Contract 2019-006-00 replaces contract 2003-017-00. CR-331720 funds the following activities:
1. Operation of the LLRTP rotary screw trap (RST), located at river kilometer nine of the Lemhi River. The LLRTP RST enumerates total juvenile emigration for spring/summer Chinook salmon and steelhead in the Lemhi River. Changes in the fraction of total annual spring/summer Chinook salmon that emigrate as presmolts (prior to the onset of winter) is a primary metric for the evaluation of habitat restoration within the Lemhi River watershed.
2. Modification of the Dam Branch Occupancy Model (DABOM) to incorporate data from Genetic Stock Identification (GSI). In its current form, DABOM decomposes total adult steelhead and spring/summer Chinook salmon enumerated at Lower Granite Dam (LGR) into population, subpopulation, and tributary estimates of escapement using PIT tag interrogation. The incorporation of GSI will enable similar estimates in locations lacking PIT interrogation infrastructure, and is likely to improve the precision of all upstream abundance estimates.
3. An evaluation of the precision of estimates of Proportion Hatchery Origin Spawners (PHOS) that could be obtained by PIT tagging a subsample of hatchery origin steelhead and spring/summer Chinook salmon at LGR. If implemented, tagging of hatchery origin adults at LGR would enable upstream estimates of escapement using the DABOM model. If feasible, this approach would enable consistent PHOS estimates across all upstream locations with sufficient PIT tag interrogation infrastructure.
4. An analysis of the conditions under which juvenile detection at In-stream PIT Tag Detection Systems (IPTDS) can be reliably used to estimate metrics such as juvenile abundance and/or survival. Briefly, juvenile detection at IPTDS is often biased owing to the fact that many juveniles pass IPTDS in sub-optimal orientation and/or outside the effective read range of IPTDS. This analysis seeks to identify the conditions under which juvenile detection data can produce statistically robust estimates.
5. Automation of post-processing to generate metrics used by the Salmonid Habitat Protocol (SHAP). Briefly, SHAP was developed in an attempt to cost-effectively generate the suite of habitat metrics required by commonly-used models that estimate life-stage specific habitat capacity estimates (e.g., Quantile Regression Forest, Habitat Suitability Index, and Bioenergetics models). SHAP relies on photogrammetry and related techniques to process imagery and generate metrics. Problematically, few of these post-processing tasks are currently automated, requiring substantial labor. We propose to use approaches such as neural-network models to automate post-processing and decrease long-term costs of data reduction.
6. Coordination. Tasks 2-5 (above) will require substantial collaboration with co-managers, sovereigns, regulatory agencies, and action agencies. Ultimately, this collaboration will increase the likelihood that the models and tools generated by this contract will be adopted  and used.