Table 4.

Binary logistic regression results predicting the occurrence of (a) protruding roots, (b) protruding flares, and (c) protruding roots and/or flares. Diameter at breast height and tree height were abbreviated as DBH and H.

		Model statistics				Effects on odds ratio
		X²	R²	Yes%	No%	Intercept	DBH	H	Lean	Pavement width	Open soil area	Setback	Pavemen material (brick)
(a) Protruding roots only	A. moluccanus	22.43	0.365	55.0	89.8	0.336	1.081	1.056	0.904	1.101	0.319	2.083	0.064
	B. ceiba	15.58	0.264	39.3	86.4	0	1.006	1.022	0.979	0.374	3.749	7.102	∞
	C. equisetifolia	15.53	0.247	15.4	99.0	∞	1.146	0.973	0.917	0.923	0	1.242	N/A
	C. burmannii	33.44	0.519	74.2	83.8	0	0.963	1.471	1.277	1.435	10.700	0.850	0.534
	D. regia	14.98	0.251	40.7	87.2	0.105	1.048	0.828	1.105	0.747	1.707	1.046	7.128
	F. altissima	11.88	0.295	93.8	27.3	∞	1.011	1.163	1.169	0.423	0.777	2.007	0
	F. microcarpa	5.01	0.061	100.0	0	3.401	0.998	0.936	1.014	1.144	1.162	1.058	0.724
	M. cajuputi	26.83	0.287	39.4	94.4	0.011	1.086	1.049	0.984	1.095	1.486	1.352	1.840
	S. campanulata	42.85	0.616	73.3	96.4	0	1.241	1.006	0.924	1.424	2.959	0.997	0.089
	All trees	273.84	0.314	38.3	94.2	0.064	1.049	1.014	1.028	0.934	1.269	1.110	0.958
(b) Protruding flares only	A. moluccanus	38.63	0.637	64.3	96.9	0	1.158	0.948	1.112	0.278	32.930	165.600	5.516
	B. ceiba	16.53	0.290	31.8	88.0	∞	1.086	0.921	0.896	0.598	2.170	1.456	0
	C. equisetifolia	35.55	0.409	41.7	94.7	0	1.262	0.927	0.960	1.001	39.110	0.966	N/A
	C. burmannii	12.88	0.239	52.2	86.7	0.196	1.130	1.018	0.942	0.766	0.357	1.475	1.471
	D. regia	17.71	0.284	64.9	64.9	0.047	1.078	1.067	0.990	0.721	1.345	1.174	2.901
	F. altissima	10.68	0.302	100.0	37.5	∞	1.020	1.464	0.950	1.326	0.949	0.708	0
	F. microcarpa	27.65	0.312	95.0	28.0	0.032	1.051	1.083	0.980	1.031	1.524	0.926	7.659
	M. cajuputi	25.11	0.296	12.5	95.9	0.007	1.137	0.988	1.018	1.101	0.713	0.952	2.061
	S. campanulata	39.96	0.518	52.4	94.8	0.002	1.075	1.177	1.080	1.001	12.800	0.901	0.111
	All trees	468.80	0.493	53.8	94.0	0.015	1.102	1.000	1.008	0.972	1287	1.056	1.222
(c) Protruding roots and/or flares	A. molp6uccanus	36.93	0.511	69.0	90.0	0.013	1.120	1.019	0.999	0.858	1.646	12.130	0.227
	B. ceiba	24.74	0.389	66.7	72.7	∞	1.008	1.119	0.899	0.188	7.540	23.080	0
	C. equisetifolia	36.11	0.385	41.9	95.4	0	1224	0.938	0.935	0.910	44.020	1.074	N/A
	C. burmannii	24.53	0.414	81.4	64.0	0.006	1.053	1.387	1.071	0.602	6.987	1.891	1.876
	D. regia	12.73	0.214	77.8	31.0	0.291	1.030	0.987	1.009	0.642	1.954	1.311	5.918
	F. altissima	17.47	1.000	100.0	100.0	∞	1.829	∞	0.565	0	0.008	0	∞
	F. microcarpa	15.37	0.570	100.0	33.3	∞	0.988	0.796	0.742	0	51.100	∞	∞
	M. cajuputi	42.88	0.406	59.1	83.3	0.007	1.111	1.107	0.981	1.033	1.488	1.283	2.457
	S. campanulata	55.91	0.634	69.2	93.1	0	1.088	1.346	1.072	1.118	20.480	1.036	0.056
	All trees	540.1	0.523	69.8	87.9	0.034	1.114	0.980	1.003	0.916	1.758	1.132	0.956

The rates of correct prediction of the presence (Yes%) and absence (No%) of protrusion were provided. Species-specific effects of DBH in centimetres, H, lean angle, pavement width, open soil area, setback, and pavement material on the odds ratio of the occurrence of the 3 scenarios were reported. “N/A” indicated a lack of variation. Significant model statistics and predictors with P < 0.05 were italicised and underlined. For comparison, the pseudo-R² values were supplied. A. alexandrae, L. speciosa, M. × alba, P. serratifolia, and X. chrysanthus were excluded due to small sample size.