In a model without extrinsic (YEAR and SITE) or intrinsic (AGE and SEX) factors, this difference was highly significant ( χ 2 1 = 11.7, P < 0.001) and retained independent significance when all 4 factors and their interactions were taken into consideration (P/A heligmosomids × P/A other nematodes, χ 2 1 = 4.69, P = 0.030).
A model with all 4 factors as main effects and with 2 interactions (SITE × YEAR and SEX × AGE) plus P/A heligmosomids converged but was marginally non-significant (GLM with negbin errors, BFM for oxyuroid nematodes + P/A of heligmosomids, LR 1,904 = 3.505, P = 0.061).
As above, the relationships for intestinal cestodes ( R 2 = 0.11%) was non-significant, but using this approach, that for larval cestodes ( R 2 = 0.15%) also failed to achieve significance.
Much the same was found when the analyses were confined to non-heligmosomid nematodes (essentially 5 species combined in this taxon, with a collective prevalence of 54.6%; see Table 1 ), although the relationships were weaker, and the abundance of non-heligmosomid nematodes did not achieve significance when intrinsic and extrinsic factors were taken into account.
tianjinensis n = 231, ß = 0.090 ± 0.0609, t 229 = 1.479, P = 0.0703, Adjusted R 2 = 0.0051) covariance was close to significance, but for all cestodes and larval cestodes covariances were not significant.