TABLE 1.

Pairwise contrasts of the marginal estimated means of recovered loci per sample between multi‐kmer and single‐kmer strategies and between both multi‐kmer strategies

Pairwise comparisons among strategies	Completeness of alignments	WGS 10× read depth		Target sequence capture
		Estimate	p‐value	Estimate	p‐value
abyss Multikmer ‐ abyss 21	Unprocessed	136	<.0001	126	<.0001
abyss Multikmer ‐ abyss 33		99	<.0001	120	<.0001
abyss Multikmer ‐ abyss 55		143	<.0001	115	<.0001
abyss Multikmer ‐ abyss 77		191	<.0001	104	<.0001
abyss Multikmer ‐ spades		68	<.0001	85	<.0001
spades ‐ abyss 21		68	<.0001	41	.012
spades ‐ abyss 33		31	.1311	35	.0535
spades ‐ abyss 55		75	<.0001	30	.1386
spades ‐ abyss 77		123	<.0001	19	.6389
abyss Multikmer ‐ abyss 21	Processed	117	<.0001	107	<.0001
abyss Multikmer ‐ abyss 33		81	<.0001	102	<.0001
abyss Multikmer ‐ abyss 55		129	<.0001	100	<.0001
abyss Multikmer ‐ abyss 77		180	<.0001	92	<.0001
abyss Multikmer ‐ spades		27	.2523	44	.006
spades ‐ abyss 21		90	<.0001	63	<.0001
spades ‐ abyss 33		54	.0002	58	<.0001
spades ‐ abyss 55		102	<.0001	56	.0001
spades ‐ abyss 77		153	<.0001	48	.0015

Contrasts are made considering an average of 10× coverage for the low‐coverage WGS and the target sequence capture data sets. Standard error = 12.4 for all pairwise comparisons; degrees of freedom = 477 for all pairwise comparisons