How big, how blue, how beautiful! Studying the impacts of climate change on big, (and beautiful) blue whales

Dr. KC Bierlich, Postdoctoral Scholar, OSU Department of Fisheries, Wildlife, & Conservation Sciences, Geospatial Ecology of Marine Megafauna Lab

The SAPPHIRE Project is in full swing, as we spend our days aboard the R/V Star Keys searching for krill and blue whales (Figure 1) in the South Taranaki Bight (STB) region of Aotearoa New Zealand. We are investigating how changing ocean conditions impact krill availability and quality, and how this in turn impacts blue whale behavior, health, and reproduction. Understanding the link between changing environmental conditions on prey species and predators is key to understanding the larger implications of climate change on ocean food webs and each populations’ resiliency. 

Figure 1. The SAPPHIRE team searching for blue whales. Top left) KC Bierlich, top right) Dawn Barlow, bottom left) Dawn Barlow, Kim Bernard (left to right), bottom right) KC Bierlich, Dawn Barlow, Leigh Torres, Mike Ogle (left to right).  

One of the many components of the SAPPHIRE Project is to understand how foraging success of blue whales is influenced by environmental variation (see this recent blog written by Dr. Dawn Barlow introducing each component of the project). When you cannot go to a grocery store or restaurant any time you are hungry, you must rely on stored energy from previous feeds to fuel energy needs. Body condition reflects an individual’s stored energy in the body as a result of feeding and thus represents the foraging success of an individual, which can then affect its potential for reproductive output and the individual’s overall health (see this previous blog). As discussed in a previous blog, drones serve as a valuable tool for obtaining morphological measurements of whales to estimate their body condition. We are using drones to collect aerial imagery of pygmy blue whales to obtain body condition measurements late in the foraging season between years 2024 and 2026 of the SAPPHIRE Project (Figure 2). We are quantifying body condition as Body Area Index (BAI), which is a relative measure standardized by the total length of the whale and well suited for comparing individuals and populations (Figure 3). 

The GEMM Lab recently published an article led by Dr. Dawn Barlow where we investigated the differences in BAI between three blue whale populations: Eastern North Pacific blue whales feeding in Monterey Bay, California; Chilean blue whales feeding in the Corcovado Gulf; and New Zealand Pygmy blue whales feeding in the STB (Barlow et al., 2023). These three populations are interesting to compare since blue whales that feed in Monterey Bay and Corcovado Gulf migrate to and from these seasonally productive feeding grounds, while the Pygmy blue whales stay in Aotearoa New Zealand year-round. Interestingly, the Pygmy blue whales had higher BAI (were fatter) compared to the other two regions despite relatively lower productivity in their foraging grounds. This difference in body condition may be due to different life history strategies where the non-migratory Pygmy blue whales may be able to feed as opportunities arrive, while the migratory strategies of the Eastern North Pacific and Chilean blue whales require good timing to access high abundant prey. Another interesting and unexpected result from our blue whale comparison was that Pygmy blue whales are not so “pygmy”; they are actually the same size as Eastern North Pacific and Chilean blue whales, with an average size around 22 m. Our findings from this blue whale comparison leads us to more questions about how environmental conditions that vary from year to year influence body condition and reproduction of these “not so pygmy” blue whales. 

Figure 2. An aerial image of a Pygmy blue whale in the South Taranaki Bight region of Aotearoa New Zealand collected during the SAPPHIRE 2024 field season using a DJI Inspire 2 drone. 
Figure 3. A drone image of a Pygmy blue whale and the length and body width measurements used to estimate Body Area Index (BAI), represented by the shaded blue region. Width measurements will also be used to help identify pregnant individuals.

The GEMM Lab has been studying this population of Pygmy blue whales in the STB since 2013 and found that years designated as a marine heatwave resulted with a reduction in blue whale feeding activity. Interestingly, breeding activity is also reduced during marine heatwaves in the following season when compared to the breeding season following a more productive, typical foraging season. These findings indicate that fluctuations in the environment, such as marine heatwaves, may affect not only foraging success, but also reproduction in Pygmy blue whales. 

To help us better understand reproductive patterns across years, we will use body width measurements from drone images paired with hormone concentrations collected from fecal and biopsy samples to identify pregnant individuals. Progesterone is a hormone secreted in the ovaries of mammals during the estrous cycle and gestation, making it the predominant hormone responsible for sustaining pregnancy. Recently, the GEMM Lab’s Dr. Alejandro Fernandez-Ajo wrote a blog discussing his publication identifying pregnant individual gray whales using drone-based body width measurements and progesterone concentrations from fecal samples (Fernandez et al., 2023). While individuals that were pregnant had higher levels of progesterone compared to when they were not pregnant, the body width at 50% of the body length served as a more reliable method for detecting pregnancy in gray whales. We will use similar methods to help identify pregnancy in Pygmy blue whales for the SAPPHIRE Project where will we examine body width measurement paired with progesterone concentrations collected from fecal and biopsy samples to identify pregnant individuals. We hope our work will help to better understand how climate change will influence Pygmy blue whale body condition and reproduction, and thus the overall health and resiliency of the population. Stay tuned! 

References

Barlow, D. R., Bierlich, K. C., Oestreich, W. K., Chiang, G., Durban, J. W., Goldbogen, J. A., Johnston, D. W., Leslie, M. S., Moore, M. J., Ryan, J. P., & Torres, L. G. (2023). Shaped by Their Environment: Variation in Blue Whale Morphology across Three Productive Coastal Ecosystems. Integrative Organismal Biology, 5(1). https://doi.org/10.1093/iob/obad039

Fernandez Ajó, A., Pirotta, E., Bierlich, K. C., Hildebrand, L., Bird, C. N., Hunt, K. E., Buck, C. L., New, L., Dillon, D., & Torres, L. G. (2023). Assessment of a non-invasive approach to pregnancy diagnosis in gray whales through drone-based photogrammetry and faecal hormone analysis. Royal Society Open Science10(7), 230452. https://doi.org/10.1098/rsos.230452

Making predictions: A window into ecological forecast models

By Dawn Barlow, PhD Candidate, OSU Department of Fisheries, Wildlife and Conservation Sciences, Geospatial Ecology of Marine Megafauna Lab

“What is the weather going to be like tomorrow?” “How long will it take to drive there, with traffic?” We all rely on forecasts to make decisions, such as whether to bring a rain jacket, when to get in the car to arrive at a certain destination on time, or any number of situations where we want a prediction of what will happen in the near future. Statistical models underpin many of these examples, using past data to inform future predictions.

Early on in graduate school, I was told that “all models are wrong, but some models work.” Any model is essentially a best approximation, using mathematical relationships, of how we understand a pattern. Models are powerful tools in ecology, enabling us to distill complex, dynamic, and interacting systems into terms and parameters that can be quantified. This ability can help us better understand our study systems and use that understanding to make predictions. We will never be able to describe every nuance of an ecosystem. Instead, the challenge is to collect enough information to build an informed model that can enhance our understanding, without over-simplifying or unnecessarily complicating the system we aim to describe. As Dr. Simon Levin stated in his 1989 seminal paper:

A good model does not attempt to reproduce every detail of the biological system; the system itself suffices for that purpose as the most detailed model of itself. Rather, the objective of a model should be to ask how much detail can be ignored without producing results that contradict specific sets of observations, on particular scales of interest.”1

Species distribution models (SDMs) are the particular branch of models that underpin much of my PhD research on blue whale ecology and distribution in New Zealand. SDMs are mathematical algorithms that correlate observations of a species with environmental conditions at their observed locations to gain ecological insight and predict spatial distributions of the species (Fig. 1)2. The model is a best attempt to quantify and describe the relationships between predictors, e.g., temperature and the observed species distribution pattern. For example, blue whale occurrence is higher in areas of lower temperatures and greater krill availability, and these relationships can be described with models3. So, a model essentially takes all the data available, and synthesizes that information in terms of the relationships between the predictors (environment) and response (species occurrence). Then, we can look at the fitted relationships to ask what we would expect from the species distribution pattern when temperature, or krill availability, or any other predictor, is at a particular value. 

Figure 1. A schematic of a species distribution model (SDM) illustrating how the relationship between mapped species and environmental data (left) is compared to describe “environmental space” (center), and then map predictions from a model using only environmental predictors (right). Note that inter-site distances in geographic space might be quite different from those in environmental space—a and c are close geographically, but not environmentally. The patterning in the predictions reflects the spatial autocorrelation of the environmental predictors. Figure reproduced from Elith and Leathwick (2009).

So, if a model is simply a mathematical description of how terms interact to produce a particular outcome, how do predictions work? To make a spatial prediction, e.g., a map of the probability of a species being present, you need two things: a model describing the functional relationships between species presence and your environmental predictors, and the values of your predictor variables on the day you are interested in predicting to. For example, you may need to obtain a map of sea surface temperature, productivity, temperature anomaly, and surface currents on a day you want to know where whales are expected to be. Your model is the applied across that stack of spatial environmental layers and, based on the functional relationships derived by the model, you get an estimate of the probability of species occurrence based on the temperature, productivity, anomaly, and surface current values at each location. By applying the model over a range of values, you can obtain a continuous surface with the probability of presence, in the form of a map. These maps are typically for the past or present because that is when we can typically acquire spatial environmental layers. However, to make predictions for a future time of interest, we need to have spatial environmental layers for the future.

Forecasts are predictions for the future. Recent advances in technology and computing have led to an emergence of environmental and ecological forecasting tools that are being developed around the world to produce marine forecasts. These tools include predictions of the physical environment such as ocean temperatures or currents, and biological patterns such as where species will be distributed in space and timing of events like salmon spawning or lobster landings4. The ability to generate forecast of marine ecosystems is of particular interest to resource users and managers because it can allow them to be proactive rather than reactive. Forecasts enable us to anticipate events or patterns and prepare, rather than having to respond in real-time or after the fact.

The South Taranaki Bight region in New Zealand is an area where blue whale foraging habitat frequently coincides with industry pressures, including petroleum and mineral extraction, exploration for petroleum reserves using seismic airgun surveys, vessel traffic between ports, and even an ongoing proposal for seabed mining5. Static spatial restrictions to mitigate impacts from these activities on blue whales may be met with resistance from industry user groups, but dynamic spatial management6–8 of blue whale habitat could be more attractive and acceptable. The key for successful dynamic management is knowing where and when to put those boundaries; and this is where ecological forecast models can show their strength. If we can predict suitable blue whale habitat for the future, proactive regulations can be applied to enhance conservation management in the region. Can we develop reliable and useful ecological forecasts for the South Taranaki Bight? Well, given that we have already developed robust models of the relationships between blue whales and their habitat3 and have documented the spatial and temporal lags between wind, upwelling, and blue whales9, we feel confident that we can develop forecast models to predict where blue whales will be in the STB region. As we continue working hard toward this goal, we invite you to check back for our findings in the future. So, consider this blog post a forecast of sorts, and stay tuned!  

Figure 2. A blue whale surfaces in front of an oil extraction platform in the South Taranaki Bight, demonstrating the overlap between whales and industry in the region. Photo by D. Elvines.

References:

1.        Levin, S. A. The problem of pattern and scale. Ecology 73, 1943–1967 (1992).

2.        Elith, J. & Leathwick, J. R. Species Distribution Models: Ecological Explanation and Prediction Across Space and Time. Annu. Rev. Ecol. Evol. Syst. 40, 677–697 (2009).

3.        Barlow, D. R., Bernard, K. S., Escobar-Flores, P., Palacios, D. M. & Torres, L. G. Links in the trophic chain: Modeling functional relationships between in situ oceanography, krill, and blue whale distribution under different oceanographic regimes. Mar. Ecol. Prog. Ser. 642, 207–225 (2020).

4.        Payne, M. R. et al. Lessons from the first generation of marine ecological forecast products. Front. Mar. Sci. 4, 1–15 (2017).

5.        Torres, L. G. Evidence for an unrecognised blue whale foraging ground in New Zealand. New Zeal. J. Mar. Freshw. Res. 47, 235–248 (2013).

6.        Hyrenbach, K. D., Forney, K. A. & Dayton, P. K. Marine protected areas and ocean basin management. Aquat. Conserv. Mar. Freshw. Ecosyst. 10, 437–458 (2000).

7.        Maxwell, S. M. et al. Dynamic ocean management: Defining and conceptualizing real-time management of the ocean. Mar. Policy 58, 42–50 (2015).

8.        Oestreich, W. K., Chapman, M. S. & Crowder, L. B. A comparative analysis of dynamic management in marine and terrestrial systems. Front. Ecol. Environ. 18, 496–504 (2020).

9.        Barlow, D. R., Klinck, H., Ponirakis, D., Garvey, C. & Torres, L. G. Temporal and spatial lags between wind, coastal upwelling, and blue whale occurrence. Sci. Rep. 11, (2021).