Summary1. Life-history theory predicts that immune responses have evolved in the context of costs and benefits. However, our understanding of the costs of mounting an immune response is limited. 2. Using four species of insects, we tested for metabolic costs of immunity by inducing a short-term immune response and ⁄ or wounding and measuring CO 2 production. Inducing an encapsulation response and ⁄ or wounding raised resting metabolic rate by up to 28% with a strong positive correlation between individual encapsulation response and metabolic rate in Tenebrio molitor, Acheta domesticus, Cotinis nitida and Periplaneta americana. Interestingly, we found that haemolymph removal increased metabolic activity relative to only wounding, suggesting a cost to sampling haemolymph. 3. We also tested for how mounting an encapsulation response and ⁄ or wounding would affect other immune components. We found that inducing an encapsulation response led to increased levels of phenoloxidase and decreased levels of lysozyme, an antimicrobial protein. 4.Our results support the growing evidence that immune responses entail specific energetic and corresponding physiological costs.

To quantify the balance between new production and vertical nitrogen export of sinking particles, we measured nitrate uptake, net nitrate drawdown, ΔO 2 /Ar-based net community production, sediment trap flux, and 234 Th export at a coastal site near Palmer Station, Antarctica, during the phytoplankton growing season from October 2012 to March 2013. We also measured nitrate uptake and 234 Th export throughout the northern western Antarctic Peninsula (WAP) region on a cruise in January 2013. We used a nonsteady state 234 Th equation with temporally varying upwelling rates and an irradiance-based phytoplankton production model to correct our export and new production estimates in the complex coastal site near Palmer Station. Results unequivocally showed that nitrate uptake and net community production were significantly greater than the sinking particle export on region-wide spatial scales and season-long temporal scales. At our coastal site, new production (105 ± 17.4 mg N m À2 d À1 , mean ± standard error) was 5.3 times greater than vertical nitrogen export (20.4 ± 2.4 mg N m À2 d À1 ). On the January cruise in the northern WAP, new production (47.9 ± 14.4 mg N m À2 d À1 ) was 2.4 times greater than export (19.9 ± 1.4 mg N m À2 d À1 ). Much of this imbalance can be attributed to diffusive losses of particulate nitrogen from the surface ocean due to diapycnal mixing, indicative of a "leaky" WAP ecosystem. If these diffusive losses are common in other systems where new production exceeds export, it may be necessary to revise current estimates of the ocean's biological pump.

Dispersal behaviour plays a key role in social organisation, demography and population genetics. We describe dispersal behaviour in a population of African wild dogs (Lycaon pictus) in Kenya. Almost all individuals, of both sexes, left their natal packs, with 45 of 46 reproductively active “alpha” individuals acquiring their status through dispersal. Dispersal age, group size and distance did not differ between males and females. However, only females embarked on secondary dispersal, probably reflecting stronger reproductive competition among females than males. When dispersing, GPS‐collared wild dogs travelled further than when resident, both in daylight and by night, following routes an order of magnitude longer than the straight‐line distance covered. Dispersers experienced a daily mortality risk three times that experienced by adults in resident packs. The detailed movement data provided by GPS‐collars helped to reconcile differences between dispersal patterns reported previously from other wild dog populations. However, the dispersal patterns observed at this and other sites contrast with those assumed in published demographic models for this endangered species. Given the central role of dispersal in demography, models of wild dog population dynamics need to be updated to account for improved understanding of dispersal processes.

Wildlife fences are often considered an important tool in conservation. Fences are used in attempts to prevent human–wildlife conflict and reduce poaching, despite known negative impacts on landscape connectivity and animal movement patterns. Such impacts are likely to be particularly important for wide-ranging species, such as the African wild dog Lycaon pictus, which requires large areas of continuous habitat to fulfil its resource requirements. Laikipia County in northern Kenya is an important area for wild dogs but new wildlife fences are increasingly being built in this ecosystem. Using a long-term dataset from the area's free-ranging wild dog population, we evaluated the effect of wildlife fence structure on the ability of wild dogs to cross them. The extent to which fences impeded wild dog movement differed between fence designs, although individuals crossed fences of all types. Purpose-built fence gaps increased passage through relatively impermeable fences. Nevertheless, low fence permeability can lead to packs, or parts of packs, becoming trapped on the wrong side of a fence, with consequences for population dynamics. Careful evaluation should be given to the necessity of erecting fences; ecological impact assessments should incorporate evaluation of impacts on animal movement patterns and should be undertaken for all large-scale fencing interventions. Where fencing is unavoidable, projects should use the most permeable fencing structures possible, both in the design of the fence and including as many purpose-built gaps as possible, to minimize impacts on wide-ranging wildlife.