Previous studies into precision potato production have focussed on describing the variability in total yield within fields. Very little literature has been published on the variability in tuber size, which is a key quality (marketability) criteria in potato production. There is very little understanding of how potato tuber size varies spatially and temporally within production systems. To examine this, intensive mid-season and harvest surveys were conducted in 9 ware production fields over three seasons (2015-17). Information on stem and plant density as well as tuber grades in 5 or 10 mm fractions were collected from 3 m plots. Typically approximately 15-50 digs were performed mid-season to understand general responses within defined management zones, while 100 digs were performed at harvest to characterise the response and to permit variogram analysis. The mid-season digs were always paired with intended harvest dig sites to allow for a temporal analysis of tuber size distribution.
Tuber size distribution within the ware production systems exhibited a strong spatial structure, that was considerably stronger (as determined using the Cambardella Index) than the spatial structure exhibited by yield and stem density. It also had a greater range (average of 250 m c.f. 100 m for yield). It appears from this survey that there is considerable structured variability in tuber size distribution within ware production fields in the UK and an opportunity for growers to better manage this. The results indicated that the percent of marketable yield in the fields also exhibited strong spatial structure and that there was 11 – 37 % of the production outside the desired range in the fields. It also indicates that correctly targeting mid-season digs to assess tuber development – a key part of potato agronomy is critical for correct management.
Plot and lab studies have indicated previously that once tuber initiation has finished (mid-season), then the population distribution is relatively stable and can be modelled during the bulking process to predict under-ground tuber development. This is the first large scale field production data set available to test this hypothesis, which is key to correctly modelling tuber size as well as yield within potato crop models. Initial analysis has indicated that this assumption holds true and more detailed spatio-temporal analysis will be presented in the final paper.
