. 2023 Jun 5;17(6):e0011407. doi: 10.1371/journal.pntd.0011407

Table 2. Effect estimates and P values for each of the terms in our generalized additive mixed models.

Response	Variable	Estimate	Lower95	Upper95	Sig	edf	Chi_or_F	P
Infection	Intercept	2.237	0.714	8.829				0.204
	Rural area ¹	1		1
	Urban area¹	0.6	0.388	0.973	*			0.03
	Urban_rural_interface¹	0.607	0.296	1.265				0.166
	Bad carcass quality²	1		1
	Good carcass quality²	1.355	0.71	2.591				0.384
	Intermediate carcass quality²	1.38	0.62	2.563				0.387
	Alouatta ³	1		1
	Callicebus ³	0.37	0.104	1.214				0.122
	Callithrix ³	0.124	0.046	0.339	*			<0.001
	Date				*	6.299	39	<0.001
	Longitude					0.766	9	0.09
	Latitude				*	2.17	9	0.001
	Long:Lat				*	4.189	16	0.002
Intensity	Intercept	0.782	0.363	1.253	*			0.001
	Rural area ¹	0
	Urban area¹	-0.636	-0.863	-0.465	*			<0.001
	Urban_rural_interface¹	-0.396	-0.77	-0.062	*			0.028
	Bad carcass quality²	0
	Good carcass quality²	0.325	-0.043	0.702				0.079
	Intermediate carcass quality²	0.195	-0.187	0.565				0.307
	Alouatta ³	0
	Callicebus ³	-0.285	-0.71	0.157				0.198
	Callithrix ³	-0.813	-1.138	-0.489	*			<0.001
	Date				*	4.849	21	<0.001
	Longitude					0.576	9	0.15
	Latitude				*	1.784	9	0.001
	Long:Lat				*	6.015	16	<0.001

Response variables included “Infection” given by binary qualitative liver RTqPCR results, denoting the presence or absence of YFV RNA, and “Intensity” given by RTqPCR Cq threshold values representing the intensity of YFV RNA load in infected individuals. The estimates for the Infection model are given as odds ratios; the estimates for the intensity model are given as linear effects. For each linear effect, we include the GAMM slope coefficient and the 95% confidence intervals on estimates (upper 95 and lower 95). For these and all smooth terms, we include p values, effective degrees of freedom (edf), and the Chi square (for infection) or F value (for intensity). An asterisk in the column “Sig” denotes whether the variable was statistically significant (P<0.05). Categorical variables are presented as [1] areas of non-human primate sampling (urban, rural or urban-rural interface), [2] the preservation status of non-human primate carcasses (good, bad, intermediate), and [3] non-human primate taxa.