Table 1.

Analysis of variance (F-test or χ² test) was performed across single and co-exposure treatments with shared timing of infection, to answer the following: (a) How does Metschnikowia respond to later arrival of Ordospora? (b) How does Metschnikowia respond to prior infection by Ordospora? (c) How does Ordospora respond to later arrival of Metschnikowia? (d) How does Ordospora respond to prior infection by Metschnikowia?

Response variable	Degree of freedom	χ2/F value	P value
(a) METS early | CO:METS early:ORDO late
Host viability	75	0.6807	0.4093
Parasite infectivity	68	0.9820	0.3217
Parasite growth	42	0.5758	04522
Net output per exposed host	75	0.8109	0.3708
(b) METS late | CO:ORDO early:METS late
Host viability	76	4.4597	0.0347
Parasite infectivity	53	0.0768	0.7817
Parasite growth	38	5.7688	0.0213
Net output per exposed host	76	0.4945	0.0291
(c) ORDO early | CO:ORDO early:METS late
Host viability	78	3.8652	0.0493
Parasite infectivity	58	2.9877	0.0839
Parasite growth	19	0.1618	0.6920
Net output per exposed host	78	6.0996	0.0157
(d) ORDO late | CO:METS early:ORDO late
Host viability	74	2.9155	0.0877
Parasite infectivity	70	1.2013	0.2731
Parasite growth	43	1.2613	0.2676
Net output per exposed host	73	1.2907	0.2596

A generalized linear model was used, assuming a binomial distribution of residuals for host viability of individual Daphnia (0 = early death, 1 = viable host) and infection status of individual Daphnia (0 = non infected, 1 = infected). A general linear model was used, assuming a normal distribution of residuals for parasite growth (rate of spore production per infected host) and the net spore output per exposed host. Significant P values (⩽0.05) are highlighted in bold.