Our objective was to evaluate the long-term sustainability of lake trout Salvelinus namaycush and rainbow trout Oncorhynchus mykiss populations subjected to a range of fishing mortality (F) in Lake Pend Oreille, Idaho, USA, while providing for bull trout Salvelinus confluentus and kokanee Oncorhynchus nerka recovery. In order to achieve our objective, we developed a density-dependent stochastic predator-prey simulation model for the three major predators (lake trout, rainbow trout, and bull trout) on kokanee in Lake Pend Oreille. As F increased from 0.0 to 1.0, lake trout numbers in 2015 declined 90% for gillnetting, 76% for angling, and 48% for trap netting. At fishing mortality rates observed in Lake Pend Oreille during 2006, all methods combined and angling alone suppressed the lake trout population, but not gillnetting or trap netting alone. As F increased from 0.0 to 0.3, rainbow trout numbers in 2015 declined by 38%. Abundance of adult bull trout increased by 5.8% per year during 1996-2006, after implementation of no-kill regulations, which met the Federal Recovery Plan criterion of a stable or increasing trend in abundance. By 2010, total consumption of kokanee by lake trout, rainbow trout, and bull trout would increase by 20% if fishing mortality on lake trout and rainbow trout declined by 30% from 1996 levels, and would decrease by 14% if fishing mortality on lake trout and rainbow trout increased by 30% from 1996 levels. At rates of fishing mortality exerted on lake trout and rainbow trout in 2006, the likelihood of kokanee collapse was 65% within the next decade. Therefore, fishing mortality would need to be at least 6% higher on both lake trout and rainbow trout to reduce the likelihood of kokanee collapse to 50%. We conclude that kokanee biomass is presently out of balance with predation in Lake Pend Oreille, because kokanee production cannot compensate for all predation loss. Our findings suggest that a combination of unusually high kokanee production and unusually low predation are likely needed for kokanee to survive the next decade in Lake Pend Oreille.

We assessed noncompliance with angling regulations on three Idaho waters using random response, a technique designed to quantify embarrassing or criminal behavior. We searched for associations between positive random response answers and angler regulation awareness across a number of demographic variables. Illegal use of bait and creeling of westslope cutthroat trout Oncorhynchus clarki lewisi within two catch‐and‐release zones ranged from –0.4 to 3.0%. Creeling of illegal‐sized cutthroat trout was a more common violation (5 to 8%) in two zones managed with a minimum size regulation. Estimated noncompliance with barbless hook regulations for the same zones was high (29%), but nearly 75% of these violations were accidental. Noncompliance with harvest restrictions was greatest on Henrys Lake where 9.5% of anglers violated the two‐trout creel limit each day. We observed statistically significant associations between the types of regulations and angler ability to correctly recite them on a given stream. Several demographic variables including age, residence, and gear type used were also associated with regulation awareness. We conclude that random response is a viable method for estimating rates of angler noncompliance with regulations. Additional analyses are needed to evaluate potential biological effects of noncompliance on the trout populations.