Tadpole growth performance data

We measured tadpoles’ growth in response to temperature conditions in a complementary set of manipulations conducted in three different experiments. First (Field Enclosures tab), within one catchment we translocated embryos from one source and reared them to metamorphosis in multiple streams with disparate thermal regimes (same experiment as Tsel Trial 1). Second (Common Garden Growth Chambers tab), we reared embryos from different catchments under controlled thermal conditions in the laboratory, and third (Common Garden tab), we used outdoor stream mesocosms to create a common-garden environment (corresponding to Tsel Trial 3). For the Field Enclosures experiment, in 2008 and repeated in 2009, we raised full-sibling tadpoles from shortly after fertilisation until front limb emergence in flowthrough stream enclosures placed in six locations in the catchment of the South Fork Eel River hereafter SF Eel, Appendix S2 in article). Details of animal husbandry methods are in Appendix S3 of article. For each replicate we measured body length and mass of tadpoles weekly and calculated the mean growth rate of all tadpoles in the enclosure (mg/day and mm/day). We used least squares linear regression to relate growth and size at metamorphosis to M30DAT. For the Common Garden Growth Chambers, we collected embryos from three of the coastal study rivers (two with dams) and four of the inland rivers (two with dams) from April to June 2010 (Table S1 of article). We transported embryos in chilled aerated river water to the laboratory and within 8 hr of collection placed them into temperature-controlled diurnal growth chambers (cycle set to 14.5 hr light and warm, 9.5 hr dark and cool). After hatching, we randomly chose five tadpoles from each clutch and placed them in 3.8-L aquaria with an aerated mixture of river water and de-chlorinated tap water (changed 29/week). To mimic warm summer conditions at locations with robust populations of frogs, one chamber was set at 18°C during the dark cycle, and 22°C during the light cycle. Daily mean temperature, calculated from 3 i-Button data loggers placed in randomly selected aquaria, was 19.5°C. To mimic conditions of the coolest occupied sites, the other incubator was set at 13°C dark cycle, 19°C light cycle, with a daily mean = 16.6°C. Each chamber held 35 aquaria (4–6 replicate clutches 9 7 rivers). We measured body and total length of tadpoles with calipers to the nearest 0.1 mm at weekly intervals after hatching. The response variable was body growth, mm/day. For the Common Garden experiment, we collected embryos in late May and early June 2010 from four rivers, and placed them in outdoor mesocosms at the University of California’s Richmond Field Station (37.913536°N, -122.333303°W). We captured tadpoles in four rivers: a coastal unregulated (South Fork Eel), a coastal regulated (Eel below Scott Dam), an inland unregulated (Middle Fork Feather) and an inland regulated (North Fork Feather). At each river, we obtained embryos from three different egg masses; furthermore, at the North Fork Feather we collected from two reaches (below the Cresta and Poe dams). After transferring embryos to Richmond and rearing them to tadpole stage, we placed 10 tadpoles in separate mesocosms consisting of re-circulating troughs (total of 15 troughs, 3 for each river except for the North Fork Feather which had 6 troughs). We mounted PVC pipe cut in half (2.2 m 9 15.3 cm diameter) on saw horses at a slight incline so water drained into a 946 L reservoir. Pumps and hoses continuously re-circulated water to the top of each trough lined with periphyton covered river rocks and loose algae. Tadpoles were reared indoors at 18°C until stage Gosner stage 25, and placed in the troughs on 21 July 2010. We measured and staged tadpoles weekly until front limb emergence (Gosner stage 42). We calculated a mean growth rate for each trough (mm of body length increase/day).