After long and detailed studies focused on production models developed in universities 1,2,3and adopted by extension services 4,5 it is now time to advance precision agriculture (PA), resulting from the developments made in the mechanical agricultural industry and from the work of scientists who produced a body of 17,756 scientific publications, indexed in the Scopus database from 2000 to 2016, throughout the world 6. The PA mainly involves: reduction of overlaps of fertilizers and pesticides during the distribution; soil tillage managements; precision in the dosage of the inputs for the best production. The theoretical potentialities of the various operating instruments now available are challenged by the “big data” generated from the fields as well as from the environment. However, the complexity of the large number of factors, and the multiple interference between their forcing activities that affect production are incomparably less “precise” than the PA definition suggests. Experimenting with a single sensor (Figure 1) is relatively simple, but it is not straightforward to include the results in the complex system of factors that influence productivity, many of which are unknown or uncontrollable.

The progress in yield raising and homogenizing by use of the fertilization maps may be quickly achieved at the beginning of the process 4,5and this fact is exemplified in the Figure 2.

Since the PA is welcome in rice cultivation, in this work we aimed to study in retrospect the productive consequences of its application on a PA Italian rice farm, oriented to obtain maximum yields. In this paper we are concerned in featuring the trends in the variability parameters, obtaining some lessons from fields.

The PA experience started in 1998 with a harvester which was equipped with a yield meter, and then with a yield monitor, a satellite receiver and a light bar to assist manual drive. The farm now spreads fertilizers using a VRT technology, following prescription maps or real time signals of vigor sensors using an OptRx by AgLeader (Figure 3), which reads both NDVI (wavelengths of 670 nm and 780 nm) and NDRE (730 nm and 780 nm).

Before N-pi administration, a drone was used to assess the mean and probable range of the NDRE readings of the sensor, and to fix the thresholds. Following an RTK signal, herbicides are sprayed by a VRT sprayer, which controls the opening and closing of the nozzles lead from a GPS signal, and yield maps are obtained by an harvester equipped with an AgLeader yield meter. All the technology had to be adapted to the soil (Table 1). The name of the farm, “Palestro” (GPS data of the farm buildings: WGS 84 zone 32 N Lat 45,371488 – Lon 8,298165), comes from the Latin “Palustris”, which means Swampy. All the soil, that is, 249 ha, of which 230 ha are cultivated, is silty-clay, and is flooded from April to August. Moreover, a rice monoculture has been grown on this farm since 1965.

Yield data from each 19x19m2 mesh plot have been elaborated for the 2012-17 interval (No. 1165 ha^-1). Yield repeatability was calculated as a Pearson correlation of the plot-yields within field between consecutive years. Several mixed models (PROC MIXED by SAS 9.0), concerning the yield and NDRE, were applied to the statistical parameters of the fields, whereby the effects of the field or cultivar were considered fixed and the year was taken as the random factor in order to estimate the linear trend. Conversely, when the years were considered fixed, and the fields were considered as the random factor, a least squares solution was found. The covariation of the variability between the different traits was enhanced by fitting a linear or non-linear regression of the estimates. A Friedman’s test for paired samples was used to compare the yield averages of the farm vs. the regional means 7. In a sample field, over a period of two years, the total N (N-tot), and N divided into N-pre-pi and N-pi, was related to the yield by developing a partial correlation with NDRE.

Place: r12; r13 and r23 then r12.3 = (r12 – r13 * r23) / ((1-r13^2)^.5* (1-r23^2)^.5)).

Place: regression b1/2 = covariance12 / stdev2 ^2; regression b2/1 = covariance 12 / stdev1 ^2 then (b1 * b2) = r12^2.

The study is detached from the usual basic models and applications of the PA, but unfortunately there are no detailed references of similar duration and width to be compared. However, PA is about, and interesting lessons, derived from real and not from model domain, featured the evolution of high production rice fields. The PA on this farm increased the yield for the same total input of N, but also its variability was increased – and not reduced, as predicted by the theory – but strengthening the repeatability, a neglected parameter.

The present elaborations of field parcel data suggest that the actual applied N rates are near, or over, the maximum of the parabolic curve that represents the non-linear correlation between N-tot and Yield. Slightly reducing the maximum N rates would probably not affect the maximum yields.

In order to reduce the variability of NDRE and of the yields, the first factor that must be considered is the plant density: having an even density of between 200 and 250 plants m^-2 should be the key to improving the results (Figure 12)

An advantage of the reduction in the N rates at PI, where NDRE is already high, is a reduction in the risk of attacks of Pyriculariagrisea, a fungal disease that is increased by N excesses. This advantage is very complex to be estimated in a farm environment.

All over the paper we have deliberately avoided representing maps of vigor and minute details of production and N-supply. The fact that the repeatability was barely +0.31 over 6 years indicates that the soil factor & N-pre-pi explains less than 10% of the variability, so we would have shown a 90% noise in the maps

This paper underlines that PA still has a huge potential that still must be explored.

The authors wish to thank the Fondazione CRT, Torino – Italy, for the financial support to the scientific activities of the “Accademia di Agricoltura di Torino”.