A Comparison of Pathogen Isolation in Culture and Injection–infiltration Bioassay of Citrus Leaves for Detecting Xanthomonas citri subsp. citri. Issue 5 (15th November 2013)

• Record Type: Journal Article
• Authors: Bock, Clive H.; Gottwald, Tim R.; Graham, James H.
• Is Part Of: Journal of phytopathology. Volume 162:Issue 5(2014:May)
• Page Start: 291
• Page End: 301
• Publication Date: 2013-11-15
• Subjects: Plant diseases -- Periodicals; 632
• DOI: 10.1111/jph.12189
• Languages: English
• ISSNs: 0931-1785

Abstract

<abstract abstract-type="main" id="jph12189-abs-0001"> <title>Abstract</title> <p>Citrus canker [caused by <italic>Xanthomonas citri</italic> subsp. <italic>citri</italic> (<italic>Xcc</italic>)] can cause yield loss of susceptible citrus and result in trade restrictions of fresh fruit. For both regulatory purposes and epidemiological studies, accurate detection and quantification of viable inoculum are critical. Two accepted methods used to detect and quantify <italic>Xcc</italic> are injection–infiltration bioassay and culture, but these two methods have not been directly compared using field‐obtained samples. The two methods were compared using washates of lesions taken from fruit, leaves and shoots in a commercial orchard in Florida in 2009–2010 and 2010–2011, with bioassay being the assumed standard. Despite some misclassifications, true positives (sensitivity) and true negatives (specificity) were the dominant classes using culture. False positives for lesions from shoots ranged from 13.1 to 21.4% in 2009–2010 and 2010–2011, respectively, and false positives for lesions from fruit and leaves ranged from 4.3 to 15.7%, in the two seasons, respectively. The false positive rate for culture compared with injection–infiltration bioassay was highest (0.16–0.55), due to more frequent recovery of <italic>Xcc</italic> by culture at ≤10<sup>3</sup> colony‐forming units (CFU) <italic>Xcc</italic> per ml. The false negative rate was consistently lower (0.02–0.21), confirming that in only a few cases did culture fail to detect <italic>Xcc</italic> when it was present. The area under the curve for receiver operator characteristic analysis ranged from 0.80 to 0.97, confirming that culture provided an accurate diagnosis in most cases. There was a higher frequency of lesions from shoots with a CFU ≤10<sup>3</sup><italic>Xcc</italic> compared with lesions from fruit or leaves, making culture more effective at detecting these. The data demonstrate that culture is a reliable way to detect and quantify <italic>Xcc</italic> compared with injection–infiltration bioassay, particularly when the CFU is ≤10<sup>3</sup><italic>Xcc</italic> per ml.</p> </abstract>