Since the arrival of humans approximately 1000 years before present (B. P.), New Zealand has lost approximately 80% of its forest cover and along with it, the contribution of wood to our aquatic ecosystems. The aim of this thesis was to undertake a large catchment-scale assessment of LW loadings, spatial distribution and morphological influence in an old-growth indigenous forest to provide some understanding on the natural characteristics of wood that would have been present in many river systems of New Zealand prior to human settlement. The second component of the thesis involved the experimental removal of wood from three small streams in order to provide some insight into what that loss of wood may have meant for fish and aquatic invertebrate communities. In the first part of the study, a catchment scale survey of large wood (LW) was completed in a 5thorder, old-growth forest river system. LW volumes ranged from 59-503 m³ ha⁻¹ and declined down the river system along with the number of LW pieces suspended across the channel and LW influence on channel morphology, whereas piece frequency, number of pieces in debris dams and length increased. Nearly half the pieces were influencing channel morphology, particularly wood accumulation, sediment storage, bank armouring, and pool formation. These key pieces were larger, longer and more stable than average. LW contribution to habitat complexity was highest in the middle to upper sections of the river system. Four key zones of wood distribution and influence were identified in the river system. Zonal boundaries were influenced by changes in transport capacity, fluvial processes and channel morphology. In the second part of the study, a field trial was established in three small forested streams to measure the influence of wood and its experimental removal on channel morphology, and indigenous fish and aquatic invertebrate communities. Prior to wood removal there were no significant differences in the total density of fish between wood pools (pools with wood cover), open pools and riffles. Total fish biomass was marginally significant with most of the fish biomass located in wood pools. At the species level, the density and biomass of banded kokopu (Galaxias fasciatus) and the weights of longfin eels (Anguilla dieffenbachii) were significantly higher in wood pools. Species richness, density and biomass of bluegill bullies (Gobiomorphus hubbsi), torrentfish (Cheimarrichthys fosteri) and the density of redfin bullies (Gobiomorphus huttoni) was highest in riffles. Differences in fish community composition were greatest between riffles and pools, whereas there was considerable overlap between the two pool types. Total invertebrate density was 70% higher in debris dams than riffles prior to wood removal, but this difference was not significant. Densities of Trichoptera (caddisfly) and Plecoptera (stonefly), and five aquatic invertebrate taxa were significantly higher in debris dams which also contained greater numbers of less common taxa (< 1% total catch) than riffles. Only Deleatidium sp. (Ephemeroptera) densities were significantly higher in riffles than in debris dams. Aquatic invertebrate communities in debris dams differed significantly from those in riffles and season had a significant influence on aquatic invertebrate community structure. Removal of wood and associated debris dams from the treatment sections in each of the three streams resulted in a simplified channel morphology, significantly increasing the length and area in riffles and reducing the area of pools. The impact on the fish community was greatest for the two larger fish, banded kokopu and large longfin eels, whose abundance declined in the treatment sections. At the reach scale, only banded kokopu biomass showed a significant decline following wood removal. Invertebrates were less affected by wood removal and associated loss of debris dams. Invertebrate composition in the remaining riffles in the treatment sections had a higher proportion of Ephemeroptera and lower proportions of Trichoptera, Plecoptera and Diptera with fewer rare species than remaining debris dams in the control sections, but there were no discernable effects on invertebrate densities and functional feeding groups at the reach scale. Public perception of wood in waterways is mainly negative and wood is managed primarily to reduce flood damage in New Zealand’s streams. With continued research and advocacy on the environmental benefits, careful planning and judicial use, there is the potential to make better use of wood to rehabilitate and enhance New Zealand’s stream environments. This thesis provides some insight into the contribution of wood to forested stream ecosystems in New Zealand and the implicit losses associated with forest removal. It also contributes to our global understanding on the role of wood, its contribution to habitat heterogeneity and influence on biological communities.