Spatial and temporal analysis and explanation of potato variability

Submitted by pytrik.reidsma on Wed, 10/03/2018 - 16:46

Potato yields vary across and within fields in the Netherlands. Until recently, data was limited to understand this yield variability across and within fields. For this research, four years dataset was used that is obtained by Van den Borne. The number of fields that Van den Borne uses annually for the cultivation of around 500 hectares of ware potatoes on around 150 fields. The collection of the data started in 2012. In 2012 only one or two observations were sampled in the fields. This number did increase over years, in 2016 six observations were sampled. The observation spots were chosen based on the median electro conductivity of the field that was determined by the dualem 38 from Van den Borne. The majority of the fields was planted with the variety Fontana.


The aim of this study was to investigate the influence of soil characteristics, crop management and weather conditions on plant characteristics, determined by Van den Borne. The first smaller subject was to investigate which factors together could explain the variability in growth curves of the plant characteristics. The relationships between the plant characteristics during the growing season and the underlying factors (soil, crop management and weather) were investigated. Plant characteristics that were analysed include tuber weight, shoot weight, root weight, underwater weight, number of leaves, number of tubers, number of stems, nitrate content, flavonoid content and chlorophyll content of the leaves. The relationship between each plant characteristic and factor (management, soil, etcetera) was analysed separately. For the different characteristics, different growth curves and therefore different mixed models were used (proc MIXED, GLIMMIX and NLMIXED). The second subject was to investigate the relation between the plant characteristics and to estimate the most optimal curve, with a generalized linear model (HP GENSELECT).


The fit of the models was good except for root weight, underwater weight and flavonoid had a R-square less than 0.512. Especially those results need to be interpreted carefully. Postponing the haulm killing had a positive influence on the tuber growth curve and stem length, due to an increase of radiation duration, but negative on underwater weight. The amount of rainfall positively influenced the yield and number of tubers, but negatively influenced underwater weight. Fertilizer application was very important, in particular potassium and nitrogen. The overall applied potassium fertilizer increased the underwater weight and stem length but had a negative effect on the number of stems. The K50 potassium fertilizer had a positive influence on the tuber growth curve and root weight, which was applied only on fields in 2013. K60 had a positive effect on the number of stems, which was only applied in 2014, 2015 and 2016. Nitrogen application had a positive influence on the root growth, underwater weight, number of tubers and a negative impact on shoot weight. Only sulfasote had a negative impact on the stem length and shoot weight, but positive on the number of stems. KAS had a positive effect on underwater weight, number of tubers, but a negative effect on nitrate content, stem length and compound leaves. Urea resulted in a lower underwater weight, number of stems, but had a positive effect on the nitrate content and yield. A mix of different nitrogen formulated (nitrate, ammonia and urea) fertilizers was better. Planting distance was negatively correlated to tuber weight, so planting the potatoes closer to each other resulted in an increase of tuber weight. There is an optimum for planting distance and seed size. Clay fraction represents the soil type of the fields. More yield is achieved on a heavier soil (more clay) but negatively affected the root weight. A sandy soil contains less water and therefore according to the results the amount of nitrogen in the plant increases. A lower soil conductivity suggested a higher number of tubers. A higher nitrogen content in the soil did cause a decrease in flavonoid. On soils with a low phosphate content, a higher stem density was obtained. Beside the macronutrients boron (micronutrient) caused a reduction in stem length.


Different plant characteristics were related to each other. All the results are therefore related not only for that particular plant characteristic but also for others. To obtain a high yield a high shoot weight and therefore a high stem length, compound leaves and nitrate content were important to intercept radiation and therefore for the yield. The number of stems was important to intercept radiation and a higher stem density implies more tubers. More stems imply also more roots which lead to a higher flavonoid level. A high stem density also results in more tubers, which implies smaller tubers. More shoot weight was related to a higher number of compound leaves and a higher root weight was negatively related to the nitrate content in the leaves and compound leaves. For the management, soil and weather variables an HPGENSELECT procedure is performed. The parameters that were important to obtain the optimal growth curve were: seed distance, total potassium, variety, field wetness, KAS and sulfasote. The most optimal curve was estimated at 133 ton/ ha, based on parameters that were inside the range that were used.


Based on this report, the importance of the radiation, plant available water content (GHG and rainfall) in the soil and plant available nutrients (soil nutrients and fertilizer) showed to be the main factors influencing the potato yield. Further research is needed to investigate the influence of the different fertilizer application, different mixes of fertilizer (also manure), soil conditions and their effect on the potato yield. A higher number of tubers was related to a lower amount of nitrate content in the leaves and compound leaves. Because more stems indicated more tubers and therefore more intraspecific competition. According to the findings flavonoid increases if the plant experiences more stress due to water and or nutrients (less nitrate in the leaves). So, to obtain a high yield with a relatively high underwater weight, a high stem density should be obtained by planting bigger seed potatoes, so the number of tubers would be relatively high.

Droevendaalsesteeg 1, Wageningen
Rick Rasenberg
potato growth, yield variability, farm data
Pytrik Reidsma, Edwin van den Heuvel