Skip to main content
. 2015 Jul 29;66(21):6551–6562. doi: 10.1093/jxb/erv359

Table 2.

Comparison of results of application of (a) RootGraph, (b) RTipC, and (c) WinRHIZO software, and manual labelling of primary and lateral root numbers extracted from subsets of images of barley (n=20) and wheat cultivar Kukri (n=18)

The columns show primary (Prim), lateral (Lat), and total (Tot) root counts accumulated over all manually labelled images. The inequalities beneath the barley data (only) refer to the relative error in lateral root count for the given method as experienced across the range of images. Negative values refer to underestimated root counts. As WinRHIZO does not explicitly differentiate between primary and secondary roots, a diameter threshold of 0.338mm subjectively applied to differentiate between primary and lateral roots. As WinRHIZO grossly overestimates root numbers generally, no effort was made to categorize the counts in primary and lateral roots.

Manual RootGraph RTipC WinRHIZO
Prim Lat Tot Prim Lat Tot Prim Lat Tot Prim Lat Tot
Barley 120 1836 1956 120 1888 2018 120 1745 1865 123 4379 4502
−7%<Δε(lat)<14% −22.3%<Δε(lat)<20% 63%<Δε(lat)<211%
Wheat 90 2097 2187 91 2321 2412 54 5450 5504 10493