A preliminary examination of an in situ dual dye approach to measuring light fluxes in lotic systems

Light is a critical parameter in aquatic ecosystems, affecting primary production and in situ photochemistry. However, measuring light exposure for suspended particles or dissolved components in a dynamic water column can be challenging with existing Eulerian approaches. Here, we assess the simultaneous deployment of two dyes differing in photolability (rhodamine WT and fluorescein) as a Lagrangian measure of sunlight exposure in a lotic system. Fluorescein is sensitive to light exposure; rhodamine WT is relatively photostable. We examined dye fluorescence at various pH, salinity, and temperature conditions. We also tested dye photolability as a function of pH and wavelength range. In conjunction with this laboratory work, we performed initial field testing of the dual-dye approach in a stream on the north shore of Lake Superior, USA. Irradiation of the dyes using long-pass filters identified wavelengths ≥ 420 nm as responsible for the vast majority of the loss of fluorescein fluorescence, with rhodamine appearing relatively photostable in these short-term studies across the wavelength ranges tested. Dye response to irradiation is pH-sensitive; the dual-dye approach will require additional calibration for acidic or basic waters and should be used with caution in aquatic systems undergoing strong (several pH unit) changes in pH. Field testing showed that the fluorescein to rhodamine WT ratio decreased approximately linearly with light exposure. The dual-dye methodology shows promise as an in situ light sensor applicable to water column species in lotic systems if temperature is recorded, and the pH range is measured and relatively stable (e.g., varies by < 1 unit). *Corresponding author: E-mail: [email protected] Acknowledgments This work was funded by a grant from Minnesota’s Office of the Vice President of Research to E.C.M. and by NSF grant OCE-1235379 to E.C.M. and J.A. and NSF grant OCE-1235005 to K.M. The authors thank Hongyu Li (from LLO) and Luni Sun (from ODU) for assistance in dyePAR calibrations. DOI 10.4319/lom.2013.11.631 Limnol. Oceanogr.: Methods 11, 2013, 631–642 © 2013, by the American Society of Limnology and Oceanography, Inc. LIMNOLOGY and OCEANOGRAPHY: METHODS Laidlaw 1977; Kasnavia et al. 1999; Klonis and Sawyer 2000; Dierberg and DeBusk 2005; Chua et al. 2007; Upstill-Goddard 2008). In order to be a good tracer, these dyes must have high fluorescence quantum yields, high solubility, negligible background concentrations, and be inert to (or predictably and reproducibly affected by) chemical and biological factors (i.e., change in pH, change in temperature, particle sorption, and microbial uptake) within the aquatic system of interest. Most tracer dyes have ionic functional groups, making them soluble in water; however, these functional groups are often pH sensitive; with the resulting change in structure causing changes in dye fluorescence and sorption characteristics (Smart and Laidlaw 1977; Kasnavia et al. 1999). Rhodamine WT (Fig. 1) was specifically designed to be an inexpensive and effective tracer for water flow (Smart and Laidlaw 1977) and is one of the most commonly used dyes for tracer studies. While often considered conservative (UpstillGoddard 2008), it has been shown to sorb to particles, especially organic-rich and metal-oxide coated ones (Vasudevan et al. 2001; Smart and Laidlaw 1977), and must be used with care in particle rich and organic rich systems, such as wetlands (Dierberg and DeBusk 2005). Complicating the sorption issue is that tracer grade rhodamine WT often consists of 2 isomers (Shiau et al. 1993; Vasudevan et al. 2001) that have different sorption characteristics (Vasudevan et al. 2001). Rhodamine WT is somewhat photoreactive (Abood et al. 1969), but its photolytic loss has been reported as less than five percent for an 11.5-d deployment in North Sea surface waters (UpstillGoddard et al. 2001) placing it in a usable range as a conservative tracer in coastal marine systems. Rhodamine WT does show pH sensitivity, with its fluorescence dropping dramatically below pH 5 (Smart and Laidlaw 1977), and its partition into organic phases (and likely sorption to organic particles) increasing with decreasing pH (Kasnavia et al. 1999). The relevant pKa for rhodamine WT in natural waters is between 4.7 and 5.1, which refers to the interchange between either the +1 charged species or the zwitterion and the –1 charged species (Shiau et al. 1993; Kasnavia et al. 1999; Vasudevan et al. 2001). Fluorescein (Fig. 1) is another dye commonly used for aquatic tracer work. It has been shown to exhibit low particle sorption to mineral phases and moderate sorption to organic material (e.g., Smart and Laidlaw 1977), has been used with success in brackish water aquifers, and appears to be unaffected by salinity (Chua et al. 2007). Fluorescein is highly susceptible to photolytic losses (Smart and Laidlaw 1977) and is thus more commonly used in groundwater or aquifer studies rather than in surface water work (Chua et al. 2007). Fluorescein is sensitive to pH, shifting among five to six different structures, with different fluorescence characteristics as the pH changes. Martin and Lindqvist (1975) report pKa1 ([H][H2Fl]/[H3Fl +] where “Fl” represents a deprotonated fluorescein molecule) to be 2.2, pKa2 ([H ][HFl]/[H2Fl]) to be 4.4, and pKa3 ([H +][Fl2–]/[HFl–]) to be 6.7, and further discuss that there are three neutral forms of fluorescein, which have varying light absorption and fluorescence properties. Both fluorescein and rhodamine WT fluorescence have also been shown to be functions of temperature, showing decreases in fluorescence as temperature increases (Smart and Laidlaw 1977). Rhodamine WT is more sensitive to temperature changes than fluorescein (e.g., rhodamine fluorescence doubles (~100% increase) as temperature decreases from 25°C to 0°C while fluorescein fluorescence increases by approximately 10% over the same temperature range, as shown by Smart and Laidlaw 1977, and our own data). Whereas both dyes have often been used as water tracers, the potential of fluorescein and rhodamine WT as light sensors is just beginning to be explored. The first attempt to apply these in a quantitative sense has been to develop inexpensive Eulerian sensors. Bechtold and coworkers (2012) placed multiple vials or plastic bags containing either fluorescein or rhodamine in fixed locations on a stream bed and the loss of fluorescence over time was related to PAR sensor data to determine fluorescence decay rates as a function of light exposure. These dyes are also beginning to be applied as light sensors in pulsed injection approaches (Austin et al. 2010; Welsh 2012; Cullin et al. 2013), but these attempts have not yet addressed critical dye parameters, such as the wavelength range of light affecting fluorescence response and pH effects on dye fluorescence, and they have been applied in more qualitative rather than quantitative approaches. The goal of this study is to further the coupling of fluorescein and rhodamine WT to make a Lagrangian light integrator. We performed preliminary lab testing to evaluate the dyes’ sensitivity to salinity, pH, temperature, and the presence of natural stream dissolved material. We also constrained the integrator’s wavelength range of response and pH effects on dye photodegradation. We then conducted field testing in Amity Creek (Duluth, Minn., USA), located on the north shore of Lake Superior, across a range of water flows that included both base flow and high flow conditions. Finally, we performed calibrations of dye response relative to PAR irradiation. This study, therefore, represents a proof of concept application of the Lagrangian dual dye approach. Materials and procedures Reagents and samples For all lab and field experiments, the following chemicals were used: fluorescein (“Uranine K Liquid,” manufactured by Keystone, item ID 801-073-42, lot number A208F209, where the fluorescein is present as the dipotassium salt) and Keyacid Rhodamine WT Liquid (“rhodamine WT,” manufactured by Keystone, item ID 703-010-27, lot number A207K221). These liquids were viscous, making precise delivery of volumes challenging; thus they were diluted into working solutions of dye in tap water or deionized water, and these solutions were then used for the experiments. For each experiment or deployment, the same working solution was used for initial, exposed, and control samples (as described below). We present concentraMinor et al. Dual dye light measurement in streams