Diving behaviour of Elseya albagula from a naturally flowing and hydrologically altered habitat M. A. Gordos 1 , M. Hamann 2 , C. S. Schauble 2 , C. J. Limpus 3 & C. E. Franklin 1 1 School of Integrative Biology, The University of Queensland, Brisbane, Qld, Australia 2 Queensland Parks and Wildlife Service, Bundaberg, Qld, Australia 3 Queensland Parks and Wildlife Service, City East, Brisbane, Qld, Australia Keywords Elseya albagula; turtle; diving; bimodal respiration; water impoundment. Correspondence Matthew A. Gordos, School of Integrative Biology, The University of Queensland, Brisbane, Qld 4072, Australia. Tel.: +61-2-6626-1395; Fax: +61-2-6626-1377 Email: matthew.gordos@dpi.nsw.gov.au Received 10 September 2006; accepted 14 November 2006 doi:10.1111/j.1469-7998.2007.00289.x Abstract This study investigated the diving behaviour and performance of the bimodally respiring turtle Elseya albagula within the Burnett River, central Qld, Australia. Diving parameters were recorded using pressure-sensitive time-depth recorders for turtles residing within a free flowing versus regulated reach. Maximum submer- gence time recorded for El. albagula (greater than 3 hours) is among the longest recorded for a voluntarily diving vertebrate, and is attributed to the turtle’s ability to respire aquatically. Median dive times logged for El. albagula within the regulated reach (6.7 min) were threefold longer than values recorded for turtles residing within a naturally flowing creek (1.9 min), with discrepancies in dive duration possibly due to the variable hydrologic flows recorded below the weir. No correlation was observed between dive duration and subsequent surfacing inter- vals, suggesting that dives remained aerobic throughout the study. Despite considerable differences in the magnitude and daily variation of water flow between the two locations, similar diel activity and surfacing trends were recorded for El. albagula. Turtles undertook deep resting dives (41.5 m) during the day before moving into shallower habitats (o1.0 m) for the night, while the crepus- cular hours were characterized by elevated surfacing frequencies attributed to periods of increased activity possibly associated with foraging. Introduction One of the main threats facing the ecological integrity of natural rivers is the loss of channel heterogeneity, as pool and riffle sequences are replaced with lentic reservoirs through the construction of dams and weirs (Ward & Stanford, 1989; Tucker, 2000; Bodie, 2001). In addition to loss of habitat, water impoundments also affect downstream biological, physical and chemical processes through altered hydrologic flows (e.g. volume and timing), decreased water quality (e.g. low aquatic temperatures and pO 2 ) and changes to channel morphology and maintenance (Stanford & Ward, 1979; Ward & Stanford, 1989; Tucker, 2000; Bodie, 2001). However, the downstream effect of regulating struc- tures on freshwater turtles has seldom been considered (Dodd, 1990; Vandewalle & Christiansen, 1996; Reese & Welsh, 1998; Moll & Moll, 2000; Tucker, 2000), despite the fact that chelonians represent a diverse and abundant component of lotic systems (Congdon, Greene & Gibbons, 1986). Elseya albagula is a large (adult size; 25–40 cm straight carapace length and weight 1.5–7.7 kg for male and female, respectively) short-neck Australian chelid that displays a distribution limited to the Burnett, Mary and Fitzroy River catchments in central Qld, Australia (Thomson, Georges & Limpus, 2006). Although relatively little is known about the ecology of this recently described species, physiological investigations have demonstrated the ability of El. albagula to respire aquatically, with adult turtles obtaining up to 40% of their total VO 2 via aquatic routes (Mathie & Franklin, 2006). Moreover, examination of El. albagula’s cloacal bursae by Legler & Georges (1993; previously identified as Elseya dentata) and Limpus, Limpus & Hamann (2002) revealed numerous well-vascularized elon- gated bifurcating papillae lining the bursae walls that closely resemble the cloacal gill systems described for Rheodytes leukops, Elseya latisternum (Legler & Georges, 1993) and Elseya georgesi (=El. latisternum of King & Heatwole, 1994). For R. leukops, aquatic O 2 uptake via the cloacal bursae supplements pulmonary stores when the turtle is immersed (Priest, 1997), thereby facilitating aerobic dives of several days in length (Gordos, Franklin & Limpus, 2003a; Gordos, Limpus & Franklin, 2006). However, the effectiveness of aquatic respiration in bimodally respiring turtles is dependent upon prevailing environmental condi- tions such as water temperature, aquatic pO 2 and hydro- logical flows (Gatten, 1980; Stone, Dobie & Henry, 1992a; Prassack, Bagatto & Henry, 2001; Priest & Franklin, 2002; Gordos et al., 2003a; Gordos, Franklin & Limpus, 2004a), three factors that are themselves adversely influenced by Journal of Zoology 272 (2007) 458–469 c  2007 The Authors. Journal compilation c  2007 The Zoological Society of London 458 Journal of Zoology. Print ISSN 0952-8369