A simple, reliable method for long‐term, in‐stream data logger installation using rock‐climbing hardware

Abstract

1. Long-term deployment of in-stream data loggers provides valuable information about stream conditions, particularly at times when streams are difficult to sample manually. However, in streams with high water velocities and substantial shear stress, the detachment of loggers from their installation points is a common problem. Thus, successful logger retrieval requires a durable installation method.2. We present a method for data logger installation using rock-climbing hardware that is simple to assemble, economical and minimally invasive. The method is applicable in streams with exposed bedrock or boulders, which serve as attachment points for loggers. Installed loggers are capable of withstanding extreme environmental conditions.After deploying 18 loggers for approximately 15 months in six different streams with snowmelt-dominated hydrographs, we retrieved 16 of the 18 loggers originally deployed. Where loggers were not recovered, hardware failure at installation sites was not likely the cause.4. In addition to application in streams, the use of rock-climbing hardware may be well-suited to securing long-term data logger installations on rock substrates in any extreme environment. K E Y W O R D Sclimbing hardware, data logger, extreme flows, installation, long-term, river, stream, temperature logger | 685 FOGG et al. Methods in Ecology and Evoluঞon

An inexpensive method for reliable recovery of stream temperature data

“…For example, whereas our recovery rate with metal housings and stakes was 73% premetal detector, Isaak et al (2013) and Olsen (2013) recovered 84% and 85% of loggers from PVC canisters glued to rocks. Fogg et al (2020) observed an 89% recovery rate when loggers were fastened to boulders with rock climbing hardware for 18 months. Our relatively poor recovery rates with traditional methods of retrieval were not unexpected, given the inconspicuous, low profile of the stake and housing we used.…”

“…Specifically, we constructed a mixed‐effects ordinal regression model wherein the dependent variable was the ordinal response of logger “fate” (1 = recovery with GPS, 2 = recovery with photo, 3 = recovery with metal detector, 4 = no recovery) following the sequence of events during recovery. Because the literature suggests that stream slope (Isaak et al 2013), stream size (Mauger et al 2015), and proximity to areas of human use (Fogg et al 2020) might affect logger recovery, we included as fixed‐effects stream slope (Slope; ft/ft), Strahler stream order (Order), distance to the nearest road (Road; feet), and distance to the nearest incorporated place or population center (Incorp; miles). These four continuous metrics were derived from the National Hydrography Dataset (NHDPlus HR; U.S. Geological Survey) and Topologically Integrated Geographic Encoding and Referencing (TIGER) system (U.S. Census Bureau) using ArcGISPro v2.7 (ESRI, Redlands, California, USA).…”

“…Loggers have been tethered or attached to anchors ranging from large boulders, logs, and concrete blocks, to metal stakes and nylon bags filled with rocks (Dunham et al 2005). In recent years, underwater epoxy and even rock‐climbing hardware have been used to attach temperature loggers to large boulders, bridge pilings, and bedrock (Isaak et al 2013; Fogg et al 2020; Zimmerman and Winkowski 2021; McCarty et al 2022). While these newer approaches work well in streams with large boulders or other fixed features to which the loggers can be fastened, such opportunities are not always available.…”

“…Polyvinyl chloride (PVC) is commonly used to construct housings because it is lightweight, easy to work with, and inexpensive (Isaak and Horan 2011; Zeigler et al 2013; Chapin et al 2014). The conspicuous nature of white PVC can be mitigated with camouflage (Isaak et al 2013; Fogg et al 2020), but at the cost of making a logger more difficult to relocate in the future (Dunham et al 2005).…”

“…Across methods for protecting and anchoring water temperature loggers, long‐term (≥1 year) recovery rates generally range from 70–90% (Isaak et al 2013; Olsen 2013; Fogg et al 2020), depending on stream gradient, time at large, risk of human tampering, and other factors (Isaak et al 2013; Fogg et al 2020). Meanwhile, researchers continue to report logger loss or impairment as a reason for data gaps that impede analyses (see Anderson et al 2011; Dieterman et al 2012; Olsen 2013; Johnson et al 2017; Huntsman et al 2020), and jeopardize our ability to use temperature to inform critical management decisions.…”