‘In this world nothing can be said to be certain, except death and taxes’ Benjamin Franklin.
For ecologists, this quote will strike a chord. For some poorly paid ecologists, perhaps it might be the issue of taxes that resonates, but for most it will be the sentiment that death is all pervading in ecological studies. Death, or mortality as it is generally termed, is a fundamental component of many ecological studies: from demographic studies that describe population dynamics to studies that focus on drivers of individual fitness. As such is it no surprise that for many decades ecologists have devised methods to estimate mortality. One approach is to measure the number of individuals of different ages within a population, from which life tables and survivorship curves of the age-specific mortality rate can be derived. This approach produces an estimate of mortality for a population. But sometimes the goal is to record the fate of known individuals and hence assess the drivers of individual fitness. In this case, biotelemetry has long been used as a tool to help estimate mortality. With conventional radiotracking, animals can be located from a few km away using a directional antenna and receiver. Sometimes, other data (e.g. motion) may also be relayed. When the received data indicate that the tag has not moved for a long period, this often implies that the animal has died which may then be confirmed by relocating the tag (e.g. Brand, Vowles & Keith 1975; McLoughlin et al. 2003).
In this issue, Klaassen et al. (2014) extend this approach by using satellite tags to estimate the mortality of three species of raptors migrating from Europe to Sub-Saharan Africa: ospreys (Pandion haliaetus), Marsh harriers (Circus aeruginosus) and Montagu's harrier (Circus pygargus). These birds perform some of the longest animal migrations, with round trips of many 1000s of km, which rival the longest travel distances seen in birds and other taxa (Hays & Scott 2013). Birds were equipped with light (<4% of body mass) solar-powered tags using a flexible harness that has been shown to be very reliable for long-term deployments. Robust procedures were then developed to infer when a migrating bird died: specifically, when the tracks ended abruptly in mid-migration and tag/harness failure could be excluded as contributory factors. In some cases, the cause of mortality was confirmed by local contacts. Perhaps the most startling finding from this work was how many individuals were tracked to their deaths, with 51 of 69 being inferred to have died. These findings show how mortality rates during migration are very high, being over six times the rates when the birds are in non-migratory phases during summer and winter.
Several causes of bird death were inferred: birds flew into power lines; they became exhausted over the ocean and died at sea; and they were killed by hunters. Across species, traversing the Sahara was identified as being particularly challenging. This study paves the way for more in depth analysis of mortality hotspots which may help targeted conservation programmes. Furthermore, this approach of using satellite tracking to infer mortality can potentially be extended to other migratory taxa. For example, we have satellite-tracked large numbers of sea turtles around the world, and sometimes, unusual data are recorded: tags come out of the water (submergence data are relayed form the tags) and travel inland to villages. Follow-up trips confirm harvesting of turtles for human consumption (Hays et al. 2003). In other cases, tracked individuals wash ashore, and the animal is then located and found to have been killed accidentally (e.g. in fishing gear or by propeller strikes). These studies across taxa highlight how for migratory species it is important to consider the threats to individuals once they travel outside protected areas. These protected areas typically cover only small areas and a small part of the range used by long-distance migrants (e.g. Schofield et al. 2013).
A challenge with the use of this technology is to distinguish mortality of the tracked animals versus tag loss or failure. However, increasingly tag attachment procedures are becoming more reliable, and tags are able to relay information which helps identify the cause of tag failure (Hays et al. 2007). So the approach of inferring mortality through satellite tagging is likely to have even greater utility in the future. Pervading throughout the use of biotelemetry to estimate mortality rates is the assumption that the tag itself is not influencing the mortality rate. This issue remains important, particularly for smaller animals travelling large distances (e.g. Saraux et al. 2011; Vandenabeele et al. 2012). So at least three important challenges remain when using satellite tracking to infer mortality: obtaining sufficiently large sample sizes to provide robust mortality estimates (Heisey & Fuller 1985), assessing device impacts so that mortality rates are not impacted by the tags themselves and distinguishing mortality from tag failure. If these challenges can be overcome, then inferring mortality from satellite tracking will have ongoing utility for both life-history and conservation studies.