Summary Large river–floodplain systems are characterised by seasonal flow variability. High flows lead to hydrological connection between the main channel and inundated off‐channel lakes, wetlands and floodplains, which provide essential habitats for riverine biota. We tested the following hypotheses: (i) that crustacean zooplankton are more abundant in connected lentic habitats such as riverine lakes and wetlands and have a different community composition compared with the main channel and (ii) that vegetation structure will moderate abundances of crustacean zooplankton in the lower reaches of the Waikato River, New Zealand. Zooplankton densities in main channel inflows and inundated floodplains showed clear seasonal changes, with cladoceran and copepod abundance peaks occurring in the majority of sites 2–3 weeks following the peak spring discharges (coinciding with the retreat of water from floodplains into the main channel). Mean densities of zooplankton were highest in the inflows originating from riverine lakes (10–20 ind. L−1) where rotifers were dominant. We recorded significantly higher abundances of copepods in peat bog and swamp wetland inflows (c. 5 ind. L−1) relative to the main river channel and riverine lake inflows (0.1–1 ind. L−1). Some lake inflows also had high numbers of cladocerans (5–10 ind. L−1). Inundated floodplains displayed heterogeneity in zooplankton community composition in relation to their structural complexity. Flooded forest accommodated higher numbers of copepods (c. 8 ind. L−1) and cladocerans (c. 17 ind. L−1) than flooded grassland, where zooplankton assemblages resembled those in the main river channel and were characterised by dominance of rotifers and low overall zooplankton densities (1–2 ind. L−1). Our results suggest that seasonal flow and flood pulses, which determine the degree of connectivity of the main channel with the floodplain and off‐channel habitats, govern zooplankton densities and community structure in this large temperate river. Furthermore, the structural complexity of floodplain habitats may play an important role in enhancing riverine zooplankton diversity. We postulate that the post‐flood peak of large‐bodied cladocerans and copepods might have historically played an important role in the provision of food for juvenile fish such as migrating Galaxiidae.

Understanding the mechanisms that facilitate establishment of non-indigenous species is imperative for devising techniques to reduce invasion rates. Passively dispersing non-indigenous organisms, including zooplankton, seemingly invade constructed waters (e.g., ornamental ponds, dams and reservoirs) at faster rates than natural lakes. A common attribute of these invaded water bodies is their relatively young age, leading to the assertion that low biotic resistance may lead to their higher vulnerability. Our aim was to determine if seeding of young water bodies with sediments containing diapausing stages of native zooplankton could accelerate community development, leading to greater biotic resistance to the establishment of new species. Twenty outdoor tanks were filled with water (1,400 L) and nutrients added to attain eutrophic conditions. Ten treatment tanks had sediments added, sourced from local water bodies. In the remaining ten, sediments were autoclaved, and received zooplankton via natural dispersal only. In an initial 12 month monitoring period, species richness increased at a greater rate in the treatment tanks (at 12 months average standing richness per tank = 3.8, accumulated richness = 8.2) than control tanks (2.6 and 5.0, P \ 0.05). Treatment tanks developed assemblages with greater proportions of species adapted to pelagic conditions, such as planktonic cladocerans and copepods, while control tanks generally comprised of smaller, littoral dwelling, rotifers. Analysis of similarities indicated community composition differed between the control and treatment groups at 12 months (P \ 0.01). Two copepod, four rotifer and one cladoceran species were intentionally added to tanks at 12 months. In the 3 month post-introduction period, five of these species established populations in the control tanks, while only two species established in the treatment tanks. The calanoid copepod Skistodiaptomus pallidus, for example, a non-indigenous species confined to constructed waters in New Zealand, established exclusively in tanks where native calanoid copepod species were absent (primarily control tanks). Our study suggests that biotic resistance could play an important role in reducing the establishment rate of non-indigenous zooplankton. It also provides evidence that seeding constructed water bodies with sediments containing diapausing eggs of native species may provide an effective management tool to reduce establishment rates of non-indigenous zooplankton.

The Agdenes peninsula, Sør-Trøndelag, Norway, 1060km2, is a heavily dissected mountainous landscape with numerous small watersheds, of mainly steep gradient, flowing separately into the sea or to fjords. Suitable habitat for permanent beaver occupation occurs mainly as isolated patches within these watersheds. Eurasian beavers were directly reintroduced to the area in 1926 and 1928. The last known individual of this population died in 1961. In 1968-69 2 pairs and a young animal were reintroduced on the Ingdalselva watershed. The current population is descended from these animals, and probably from the later 1990s by immigrants from the adjacent Orkla river system. In 2010-11 the area was surveyed and 24 beaver family group home ranges located, 20 of which were currently active and 4 abandoned; the population size was estimated at about 80 individuals within family territories plus in any year a number of dispersing individuals. Eighteen of the active territories were located on just four watersheds, Ingdalselva and three immediately adjacent to it. The remaining two territories were isolated on different watersheds distant from any other known group, requiring multiple crossings between watersheds and/or considerable movements through salt water to reach from them. Signs of vagrant individuals were found widely, including on a number of watersheds not occupied by any family group, though containing suitable habitat for permanent colonisation. Known data on the date of establishment of each family group is given, and the pattern of recolonisation to date discussed. An isolated population of beavers on a section of the Orkla river system, first noted in 1933, has been attributed to spread from the first study area reintroductions. However, there are grounds to suspect that this population may have had a different origin. Genetic studies would be useful to elucidate this point

No abstract