A network meta-analysis of the susceptibility of apple genotypes to powdery mildew under organic management
Introduction
Powdery mildew (Podosphaera leucotricha) is a major disease of apple trees. The fungus affects plant foliage, producing white mycelia and gradually, leaf wrinkling. In more severe cases, the disease can cause substantial defoliation and the fruit to develop russetting (Gubler and Koike, 2011). Besides the impact on the yield and quality of the fruit, frequent presence of the fungus may affect the longevity of the orchard (Marine et al., 2010).
The most common way of controlling powdery mildew is by applying fungicides, such as triazoles. However, their use can be expensive and can lead to detrimental effects on human health and the environment (Rosenberger, 2007). Also, the repeated use of fungicides contributes to the development of resistant strains of the pathogen (Borlaug and Dowsell, 2002).
The use of apple cultivars that are resistant to powdery mildew, on the other hand, may often be the most desirable method to prevent damage by the fungus, as this can avoid the disadvantages associated with other methods of control, without necessarily reducing fruit yield or the attractiveness of the fruit to the consumer.
The susceptibility of different varieties to powdery mildew may be evaluated in the field, where the incidence and severity of the native strains of the pathogen present at the site may be determined in trees exposed to natural infection. Indeed, many researchers have published their findings from these kinds of studies. Nevertheless, the way that the incidence or the severity of powdery mildew is measured varies substantially between these studies (Blazek, 2004; Bus et al., 2006 and Paz-Cuadra et al., 2014).
Moreover, even when the same measurement system is used, at least three difficulties need to be overcome to be able to adequately utilize and generalize the results of any given study. First, these results may be subject to a substantial statistical error. This is especially the case when the sample size is small, or information on the likely sampling error is not provided in the reports of these studies. Second, the virulence of the fungus can change substantially from one year to the next, which complicates using past results to predict outcomes in future years at the same site. Third, results obtained in one site will generally not be directly applicable to other sites, especially when these sites are in other regions or countries. These problems are compounded by the fact that the collection of genotypes that are evaluated generally differs greatly between studies, which prevents an easy corroboration of the findings.
The aim of the present study is to try to overcome the aforementioned difficulties and form a consensus about the resistance of apple genotypes to powdery mildew under organic management conditions by adapting a methodology known as network meta-analysis to an agricultural setting.
Meta-analysis represents a means by which the results of studies which attempt to address the same or similar scientific questions can be combined to obtain a clearer overview of what is currently known in the area of research concerned (Borenstein et al., 2009). The use of meta-analysis is now common in many disciplines that rely on quantitative observational or experimental data (Koricheva et al., 2013; Davis et al., 2014 and Qadir et al., 2017) and interesting extensions to the basic methodology have been proposed (Stanley and Jarrell, 1989; Sterne et al., 2002; Bowater and Escarela, 2013). More specifically, in medical research, the particular variant of meta-analysis that will be adapted for use in the present study, namely network meta-analysis (Bucher et al., 1997; Caldwell et al., 2005 and Tonin et al., 2017), has been recently recognized as being a powerful methodology for combining the results of studies in which only a small collection of the experimental units of interest are evaluated in any given one of the studies under consideration.
However, the application of network meta-analysis, or even just standard meta-analysis, has been limited in agricultural research (Fisher, 2015). Indeed, there are very distinct challenges to be confronted in applying network meta-analysis to agricultural rather than medical research. The present study will therefore attempt to break relatively new ground by exploring the benefits of network meta-analysis in addressing an issue that is of interest in horticultural disease management.
Section snippets
Literature search and inclusion criteria
A systematic search of the Web of Science, PubMed and Google Scholar was conducted for articles published in the 20-year period from January 1998 to December 2017 that report experimental results of the resistance of apple tree genotypes to powdery mildew. For the searches of the Web of Science and PubMed the following combinations of words were sought in the abstracts: ‘apple’ and [‘mildew’ or ‘podosphaera’]. For the search of Google Scholar, the following combinations of words were sought in
Literature search and included studies
The implementation of the systematic search for papers, and the division of studies according to site and year produced 22 studies that satisfied the already-stated inclusion criteria of the meta-analysis. These studies are listed in Table 1 with each being identified by the article in which the study was published, geographical location and year(s) in which the incidence of powdery mildew was evaluated. The number of genotypes evaluated in each study is given in column 6 of this table.
Base study and adjustment factors
The
Interpretation of results
The less well-known genotypes of the 12 genotypes that have the lowest mean standardized incidence are ‘Earlidel’ (‘Delicious’ x ‘Early McIntosh’), ‘Florina-Quérina’ (‘612-1’ x ‘Jonathan’), ‘Lourdes’ (origin unknown), ‘Piros’ (‘Helios’ x ‘Apollo’), ‘Rajka’ (‘Sampion’ x ‘Katka’), ‘Topaz’ (‘Vanda’ x ‘Rubin’), ‘436’ (‘Anna’ x ‘Gala’) and ‘468’ (‘Anna’ x ‘Liberty’). The physical appearance, and the taste and texture for fresh eating of the fruits of 7 of these 8 genotypes could be classified as
Conclusion
Network meta-analysis is a useful methodology for combining the results of studies conducted in different locations or years that evaluate the susceptibility of apple genotypes to powdery mildew under organic management conditions. In particular, by applying this methodology it has been shown that, over the studies included in the meta-analysis, the genotypes ‘Delicious’, ‘Delicious (Bisbee)’, Malus micromalus, ‘Red Fuji (Naga Fu 2)’, ‘Earlidel’, ‘Florina-Quérina’, ‘Lourdes’, ‘Piros’, ‘Rajka’,
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of interest
The authors declare that they have no conflicts of interest to report.
References (25)
- et al.
The results of direct and indirect treatment comparisons in meta-analysis of randomized controlled trials
J. Clin. Epidemiol.
(1997) - et al.
Evaluation of the susceptibility of apple genotypes in Querétaro, Mexico to powdery mildew (Podosphaera leucotricha)
Sci. Hortic.
(2014) - et al.
Relative susceptibility of selected apple cultivars to powdery mildew caused by Podosphaera leucotricha
Plant Health Prog.
(2009) Response to diseases in new apple cultivars from the Czech Republic
J. Fruit Ornam. Plant Res.
(2004)- et al.
Introduction to Meta-Analysis
(2009) - et al.
Perspectivas de la agricultura mundial para el siglo XXI
Manejo Integrado de Plagas y Agroecología
(2002) - et al.
Heterogeneity and study size in random-effects meta-analysis
J. Appl. Stat.
(2013) - et al.
Three years of experience with the apple disease control in an organic orchard
Zemdirbyste– Agric.
(2008) - et al.
A partial diallel study of powdery mildew resistance in six apple cultivars under three growing conditions with different disease pressures
Euphytica
(2006) - et al.
Simultaneous comparison of multiple treatments: combining direct and indirect evidence
Br. Med. J.
(2005)