Determining sources of variation in yield assessments of fodder beet crops in New Zealand: How many samples are needed?
2015
Gibbs, SJ | Hodge, S | Saldias, B | Walsh, D | Hastings, C | Williams, N | Davis, P | Trotter, Craig | de Ruiter, JM | Waugh, S | Williams, D
The rapid increase in the area of fodder beet grazed and sold in the South Island of New Zealand has brought to attention the need for improved yield assessment methods. As a precision planted crop with variable germination and plant survival rates, the number of plants per row metre varies significantly. This makes it more difficult to get a representative sample for yield assessment. Currently there are differing methods for performing yield assessments involving row lengths and multiple rows. These methods are used commercially without any published validation. Therefore, the objective of this study was to determine the effects of increasing row length and increasing sampling sites on measurements of plant counts and dry matter (DM) yields. A novel ‘bootstrap’ approach was used to assess variation as row lengths and site numbers increased. The sampling area was within a 70 ha crop with seven sites selected at random within a 1 ha block. At each site, two x 7 m adjacent rows were sampled. The number of plants and total fresh weight were recorded for each metre within each row and DM content determined. The results of this study showed strong variation in plant count and DM yield between individual metre lengths sampled, with lower variation in DM content of the individual plants. There was higher variation within sampling rows than between different sites in plant count, DM yield and DM content. A sequential increase in row length up to 5 metres strongly reduced the variation in plant count and DM yield measured, with stable variation beyond that length. A double row of 5 m row lengths x 5 sampling sites provided estimates of DM yield (15.88±1.13 t DM/ha) that were similar to that produced using all samples in the study (15.38±1.17 t DM/ha), suggesting that the methodology provided a suitable balance between accuracy and increased investment for estimation of DM yield in fodder beet crops.
显示更多 [+] 显示较少 [-]