### Meta-analysis of the application effects of a biostimulant based on extracts of yeast and amino acids on off-season corn yield

#### Abstract

The tests were performed with a biostimulant (GAAP) containing yeast extract and amino acids. The yield data of the off-season corn for meta-analysis were collected from 41 trials conducted in the states of Paraná, São Paulo, Mato Grosso, Minas Gerais, and Goiás during the 2013/2014 crop season. The tests consisted of eight treatments, with four replicates per treatment, and were conducted on 3.6 × 6.0 m plots. The treatments consisted of application of biostimulant at 2.0 L ha-1 at different times and the control (no biostimulant). The time of application corresponded to the growth stages, V8, VT, R1, (V8 + VT), (V8 + R1), (VT + R1), and (V8 + VT + R1). The influence of biostimulant application was quantified as the difference in yield, expressed as kilogram per hectare (kg ha-1), between treatments and the control (effect measurements). Meta-analysis was used to study the effects of the treatments and to calculate the probability of yield increase with product use. The meta-analysis was performed using the software R. The random effects model was used for meta-analysis because of the high heterogeneity among the studies. Next, the mixed effect model was applied to explain the high heterogeneity, considering the following subgroups: the number of applications, the timing of applications, the presence of water stress, and the region where the tests were conducted. The probability of yield increase was calculated at the levels of 2, 5, and 10 bags, each of 60 kg ha-1. The meta-analysis results for the variable "General" and the subgroups were significantly positive (p < 0.0001), with a meta-analytic estimate of 342.1 kg ha-1 and the confidence interval for 95% probability ranging between 301.2 kg ha-1 and 383.0 kg ha-1. The probability for yield greater than zero or equal to 2, 5, and 10 bags of 60 kg ha-1 in subgroup "three applications" was 91.7%, 85.4%, 71.0%, and 38.9%, respectively. These same values were estimated at 91.7%, 85.4%, 71.0%, and 39.0% for the variable "applications in V8 + VT + R1"; 79.1%, 69.3%, 50.1%, and 21.1% for the variable "trials under stress condition"; and 84.2%, 75.1%, 57.7%, and 26.9% for the variable "investments made in Southern Brazil," respectively. The meta-analysis of the data from 287 effect measurements generated in 41 trials demonstrated that foliar application of GAAP biostimulant increases corn yield by 342.1 kg ha-1 with 83.7% probability of positive response.

#### Keywords

Foliar application; Forest plot; Meta-analytic estimate; Quantis.

#### Full Text:

PDFDOI: http://dx.doi.org/10.5433/1679-0359.2017v38n4Supl1p2293

**Semina: Ciênc. Agrár.**Londrina - PR

E-mail: semina.agrarias@uel.br