Mathematical methods to compare growth curves of Pythium arrhenomanes and eleven other Pythium species as an aid to their identification

Abstract

A mathematical function with parameters maximum growth rate, optimum temperature and shape of the temperature/growth curve was used to compare growth curves of 12 Pythium species ( P. aristosporum, P. arrhenomanes, P. coloratum, P. graminicola, P. irregulare, P. myriotylum, P. periilum, P. pyrilobum, P. rostratum, P. ultimum, P. vanterpoolii and P. volutum ). Sixty-one isolates were compared at 13 temperatures using measurements from live cultures and previously published data. The usefulness of temperature growth relations as an adjunct to identification appears to be species dependent. Using MANOVA to compare growth curves was a valid technique for identification of non oogonia-forming isolates of P. arrhenomanes from rice but was unsuitable for testing for species differences because within-species variation was often as great as between-species variation. Results were confirmed by principal component analysis (PCA). Due to the clustering of temperature vectors on the biplot, PCA using data from only five temperatures (4°, 28°, 31°, 34° and 37 °C) was adequate to distinguish non oogonia-forming isolates of P. arrhenomanes. P. arrhenomanes isolates from wheat and some of the P. arrhenomanes literature data had growth curves similar to the rice isolates but others more closely resembled P. aristosporum . The lack of clear distinction between P. arrhenomanes and P. graminicola was emphasized by the spread on the principal component biplots with only some isolates clustering with P. arrhenomanes . PCA and biplot methods were also unable to separate species into distinct clusters but this might be achieved by using larger numbers of isolates. The technique based on temperature/growth response can be used to separate the morphologically similar species P. arrhenomanes, P. aristosporum, P. myriotylum and P. volutum .