Using GPS tracking to monitor the breeding performance of a low-density raptor improves accuracy, and reduces long-term financial and carbon costs

Abstract

Traditionally, demographic monitoring of birds has been undertaken by intensive monitoring of nesting sites. However, this is challenging for low-density species, whereby the effort and costs involved in locating and monitoring remote sites can be prohibitive or even bias research findings. We show that Global Positioning System (GPS) tracking can overcome these challenges for a low-density raptor. Field monitoring of martial eagles Polemaetus bellicosus from 2013 to 2021 showed consistently poor breeding performance, with a mean productivity of 0.22 (±0.04) fledged young/pair/year. Using GPS tracking data to infer breeding performance gave a significantly higher productivity of 0.46 (±0.10) fledged young/pair/year. Breeding rate and success were also underestimated by field monitoring. These differences were likely due to logistical constraints of field monitoring, particularly relating to finding alternative nests. Comparing costs between approaches, we estimated that GPS monitoring was financially cheaper than field monitoring per sample after 10 years. Carbon costs per sample were lower for GPS-based approaches than field monitoring from the second year, and over a 10-year period GPS monitoring produced considerable savings (200% less carbon). We recommend that despite high initial costs, for long-term demographic monitoring of low-density species, or where logistical constraints make traditional field monitoring inaccurate, remote monitoring options should be considered.

Keywords: GPS tracking; biodiversity monitoring; carbon emissions; funding; productivity; raptors.

Figure 1.

Map of Kruger National Park (green) in South Africa, showing the distribution of martial eagle nests monitored from 2013 to 2021. Nests were monitored by field monitoring (blue) and remotely by GPS tracking (red). Boxes show approximate areas covered by aerial surveys. No nests in the north of the park (above dashed line) were monitored in 2020 and 2021.

Figure 2.

Comparisons of three measures of martial eagle breeding performance recorded by three methods (i.e. ‘field’ = field monitoring, ‘gps’ = using GPS tracking data to infer breeding performance, ‘both’ = corrected GPS tracking data). Dots ±95% confidence intervals show (a) breeding rate (the proportion of territories where an attempt was recorded), (b) breeding success (proportion of attempts which were successful) and (c) breeding productivity (proportion of monitored territories which were successful). Bars (z-axis) show number of territory years in each group for each analysis. P-values indicating differences in breeding performance between methods are given where they are less than 0.1; no other between-group differences approached significance.

Figure 3.

Tracking data for adult martial eagles which were found to be territory holding. Transparent bars are data excluded from our analyses either due to insufficient annual tracking duration (e.g. eagle 180 012 during 2018), too few average points per day for the given year (e.g. eagle 34 608 during 2020 and 2021), or to avoid pseudo replication when the tracked eagles were part of the same pair (e.g. eagle 3007 is the mate of eagle 3097).

Figure 4.

Cost analysis showing the annual cost per sample for (a) the financial cost and (b) the carbon cost of monitoring martial eagle breeding performance using either fieldwork or remote monitoring via tracking study, or a hybrid approach. See electronic supplementary material, table S1, for full costs breakdown and a reproducible framework for calculating relative costs.