Minimizing errors in the analysis of dive recordings from shallow-diving animals

Knowledge of the diving behaviour of aquatic animals expanded considerably with the invention of time-depth recorders (TDRs) in the 1960s. The large volume of data acquired from TDRs can be analyzed using dive analysis software, however, the application of the software has received relatively little attention. We present an empirical procedure to select optimum values that are critical to obtaining reliable results: the zero-offset correction (ZOC) and the dive threshold. We used dive data from shallow-diving coastal dugongs (Dugong dugon) and visual observations from an independent study to develop and test a procedure that minimizes errors in characterizing dives. We initially corrected the surface level using custom software. We then determined the optimum values for each parameter by classifying dives identified by an open-source dive analysis software into Plausible and Implausible dives based on the duration of dives. The Plausible dives were further classified as Unrecognized dives if they were not identified by the software but were of realistic dive duration. The comparison of these dive types indicated that a ZOC of 1 m and a dive threshold of 0.75 m were the optimum values for our dugong data as they gave the largest number of Plausible dives and smaller numbers of other dive types. Frequency distributions of dive durations from TDRs and independent visual observations supported the selection. Our procedure could be applied to other shallow-diving animals such as coastal dolphins and turtles.     

Diving pattern and performance in the macaroni penguin Eudptes chrysolophus

The pattern and characteristics of diving in two female macaroni penguins Eudyptes chrysolophus was studied, during the brooding period, using continuous-recording time-depth recorders, for a total of I8 days (15 consecutive days) during which the depth, duration and timing of 4876 dives were recorded. Diving in the first 11 days was exclusively diurnal, averaging 244 dives on trips lasting 12 hours. Near the end of the brooding period trips were longer and included diving at night. About half of all trips (except those involving continuous night-time diving) was spent in diving and dive rate averaged 14–25 dives per hour (42 per hour at night). The duration of day time dives varied between trips, and averaged 1.4–1.7 min, with a subsequent surface interval of 0.5–0.9 min. Dive duration was significantly directly related to depth, the latter accounting for 53% of the variation. The average depths of daytime dives were 20–35 m (maximum depth 11 5 m). Dives at night were shorter (average duration 0.9 min) and much shallower (maximum 11 m); depth accounted for only 6% of the variation in duration. Estimates of potential prey capture rates (3–5 krill per dive; one krill every 17–20 s) are made. Daily weight changes in chicks were directly related to number of dives, but not to foraging trip duration nor time spent diving. Of the other species at the same site which live by diving to catch krill, gentoo penguins forage exclusively diurnally, making longer. deeper dives; Antarctic fur seals, which dive to similar depths as macaroni penguins, do so mainly at night.     

Diving behaviour of the Shy Albatross Diomedea cauta in Tasmania: initial findings and dive recorder assessment

The diving behaviour of the Shy Albatross Diomedea cauta was investigated using archival time-depth recorders (TDRs) and maximum depth gauges (MDGs). Data from birds carrying multiple devices and from diving simulations indicated that the degree of correspondence between TDRs and MDGs varied with the dive depth, duration and frequency, as well as with body placement. The MDGs were the most reliable when the diving depth was greater than 0.5 m, when the diving frequency was low and when gauges were placed on the birds' backs. The TDRs were used during late incubation and early chick rearing in 1994. Fifty-two dives (0.4 m) were recorded during 20 foraging trips of 15 individuals. The majority of dives were within the upper 3 m of the water column and lasted for less than 6 s. However, dives to 7.4 m and others lasting 19 s were recorded. The albatrosses dived between 07.00 h and 22.00 h, with peaks in their diving activity near midday and twilight. Mean diving depth varied throughout the day. with the deepest dives occurring between 10.00 h and 12.00 h. Two dive types were identified on the basis of the relationship between dive depth and descent rate. Plunge dives were short (5 s), and the birds reached a maximum depth of 2.9 m. Swimming dives were both longer and deeper. The characteristics of Shy Albatross plunge dives were similar to those of gannets Morus spp., which are known to be proficient plunge divers. Swimming dives suggest that Shy Albatrosses actively pursue prey underwater.     

Diving pattern and performance in relation to foraging ecology in the gentoo penguin, Pygoscelis papua

We present data on diving pattern and performance (dive depth, duration, frequency and organization during the foraging trip) in gentoo penguins Pygoscelis papua , obtained using time-depth recorders ( n = 9 birds, 99 foraging trips). These data are used to estimate various parameters of foraging activity, e.g. foraging range, prey capture rates, and are compared in relation to breeding chronology. Foraging trip duration was 6 h and 10 h, and trip frequency 1.0/day and 0.96/day, during the brooding and creche periods, respectively. Birds spent on average 52%of each foraging trip diving. Dive depth and duration were highly bimodal: shallow dives (< 21 m) averaged 4 m and 0.23 min, and deep dives (> 30 m) 80 m and 2.5 min, respectively. Birds spent on average 71%and 25%of total diving time in deep and shallow dives, respectively. For deep dives, dive duration exceeded the subsequent surface interval, but shallow dives were followed by surface intervals 2–3 times dive duration. We suggest that most shallow dives are searching/exploratory dives and most deep dives are feeding dives. Deep dives showed clear diel patterns averaging 40 m at dawn and dusk and 80–90 m at midday. Estimated foraging ranges were 2.3 km and 4.1 km during the brood and creche period, respectively. Foraging trip duration increased by 4 h between the brood and creche periods but total time spent in deep dives (i.e. time spent feeding) was the same (3 h). Of 99 foraging trips, 56%consisted of only one dive bout and 44%of 2–4 bouts delimited by extended surface intervals > 10 min. We suggest that this pattern of diving activity reflects variation in spatial distribution of prey rather than the effect of physiological constraints on diving ability.     

Diving behaviour in relation to water temperature in the southern elephant seal: foraging implications

Summary A time-depth-temperature recorder provided a continuous record of diving by a female southern elephant seal in relation to water temperature for 27 days (1939 dives) after completion of moult. Mean maximum dive depth was 391±2.6 m and the overall maximum was 775 m. Dives lasted on average 17.5±0.09 min. Most dives showed a rapid descent to the discontinuity between the cold surface water and warmer deep water. Consequently the seal spent 57% of its time while diving at a depth of 200–400 m when it may have been foraging. This strongly suggests that the seal was exploiting a food source at the discontinuity between vertically stratified water masses. The water temperature data also indicated that the seal was diving in waters south of the Antarctic Polar Front and at some distance from the northern edge of the pack ice. The seal spent 88% of its time under water. Normal surface intervals between dives lasted an average of 2.1 ± 0.1 min whereas 16 extended surface intervals (>10 min duration) lasted 32.7±4.6 min. Dives were deeper during the day than at night and all but one extended surface interval occurred at night. The pattern of dives was similar to records from northern elephant seals but this is the first study to show how diving behaviour relates to water temperature.     

DIVE DATA FROM SATELLITE TAGS AND TIME-DEPTH RECORDERS: A COMPARISON IN WEDDELL SEAL PUPS

The diving behavior of juvenile Weddell seals, Leptonychotes weddellii , was monitored simultaneously with time-depth recorders (TDRs) and satellitelinked time-depth recorders (SLTDRs). Recovered TDRs provided a complete record of the depth and duration of all dives, while data received from SLTDR tags via the ARGOS satellite system were compressed into the number of dives in each of six depth or duration bins. The dive information from the two types of tags was compared to determine if data compression, processing, and transmission influenced the data received.
While only half of the dive data collected by TDRs was also received from the SLTDR tags, the chance of receiving SLTDR data was independent of when diving occurred, when data was transmitted, and the subsequent dive activity. In addition, the number of dives in each depth and duration bin was an accurate representation of the actual dive behavior. Therefore, SLTDR tags were judged to provide data qualitatively similar to that provided by TDRs. The accuracy of seal locations provided by Service ARGOS was estimated by comparison to Global Positioning System (GPS) locations, and the average position error found to be significantly greater than predicted by Service ARGOS or reported in other studies (LCO locations ± 11.4 km, LC1 ± 5.0 km).     

Diving patterns and performance in the Antarctic blue-eyed shag Phalacrocorax atriceps

The pattern and characteristics of diving of two male blue-eyed shags Phalacrocorax atriceps were studied, using continuous-recording time-depth recorders, for a total of 15 consecutive days during which the depth, duration, bottom time, ascent and descent rates and surface intervals of 674 dives were recorded. Deep dives (> 35 m, averages80–90 m, max. 116 m) were twice as common (64% versus 34%) as shallow dives (< 21 m and 90% < 10 m). Deep dives were long (averages 2.7-4.1 min, max. 5.2 min) with half the time spent near maximum depth and fast travel speeds (averages 1.0-2.4 m s⁻¹). Shallow dives were short (average 0.5 min, max. 1.3 min), without bottom time and with slow travel speeds (0.1–0.6 m s⁻¹). The time spent at depth and the diet (mainly benthic fish and octopus) is consistent with benthic foraging; the function of shallow dives is uncertain. Male shags forage mainly in the afternoon in3–5 distinct bouts of diving. Within bouts (and shorter homogeneous sequences of diving) surface intervals are consistently2–3 times the preceding dive duration; in other shags the reverse is the case. Blue-eyed shag diving depth, duration and pattern is extreme amongst shags; and the relationship between dives and surface intervals suggests that they may regularly exceed their aerobic dive limit.     

Diving behaviour, dive cycles and aerobic dive limit in the platypus Ornithorhynchus anatinus

We investigated the diving behaviour, the time allocation of the dive cycle and the behavioural aerobic dive limit (ADL) of platypuses (Ornithorhynchus anatinus) living at a sub-alpine Tasmanian lake. Individual platypuses were equipped with combined data logger-transmitter packages measuring dive depth. Mean dive duration was 31.3 s with 72% of all dives lasting between 18 and 40 s. Mean surface duration was 10.1 s. Mean dive depth was 1.28 m with a maximum of 8.77 m. Platypuses performed up to 1600 dives per foraging trip with a mean of 75 dives per hour. ADL was estimated by consideration of post-dive surface intervals vs. dive durations. Only 15% of all dives were found to exceed the estimated ADL of 40 s, indicating mainly aerobic diving in the species. Foraging platypuses followed a model of optimised recovery time, the optimal breathing theory. Total bottom duration or total foraging duration per day is proposed as a useful indicator of foraging efficiency and hence habitat quality in the species.     

Diving at the shallow end: Green turtle behaviour in near-shore foraging habitat

Green turtles Chelonia mydas of immature and adult size (n = 19, curved carapace length 49 to 118 cm) were equipped with time-depth recorders for short periods (≤ 7 d) to investigate diel and seasonal variation in diving behaviour. Research sessions were distributed over 2 years to cover seasonal variation in sea temperature from 14 °C to 30 °C. Diurnal dives were shallower and shorter than nocturnal dives, with diel patterns also evident in dawn and dusk peaks in occupation of depths within 1 m of the surface, elevated diurnal occupation of depths 1 to 2 m below the surface and elevated nocturnal occupation of depths > 2 m. Dive duration increased as sea temperature decreased, showing strong negative correlation by day and by night. Study turtles made resting dives that were 3 to 4 times longer in median duration, and six times longer in maximum duration, at cool temperatures than they were at warm temperatures, but there was no evidence of winter diapause or location shift to avoid cold water. The large majority of turtles spent 89 to 100% of their time at depths ≤ 5 m below the surface, three individuals did not exceed 3 m and the maximum depth recorded by any turtle was 7.9 m, although deeper water was available. Furthermore, the dive data indicated that study turtles collectively spent more than 80% of their time at charted (low tide) depths of 3 m or less, indicating that they consistently used the shallow margins of the bay where human activities tend to be concentrated, thereby potentially increasing their exposure to anthropogenic threats.     

Long-term monitoring of leatherback turtle diving behaviour during oceanic movements

The diving behaviour of four leatherback turtles (Dermochelys coriacea) was recorded for periods of 0.5-8.1 months during their postnesting movements in the Indian and Atlantic Oceans, when they covered 1569-18,994 km. Dive data were obtained using satellite-linked transmitters which also provided information on the dive depths and profiles of the turtles. Turtles mainly dove to depths < 200 m, with maximum dive durations under 30-40 min and exhibited diel variations in their diving activity for most part of the routes, with dives being usually longer at night. Diurnal dives were in general quite short, but cases of very deep (> 900 m) and prolonged (> 70 min) dives were however recorded only during daytime. The three turtles that were tracked for the longest time showed a marked change in behaviour during the tracking, decreasing their dive durations and ceasing to dive deeply. Moreover, diel variations disappeared, with nocturnal dives becoming short and numerous. This change in turtle diving activity appeared to be related to water temperature, suggesting an influence of seasonal prey availability on their diving behaviour. The turtle diving activity was independent on the shape of their routes, with no changes between linear movements in the core of main currents or looping segments in presence of oceanic eddies.     

Response of dugongs to boat traffic: The risk of disturbance and displacement

Disturbance from boats has been documented for many species of marine mammals, especially cetaceans, but has never been quantified for dugongs. Dugongs depend on seagrass for food. This food mostly occurs in shallow coastal areas where boat traffic is high. Thus there is potential for boats to alienate dugongs from critical habitat areas. Using an overhead video observation system (‘blimp-cam’), we observed the behaviour of focal dugongs during controlled boat pass experiments and while no boats were present. The percentage of time focal dugongs spent feeding and travelling was unaffected by boat presence, the number of boat passes and whether a pass included a stop and restart (pass continuity). The duration, distance and direction of a focal dugong's subsurface behaviour were unaffected by number, continuity or distance of boat passes. However, focal dugongs were less likely to continue feeding if the boat passed within 50 m, than if the boat passed at a greater distance. Mass movements of dugong feeding herds in response to experimental and opportunistically observed boats were timed on 42 occasions but only lasted an average of 122 s. These movements occurred in response to boats passing at a range of speeds, and at distances of less than 50 m to over 500 m. The levels of boat traffic we observed may reduce dugongs' feeding time budget by a maximum of 0.8-6%. Thus at present boats appear unlikely to be having a substantive effect on the energy intake of dugong populations at our study site on the Moreton Banks near Brisbane, Australia. However, boat traffic is likely to increase in this fast growing region, raising concern about the future impact of boats on this and other dugong populations.     

Improving population estimates by quantifying diving and surfacing patterns: A dugong example

Diving animals are available for detection from above the water when environmental conditions are favorable and the animals are near the surface. The number of animals that are unavailable for detection needs to be estimated to obtain unbiased population estimates. The current availability correction factors used in aerial surveys for the dugong (Dugong dugon) allow for variation in environmental conditions but use the average time dugongs spend near the surface (i.e., constant availability corrections). To improve availability estimates, we examined location and dive data from nine dugongs fitted with satellite telemetry units and time‐depth recorders (TDRs) in eastern Australia. The effects of water depth, tidal conditions, and habitat types on dugong surfacing time were examined using generalized linear mixed models (GLMMs). We found that availability for detection differed with water depth, and depth‐specific availability estimates were often lower than the constant estimates. The habitat effect was less influential, and there was no tidal effect. The number of dugongs estimated using depth‐specific availabilities were higher than those obtained using constant availabilities across water depth. Hence, information on water depth can refine availability estimates and subsequent abundance estimates from dugong aerial surveys. The methodology may be applicable to other aquatic wildlife.     

THREE-DIMENSIONAL DIVING BEHAVIORS OF RINGED SEALS (PHOCA HISPIDA)

Dives of five freely diving ringed seals were classified into three-dimentional movement types. Horizontally convoluted dives, defined as dives with angular velocity > 15°/sec, appeared to be foraging or social dives. Simple dives that did not include convoluted movements (angular velocity < 10°/sec) were considered to be exploration dives. Directional dives with nearly linear horizontal travel (horizontal directionality >0.6, on a scale of 0–1) were presumed to be travel dives. Each three-dimensional dive type was observed with similar frequency in dives with two distinct time-depth profiles: V-shaped profiles in which ascent immediately followed descent, and U-shaped profiles in which >7 sec were spent at depth between descent and ascent. The lack of behavioral differences between dives with distinct time-depth profiles suggested that time-depth profiles are not a reliable means of inferring dive behaviors for ringed seals.     

ANALYSIS OF SWIM VELOCITIES DURING DEEP AND SHALLOW DIVES OF TWO NORTHERN FUR SEALS, CALLORHINUS URSINUS

Swim velocities at 15-sec intervals and maximum depth per dive were recorded by microprocessor units on two "mixed diver" adult female northern fur seals during summer foraging trips. These records allowed comparison of swim velocities of deep (>75 m) and shallow (<75 m) dives.
Deep dives averaged 120 m depth and 3 min duration; shallow dives averaged 30 m and 1.2 min. Mean swim velocities on deep dives were 1.8 and 1.5 m/sec for the two animals; mean swim velocities on shallow dives were 1.5 and 1.2 m/sec. The number of minutes per hour spent diving during the deep and shallow dive patterns were 11 and 27 min, respectively.
Swim velocity, and hence, relative metabolic rate, did not account for the differences in dive durations between deep and shallow dives. The long surface durations associated with deep dives, and estimates of metabolic rates for the observed swim velocities, suggest that deep dives involve significant anaerobic metabolism.     

SUMMER DIVING BEHAVIOR OF MALE WALRUSES IN BRISTOL BAY, ALASKA

Pacific walruses ( Odobenus rosmarus divergens ) make trips from ice or land haul-out sites to forage for benthic prey. We describe dive and trip characteristics from time-depth-recorder data collected over a one-month period during summer from four male Pacific walruses in Bristol Bay, Alaska. Dives were classified into four types. Shallow (4 m), short (2.7 min), square-shaped dives accounted for 11% of trip time, and many were probably associated with traveling. Shallow (2 m) and very short (0.5 min) dives composed only 1% of trip time. Deep (41 m), long (7.2 min), square-shaped dives accounted for 46% of trip time and were undoubtedly associated with benthic foraging. V-shaped dives ranged widely in depth, were of moderate duration (4.7 min), and composed 3% of trip time. These dives may have been associated with navigation or exploration of the seafloor for potential prey habitat. Surface intervals between dives were similar among dive types, and generally lasted 1–2 min. Total foraging time was strongly correlated with trip duration and there was no apparent diel pattern of diving in any dive type among animals. We found no correlation between dive duration and postdive surface interval within dive types, suggesting that diving occurred within aerobic dive limits. Trip duration varied considerably within and among walruses (0.3–9.4 d), and there was evidence that some of the very short trips were unrelated to foraging. Overall, walruses were in the water for 76.6% of the time, of which 60.3% was spent diving.     

Recording the free-living behaviour of small-bodied, shallow-diving animals with data loggers

1. Time-depth data recorders (TDRs) have been widely used to explore the behaviour of relatively large, deep divers. However, little is known about the dive behaviour of small, shallow divers such as semi-aquatic mammals. 2. We used high-resolution TDRs to record the diving behaviour of American mink Mustela vison (weight of individuals 580-1275 g) in rivers in Oxfordshire (UK) between December 2005 and March 2006. 3. Dives to > 0.2 m were measured in all individuals (n = 6). Modal dive depth and duration were 0.3 m and 10 s, respectively, although dives up to 3 m and 60 s in duration were recorded. Dive duration increased with dive depth. 4. Temperature data recorded by TDRs covaried with diving behaviour: they were relatively cold (modal temperature 4-6 degrees C across individuals) when mink were diving and relatively warm (modal temperature 24-36 degrees C across individuals) when mink were not diving. 5. Individuals differed hugely in their use of rivers, reflecting foraging plasticity across both terrestrial and aquatic environments. For some individuals there was < 1 dive per day while for others there was > 100 dives per day. 6. We have shown it is now possible to record the diving behaviour of small free-living animals that only dive a few tens of centimetres, opening up the way for a new range of TDR studies on shallow diving species.     

Synchronous diving behavior of Adélie penguins

Synchronizing behavior with other conspecifics has been suggested as serving a function of increased foraging efficiency. However, the potential costs associated with synchronization of behavior have rarely been studied. Adélie penguins Pygoscelis adeliae sometimes dive synchronously in small open waters surrounded by fast sea ice. We examined the diving behavior of three couples and one trio, which were observed to dive synchronously among groups of 12–47 birds for 1.7–4.5 h duration, with time-depth recorders. Timing of diving and surfacing differed slightly between individuals, and one bird tended to initiate diving earlier than the other. Although the duration of the dives differed only slightly between these birds, the maximum depth of the dives differed to a large extent, with one member tending to dive consistently deeper than the other bird in two out of the four cases. Vertical distances between tagged birds in the undulatory phases of the dives (presumed feeding time) were greater than those in the descent and ascent phases, suggesting independent foraging by group members. Duration of the undulatory phase of the dives tended to be shorter in deeper-diving individuals than the others in the synchronously diving group, suggesting a potential cost of reduced feeding time to synchronize diving and surfacing with other birds. A digital video image relating to the article is available at .     

Movement heterogeneity of dugongs, Dugong dugon (Müller), over large spatial scales

Seventy dugongs were fitted with satellite PTTs and/or GPS transmitters in sub-tropical and tropical waters of Queensland and the Northern Territory, Australia. Twenty-eight of the 70 dugongs were also fitted with time-depth recorders. The dugongs were tracked for periods ranging from 15 to 551 days and exhibited a large range of individualistic movement behaviours; 26 individuals were relatively sedentary (moving < 15 km) while 44 made large-scale movements (> 15 km) of up to 560 km from their capture sites. Male and female animals, including cows with calves, exhibited large-scale movements (LSM; > 15 km). Body length of travelling dugongs ranged from 1.9 to 3 m. At least some of the movements were return movements to the capture location, suggesting that such movements were ranging rather than dispersal movements. LSMs included macro-scale regional movements (> 100 km) and meso-scale inter-patch local movements (15 to < 100 km) and were qualitatively different from tidally-driven micro-scale commuting movements between and within seagrass beds (< 15 km). The mean ± S.E. macro-scale movement distance per individual was 243.8 ± 35.4 km (N = 14 individuals that travelled > 100 km), with a mean ± S.E. travel time of 179.8 ± 29.0 h. The mean ± S.E. meso-scale movement distance per individual was 49.7 ± 3.3 km (N = 28 individuals that made movements of 15-100 km), with a mean ± S.E. travel time of 52.3 ± 7.1 h. LSMs were rapid and apparently directed (mean ± S.E. travel speeds for GPS tagged animals; meso-scale movements = 1.3 ± 0.11 km/h, min = 0.3, max = 3.0; macro-scale movements = 1.6 ± 0.16 km/h, min = 0.8, max = 1.3). Tracked dugongs rarely travelled far from the coast (mean ± S.E. max distance = 12.8 ± 1.3 km). Dive profiles from the time-depth recorders suggest that dugongs make repeated deep dives while travelling rather than remaining at the surface, increasing their likelihood of capture in bottom set gill nets. Some animals caught in the high latitude limits of the dugongs' range on the Australian east coast in winter apparently undertook long distance movements in response to low water temperatures, similar to migrational movements by Florida manatees. Our findings that dugongs frequently undertake macro-scale movements have implications for management at a range of scales, and strengthen the aerial survey and genetic evidence for management and monitoring at ecological scales that cross jurisdictions.     

SUCCESSFUL USE OF A TRANSLOCATION PROGRAM TO INVESTIGATE DIVING BEHAVIOR IN A MALE AUSTRALIAN FUR SEAL, ARCTOCEPHALUS PUSILLUS DORIFERUS

This study reports some of the first foraging behavior data collected for male fur seals. A nonbreeding male Australian fur seal, Arctocephalus pusillus doriferus , captured at a commercial salmon farm in southern Tasmania, Australia, was relocated 450 km from the site of capture. The animal was equipped with a geolocating time-depth recorder that recorded diving behavior and approximate location for the 14.4 d that it took the seal to travel down the east coast of Tasmania and be recaptured at the salmon farm. During its time at sea, the seal spent most of its time over the relatively shallow shelf waters. It spent 30% of its time ashore on a number of different haul-out sites. The deepest dive was 102 m and the maximum duration was 6.8 min. "Foraging" type dives made up 31.2% of the time at sea and had a median duration of 2.5 min and a median depth of 14 m. The seal performed these dives more commonly during the latter part of its time at sea, while it was on the east coast. Unlike other fur seal species studied to date, there was no evidence of a diurnal foraging pattern; it made dives at all times of the day and night.