Table 3.

Comparison of trematode infections by statistical methods

		Parasite species	Classical method detection	Duplex PCR detection		McNemar		Cohen’s Kappa
				+	-	X 2 -test	P-value	K	P-value
Ebro samples	2011	C. labracis	+	30	7	6.76	0.01	0.37	<0.001
			-	22	30
		M. obovata	+	24	4	3.06	0.08	0.61	<0.001
			-	12	49
		Double infection	+	3	5	18.27	<0.001	-0.01	0.53
			-	32	49
	2013	C. labracis	+	49	0	63.02	<0.001	0.33	<0.001
			-	65	54
		M. obovata	+	11	5	2.72	0.1	0.49	<0.001
			-	13	139
		Double infection	+	3	1	12.5	<0.001	0.22	0.13
			-	17	147
Otago samples	LP	M. novaezealandensis	+	87	1	11.53	<0.001	0.78	<0.001
			-	16	57
		Philophthalmus sp.	+	16	1	3.13	0.08	0.77	<0.001
			-	7	137
		Double infection	+	6	1	2.29	0.13	0.61	0.003
			-	6	148
	OB	M. novaezealandensis	+	62	1	19.36	<0.001	0.6	<0.001
			-	24	39
		Philophthalmus sp.	+	4	0	4.17	0.04	0.55	0.02
			-	6	116
		Double infection	+	0	1	0.8	0.37	-0.01	0.51
			-	4	121

Comparison of trematode infections in snails detected by the classical method (emission of parasites) and the duplex PCR method. i) Ebro samples: C. labracis, M. obovata and double infections in G. adansonii, from the Ebro Delta (Spain), years 2011 and 2013; and ii) Otago samples: M. novaezealandensis, Philophthalmus sp. and double infections in Z. subcarinatus, from Otago Habour (New Zealand), sampling sites (LP: Lower Portobello Bay, OB: Oyster Bay). Data transformed in two-dimensional contingency tables (2×2). Results from McNemar’s Chi-squared test for paired proportions (χ2) and Cohen’s Kappa Statistic (K) for agreement between both methods. Kappa values can range from <0 (no agreement) to 1 (perfect agreement).