Evaluating monitoring methods for cetaceans

Abstract

With increasing human pressures on wildlife comes a responsibility to monitor them effectively, particularly in an environment of declining research funds. Scarce funding resources compromise the level and efficacy of monitoring possible to detect trends in abundance, highlighting the priority for developing cost-effective programs. A systematic and rigorous sampling regime was developed to estimate abundance of a small, genetically isolated spinner dolphin (Stenella longirostris) population exposed to high levels of human activities. Five monitoring scenarios to detect trends in abundance were evaluated by varying sampling effort, precision, power and sampling interval. Scenario 1 consisted of monthly surveys, each of 12 days, used to obtain the initial two consecutive annual abundance estimates. Scenarios 2, 3 and 4 consisted of a reduced effort, while Scenario 5 doubled the effort of Scenario 1. Scenarios with the greatest effort (1 and 5) produced the most precise abundance estimates (CV=0.09). Using a CV=0.09 and power of 80%, it would take nine years to detect a 5% annual change in abundance compared with 12 years at a power of 95%. Under this best-case monitoring scenario, if the trend was a decline, the population would have decreased by 37% and 46%, respectively, prior to detection of a significant decline. With the potential of a large decline in a small population prior to detection, the lower power level should be used to trigger a management intervention. The approach presented here is applicable across taxa for which individuals can be identified, including terrestrial and aquatic mammals, birds and reptiles.3

Best Practice Framework and Principles for Monitoring the Effect of Coastal Development on Marine Mammals

“…Frontiers in Marine Science | www.frontiersin.org1999; Taylor et al, 2007;Tyne et al, 2016). Information from empirical research on the same or similar species can provide Supplementary Data to inform power analysis.…”

“…Detection probabilities can be influenced by species, sample effort, environmental conditions, time, observers, abundance and sites (Moore and McCarthy, 2016). Determining the probability of detection against costs/budget, time and effort, can assist in determining the necessary trade-offs between these factors to achieve the best outcomes (see Hauser and McCarthy, 2009;Moore and McCarthy, 2016;Tyne et al, 2016). The optimal sampling effort will be driven by the specific objectives and the indicators selected for monitoring.…”

“…As Lindenmayer et al (2013) emphasize, changes in indicators should be rigorously assessed as early as possible to avoid a population decline remaining undetected due to ineffective monitoring or slow and/or limited intervention. For example, survey effort may be required over multiple years to detect declines in populations, however, declines may occur well before sufficient power is ever attained Tyne et al, 2016). Changes to decision-making criteria may be necessary in the absence of sufficient resources or time to TABLE 2 | Examples of potential triggers and indicators for assessment and monitoring programs associated with determining the impact of coastal developments on marine mammals.…”

“…Sensitivity analysis is an alternative to direct measures. Sensitivity analysis can be employed to forecast the probability that changes across time and space will have a certain impact (Tyne et al, 2016). Sensitivity analysis can be used to predict the point at which thresholds of triggers are being approached or crossed and therefore when intervention is required.…”

“…Inappropriate application of methods may also occur (Bejder et al, 2012). For example, making assumptions regarding the presence and abundance of a dolphin species using a low number of samples obtained from a few opportunistic sightings is not likely to confidently determine the impacts of a coastal development (Tyne et al, 2016).…”