Hierarchical stochastic modelling of large river ecosystems and fish growth across spatio-temporal scales and climate models: the Missouri River endangered pallid sturgeon example

Abstract

We present a hierarchical series of spatially decreasing and temporally increasing models to evaluate the uncertainty in the atmosphere-ocean global climate model (AOGCM) and the regional climate model (RCM) relative to the uncertainty in the somatic growth of the endangered pallid sturgeon (Scaphirhynchus albus). For effects on fish populations of riverine ecosystems, climate output simulated by coarse-resolution AOGCMs and RCMs must be downscaled to basins to river hydrology to population response. One needs to transfer the information from these climate simulations down to the individual scale in a way that minimizes extrapolation and can account for spatio-temporal variability in the intervening stages. The goal is a framework to determine whether, given uncertainties in the climate models and the biological response, meaningful inference can still be made. The non-linear downscaling of climate information to the river scale requires that one realistically account for spatial and temporal variability across scale. Our downscaling procedure includes the use of fixed/calibrated hydrological flow and temperature models coupled with a stochastically parameterized sturgeon bioenergetics model. We show that, although there is a large amount of uncertainty associated with both the climate model output and the fish growth process, one can establish significant differences in fish growth distributions between models, and between future and current climates for a given model.

Reproductive strategy, spawning induction, spawning temperatures and early life history of captive sicklefin chub Macrhybopsis meeki

“…Egg and larval drift distance for M. meeki in the lower Missouri River using the laboratory egg and larva development times determined in this study was estimated at 4·58 ± 0·57 days (n = 13; Table II) for at a temperature of 21·30 ± 0·78 ∘ C, which is similar to the mid temperature value during the spawning season (16-28 ∘ C; Grisak, 1996;Dieterman et al, 2006). The channel velocity estimated from discharge during the Macrhybopsis spawning season (March-August) for the past 63 years at Boonville, MO and Nebraska City, NE was 1·18 ± 0·14 and 1·49 ± 0·13 m s −1 , respectively (Wildhaber et al, 2015). Assuming M. meeki egg and larvae are neutral to positively buoyant like H. amarus (Medley & Shirey, 2013) and stayed in the main channel until larvae were swimming horizontally, M. meeki would drift 468 ± 80 km using the Boonville, MO gauge and 592 ± 91 km using the Nebraska City, NE gauge before they would have hatched, absorbed their yolk sac and began swimming horizontally.…”

“…meeki in the lower Missouri River using the laboratory egg and larva development times determined in this study was estimated at 4·58 ± 0·57 days ( n = 13; Table ) for at a temperature of 21·30 ± 0·78° C, which is similar to the mid temperature value during the spawning season (16–28 ° C; Grisak, ; Dieterman et al ., ). The channel velocity estimated from discharge during the Macrhybopsis spawning season (March–August) for the past 63 years at Boonville, MO and Nebraska City, NE was 1·18 ± 0·14 and 1·49 ± 0·13 m s −1 , respectively (Wildhaber et al ., ). Assuming M .…”

Introduction to integrated environmental modelling to solve real world problems: methods, vision and challenges

“…The issues in estimation of uncertainty are further discussed by Wildhaber et al (2015a), who INTRODUCTIONoutline a process of downscaling from regional climate models to the river scale, and the impact of the uncertainties involved in linking hydrological and temperature models with a bioenergetics model (Wildhaber et al 2015b) for the pallid sturgeon species of fish. This work provides an excellent case study of the application of uncertainty considerations.…”