Predicting the effects of invasive ecosystem engineering species in new bioregions has proved elusive. In part this is because separating biological effects from purely physical mechanisms has been little studied and yet could help predict potentially damaging bioinvasions. Here we tested the effects of a large bio-engineering fanworm Sabella spallanzanii (Sabella) versus worm-like structures (mimics) on gas and nutrient fluxes in a marine soft bottom sediment. Experimental plots of sediment in Hauraki Gulf (New Zealand) were used to test the hypothesis that ecosystem engineers negatively influence benthic ecosystem function through autogenic mechanisms, facilitating activity by biofouling organisms and competitive exclusion of native infauna. Enhanced physical structure associated with Sabella and mimics increased nitrogen fluxes, community metabolism and reduced denitrification from 23 μmol m-2 h-1 to zero at densities greater than 25 m2. Sabella plots on average had greater respiration (29%), NH4 release (33%), and greater NO3 release (52%) compared to mimics, suggesting allogenic (biological) mechanisms occur, but play a secondary role to autogenic (physical) mechanisms. The dominance of autogenic mechanisms indicates that bio-engineers are likely to cause significant impacts when established, regardless of fundamental differences in recipient regions or identity of the introduced bio-engineer. In the case of Sabella spallanzanii, compromised denitrification has the potential to tip the balance of net solute and gas exchanges and cause further ecological degradation in an already eutrophic system.