exchanges, as well as the photosynthetic activity, have a direct relationship with the leaf area (Mokhtarpour et al., 2010). For this reason, among the variables that characterize the leaves, the leaf area and the parameters associated with it (leaf area index, net assimilation rate, specific leaf area, specific leaf weight) are the most representative, especially in relation to physiological and environmental factors (Sala et al., 2015).
Leaf area is a valuable index to identify and understand many agronomic and physiological processes, such as photosynthetic efficiency, evaporation, respiration, water balance, transpiration, energy balance, yield potential, response to irrigation and fertilizers (Blanco and Folegatti, 2005, Fotis et al., 2018). Although the precise estimation of the foliar area is very important in all crops, it is crucial in fruit trees (Santesteban et al., 2006) due to its impact on the size and filling potential of the fruit (Demirsoy, 2009, Keramatlou et al., 2015). In this sense, Grantz and Williams (1993) pointed out early that the distribution and density of the foliar area affects, although indirectly, the quality of the fruit and the incidence of diseases. This is so because the amount of solar radiation intercepted by the canopies influences the microclimate (light, temperature and humidity) within them (Jianga et al., 2017). Hence, the measurement of the leaf area is of particular importance both for the study of plant physiology, and for analyzing the vegetative and reproductive responses of plants to climatic conditions and to different agronomic and management practices.
In general terms, the leaf area can be estimated through direct or indirect methods (Weiss et al., 2004). Direct methods, although considered the most accurate, are destructive and require expensive instruments, which gives them limited applicability (Kumar et al., 2017). Through direct methods, the total area is usually obtained by measuring the area of all the excised leaves of the plant (Fallovo et al., 2008). By destroying the leaf, it is not possible to make successive measurements when it is necessary, for example, to verify the evolution of the plant during the growing season (Jonckheere et al., 2004).
Indirect methods allow in situ estimation of the leaf area, not require the leaves to be detached, reduce the variability associated with destructive sampling procedures and allow repeated measurements during the growth period of the plant. They are fast, non-destructive, simple, reliable, inexpensive and susceptible to automation (Keramatlou et al., 2015). In indirect methods, the leaf area is inferred directly from observations of some proxy variables, such as leaf length, leaf width or some combinations of these variables (Fascella et al., 2015). The measured variables constitute inputs for the development of mathematical models to predict the leaf area.
The non-destructive prediction of the leaf area using simple equations has become a common tool in agronomy. In general, leaf area estimation models consist of performing a regression analysis in which the leaf area acts as a dependent variable, and the length and width of the leaf as independent variables (Kumar, 2009). Simple mathematical models allow the measurement of the leaf area in the same plants during the growth period and help reduce variability in experiments (Khan et al., 2016). Different mathematical models can be elaborated for the indirect estimation of the leaf area for several cultivars, species and genotypes, or the same model can be applied for several cultivars and different species.
The review of the scientific literature shows an important number of predictive models of the foliar area for fruit crops. Such profusion requires some systematization for the purpose of offering practical and agile consultation tool. Therefore, the general objective of this study was to carry out a systematic review of the studies on the non-destructive estimation of the leaf area in fruit trees published during the first 20 years of this century (from January 2000 to January 2020).
MATERIALS AND METHOD
For the execution of this study, the guidelines of the standard systematic review were followed, as described in Preferred Reporting Items for Systematic Review and Meta-Analyzes (PRISMA) (Moher et al., 2009). Figure 1 shows the search and selection sequence of the analyzed articles.
Figure 1 Flow diagram of the information through the different phases of the systematic review procedure
Phase 1 - Identification and selection of databases
Specific and multidisciplinary databases were consulted, namely: Agricola, Science Direct, Scielo, Google Scholar, Scopus and Web of Sciences. The search equation was constituted from the terms "leaf area" AND "non-destructive estimation" AND "fruit", NOT "cereal", NOT "vegetable", NOT "ornamental", NOT "medicinal", taking into account that the latter are not precise descriptors of the object of study addressed here. The search covered research published between January 2000 and January 2020. This decision on temporality was due to the fact that other publications (Demirsoy, 2009, Khan et al., 2016) had already analyzed the previous scientific production. Only articles published in academic journals were considered, disregarding books and chapters, doctoral theses, presentations at conferences, popular science magazines, newspapers and commercial publications.
Phase 2 - Exclusion of duplicate articles and filtering the initial results
The initial results were exported to the EndNote X.7 software package for processing. First, a filter was carried out to identify duplicate results. The resulting articles were examined from the information contained in the fields title, abstract, keywords, type of article, publication date and language. As a result of this analysis, a second filtering process was carried out, excluding brief reports, conference proceedings, letters, essays and works written in languages other than English, Spanish or Portuguese.
Phase 3 - Analysis of preliminary results
The remaining articles were qualitatively analyzed. The inclusion criteria adopted in this instance were established based on three axes: (a) measured variables (leaf length, leaf width, etc.); (b) measurement instruments used (millimetric ruler, leaf area meter, etc.); (c) estimation model proposed. As a result of this analysis, publications were eliminated which, even having the terms “leaf area” and “non-destructive estimation” among their descriptors, titles and abstracts, measured variables such as leaf dry weight; they used sophisticated and unusual measuring instruments, and implemented methodologies to estimate leaf area that were not available since the beginning of this century, such as artificial neural networks.
Phase 4 - Quantitative-qualitative synthesis of articles included
The 66 publications selected through the filtering processes were analyzed in order to achieve the proposed objectives. For this purpose, a database was created with the following categories: (a) common and scientific name of the fruit tree; (b) variety / cultivar; (c) type of proposed model (simple linear, quadratic, cubic, exponential, polynomial); (d) R2 value; (e) bibliographic reference; (f) continent / country where the study was developed; (g) year of publication of the article; (h) variables measured to develop the model (leaf width, leaf height, central nervure length, lateral nervure length, mean primary leaf area per shoot, number of leaves, age of plants, etc.); (i) measuring instrument used to estimate leaf area; (j) validation of the model in the same study.
RESULTS
Next, the general characteristics of scientific production on non-destructive methods of leaf area estimation in fruit trees are presented and analyzed. Table 1 presents the following data: scientific and common name of the fruit tree, cultivars analyzed, models developed, regression coefficient obtained, and bibliographic reference. The remaining aspects are discussed qualitatively or quantitatively.
Table 1 shows that non-destructive estimation models of leaf area have been developed in the last 20 years for more than 40 fruit plants and for almost 90 different cultivars. Linear models (62%) predominate over polynomials (25%) and exponentials (13%) models, with high regression rates (varying from 0.77 to 1.00) which indicates the goodness of fit of the proposed models. Regarding the origin of the studies, it is observed that South America is the region in which the most research (51%) have been conducted during the period considered, followed by Asia, Europe and North America, in that order. A detailed analysis shows that Brazil emerges as the country in South America where the greatest number of investigations have been carried out (45%), leaving the remaining 6% distributed among Colombia, Argentina, Cuba and Costa Rica. In Asia, India and Iran stand out as prolific countries (27%) in the production of scientific knowledge on the subject, while Europe and North America together represent 22% of the total.
With respect to the periodicity of the publications it is observed that the research has developed in a markedly irregular way. In this sense, an upward trend has been observed since the beginning of the 21st century, reaching its maximum level in 2012. This boom the publications was followed by a saw-type distribution with peaks corresponding to the years 2015 and 2017. In the last two years, there has been a significant drop in the number of publications on the subject.
When the strictly operational aspects of the articles that make up the portfolio studied are considered, it is noted that regarding the variables measured, the length and width of the leaves (84%) predominate; regarding measuring instruments used, protrude portable leaf area meter, type LI-COR, model 3000 or 3100 (41%); in turn, Excel, SAS and SPSS emerge as the main computer programs used for the calculation of the regression equations and the development of the models. Finally, it should be noted that there is a very important percentage of studies (77.50%) that have validated the model, either with a new sample of the same fruit trees or with different cultivars to those used to develop the model.
DISCUSSION
The general objective of the study was to provide a systematic review of articles on non-destructive estimation of leaf area in fruits, published between January 2000 and January 2020. For this purpose, a bibliographic portfolio, composed of 66 scientific articles that responded to the defined search profile, was examined. The detailed analysis of the selected articles offers an overview of the most recent scientific evidence on this problem, being able to become a reference material for technicians, professionals and researchers interested in this subject. Because it is a descriptive-retrospective work, the study carried out includes the main trends that have characterized the published research on non-destructive estimation of leaf area in fruits during the last 20 years, which can be summarized as follows:
most of the studies have been carried out in emerging or developing countries. Possibly because they are relatively inexpensive studies, which do not require sophisticated instruments or facilities equipped with the latest technology,
models have been developed for almost all fruit crops, from the most popular to the least traditional, such as durian (Kumar et al., 2017). What gives horticultural researchers a considerable competitive advantage to the equip them with simple and reliable methods to measure leaf area in a non-destructive way,
most models developed employing as input variables proxy such as the length and/or width of leaves (Rouphael et al., 2010). In all cases, these are simple and non-destructive methodologies that preserve the canopy, allowing the reuse of the same leaves. This also makes it easier to measure the leaf area of the same plants during the growth process (Kalacska et al., 2005),
in general terms, the proposed models are based on simple equations, which can be calculated with statistical resources available to all users. This transforms them into ductile tools, easy to handle and reliable when making decisions,
finally, it is noteworthy that almost all the models analyzed in this review have been validated within the same study. This gives them a strong consistency between the observed and estimated leaf area, increasing prediction accuracy.
CONCLUSSION
Leaf area is a key parameter in various agronomic processes and physiological studies (Fanourakis et al., 2016). Among other aspects, it plays a key role on the size and filling of the fruit (Keramatlou et al., 2015). For this reason, having simple, accurate and inexpensive methodologies to estimate the foliar area is key in the physiology of fruit trees. This study presents a review of 66 models, developed around the world during the last 20 years, to measure the foliar area in fruit crops in situ and in a non-invasive or non-destructive way. It is expected to constitute a useful contribution to technicians, academics and researchers interested in this topic.
