The Art of Conference Abstracts

I didn’t have a good starting point when I wrote my first conference abstract. I had read many publication abstracts but wasn’t sure how I could cram a 9,000 word research paper into just 250 words – it took me about 15 drafts to come up with something I didn’t hate. Through many years of trial-and-error, I’ve learned to summarize my research in a concise and interesting way. My resulting abstract “cookbook” has helped me, and I wanted to pass it along to others. Depending on the specific research and conference, your approach might vary a bit (e.g. if you are going to a technology-specific conference like Biologging, focus on the technical details; if you are going to a taxon-specific conference like Marine Mammals, focus on the biology background and applications; if you are going to a topic-specific conference like Animal Behavior, focus on the relevance) – but I think it’s a nice place to start.

Motivation: Why do we care? (1 broad sentence, 1 specific sentence)
Objective: Which question are you trying to answer? (1 sentence)
Approach: How did you go about answering your question? (2 sentences)
Results: What’s the answer? (3 sentences)
Conclusion: What are the implications of your answer? (1 specific sentence, 1 broad sentence)

Do include: Sample sizes, study species, study location, a cohesive story
Do not include: Citations, a million acronyms, jargon, grammar and spelling mistakes
 
Below are a handful of my proudest abstracts so far. Hope you find them useful! Next blog post will be about putting together an aesthetically pleasing conference poster. Stay tuned 🙂


Reproductive success and molt phenology affect colony attendance in Weddell seals
2018 POLAR conference, Switzerland
 
Mammalian species have been shown to shift life history event phenology in response to environmental change, but this response is highly variable across species. The replacement of hair (i.e. molt) is an intermediate life history event that may affect subsequent breeding attempts; however, the magnitude of its impact is poorly understood. Our aim was to understand the carry-over effects of molting and pupping phenology in Weddell seals. To do so, we conducted demographic surveys of 4,252 uniquely identified seals in Erebus Bay, Antarctica during the austral summers of 2013-2017. The start date of each animal’s molt was back-calculated based on hair loss progression, and pupping success and dates were obtained for the breeding season prior to and following the molt. The molt in Weddell seals lasted 29 ± 8 days. Within adult females, molt phenology was markedly different across reproductive and non-reproductive individuals, with non-reproductive seals beginning molt 16 days earlier than reproductive seals (mean start date Jan06 and Jan22, respectively, t-test p<0.0001). Animals that molted later were less likely to be seen in the breeding colonies during the following pupping period, suggesting they either did not pup that year or were deceased. By systematically manipulating estimated transition probabilities between pupping and molting categories using a simulation model, we demonstrate the importance of intermittent skipping of pupping to reset life history phenology.

Weddell seal dives suggest a vertical ecosystem shift concurrent with seasonal phytoplankton blooms
2017 Society for Marine Mammalogy conference, Canada
Warming temperatures and associated sea ice loss are predicted to impact the availability of prey resources for polar marine mammals. One way to predict how such changes might influence predator behavioral patterns is to study their responses to seasonal fluctuations in prey availability. While the Southern Ocean is one of the most seasonally productive marine ecosystems in the world, little is known about the distributions of its many prey species. Data from our deployment of time-depth recorders on 55 adult female Weddell seals over 4 years demonstrated a remarkably consistent seasonal shift in dive depths, in which dives gradually shallowed in January from >400meters to <150meters and returned to >400meters by mid-February. To investigate feeding activity during this period, we deployed jaw-accelerometers on 5 Weddell seals. Depth, bottom time, and number of dive-depth-wiggles were quantified for individuals that showed the deep-shallow-deep pattern (n=47 seals, 135,000 dives). For dives with jaw-accelerometer data, we detected jaw motion events using a surge acceleration amplitude threshold of 0.3g. The number of depth-wiggles was a strong predictor of jaw motion events in each dive (n=519 dives, R2=0.42) and therefore provides a proxy for feeding effort. Dive metrics from 10 sequential “shallow-period” days (mid-January) and 10 sequential “deep-period” days (rest of summer) were compared within individuals using a paired t-test. Depth-wiggles were more frequent during “shallow-period” than “deep-period” dives (mean±SD 3.0±0.6 and 2.3±0.5 wiggles min-bottom-time-1 respectively; p<0.05), suggesting higher foraging efficiency. We hypothesize that the mechanism driving the pattern in seal diving depth is a vertical migration of fishes, coinciding with the seasonal phytoplankton bloom in late December. Fish distribution changes of several hundred vertical meters could strongly impact short-term ecosystem dynamics. Placing predator feeding behavior in the broader context of ecosystem dynamics allows us to track a critical yet elusive component of global change.

Reproductive effort affects the annual dive behavior of adult, female Weddell seals
2017 SCAR Biology conference, Belgium
Given forecasted changes in sea-ice cover, understanding the baseline and seasonal activities of apex predators is a scientific priority. Our objective was to characterize the annual foraging behavior of adult, female Weddell seals (Leptonychotes weddellii) in the Ross Sea, Antarctica during 2013-2017. We hypothesized that following the lactation period, post-parturient seals would increase diving effort to regain body mass prior to their annual molt. To test this hypothesis, we deployed time-depth recorders (TDRs) on the flippers of 57 seals during the austral summer, including seals that had produced pups in a given year (post-parturient; n=34) and seals that had not (non-parturient; n=23). The tags were recovered between 39 and 436 days later and together comprise more than 135,000 dives from 5,642 seal days. During summer, seals that pupped were active divers (mean ± standard deviation 48±11dives day-1, with dives averaging 12±2min, 177±73m) whereas seals that had not pupped spent less time foraging (22±7dives day-1, 18±5min, 158±62m). Coincident to the higher diving activity, seals that pupped gained mass over summer (0.6±0.5 kg day-1) whereas seals that had not pupped lost mass (-1.0±0.4kg day-1). Despite marked inter-annual differences in sea ice extent (range 17 to 110km from McMurdo research station to ice edge), behavioral patterns and mass change dynamics were consistent across years, suggesting the relative resilience of this species to changes in sea ice extent. In all years, seals showed a clear pattern during late summer where dive depth gradually shallowed from >400m to <150m across two weeks and subsequently returned to >400m. Because the number of dive-depth-wiggles (a proxy for prey capture events) was significantly higher during this brief period, shallow diving could correspond with altered prey dynamics following the seasonal phytoplankton bloom. In the seven over-winter records, seals dove 3x more frequently and 50% deeper than summer, although dive durations were similar. Differences in dive metrics between post-parturient and non-parturient seals were not evident in the winter records. Our utilization of flipper-tag TDRs provides the first look at year-round Weddell seal behavior. These behavioral data are crucial for understanding predator responses to Antarctic change.

Using agent-based models to predict behavioral and physiological responses of top predators to environmental change: a case study with Weddell seals
2017 International Society for Ecological Modeling, South Korea
One of the crucial scientific challenges of this century is characterizing the vulnerability of ecosystems to global change. Bioenergetic models can be used to estimate total energy requirements by extrapolating simple physiological calculations to population-level metrics; however, these models often fail to link energy deficiencies with reproductive consequences or consider behavioral plasticity, and thus cannot be used to predict population-level consequences. An alternative approach is to use agent-based models which permit unique individuals to interact with variable extrinsic conditions (e.g. weather, prey), link energy deficiencies to reproductive and survival consequences, and allow individuals to adapt their behaviors. Here, we present an agent-based, ecophysiological model that simulates the energy balance of adult, female Weddell seals (Leptonychotes weddellii). The inputs include physiological parameters and population-wide ranges for the duration and phenology of life history events. Energy intake depends on foraging effort and stochastic prey availability at each timestep, whereas energy expenditure is calculated from time- and behavior-specific demands. The simulated animals select their activities (forage, nurse pup, molt, rest) based on body condition and life history constraints (i.e. dependent pup). Following model development and validation with empirical data, we ran simulations and compared the responses of individuals under baseline conditions to scenarios with reduced prey availability. A 10% reduction in prey availability resulted in seals foraging more and resting less (from 52.2%±6.2% resting to 40.3±8.4% resting). At the end of the year-long simulations, animals in the baseline simulation were in significantly better condition than animals with reduced prey availability (T-test, t28=5.6, p<.0001). Our model successfully explored decision-based energy allocation strategies that occur under energetic stressors and elucidated how extrinsic conditions may impact individual fitness. Predicting the behavioral and physiological responses of predators is valuable for the study of global change biology and can be used to inform management decisions in polar regions.

Oceanographic drivers of foraging effort and success in the northern elephant seal
2017 Biologging Conference, Germany
Recent initiatives have facilitated extensive animal tracking efforts around the globe, resulting in efforts to explain the distributions and behaviors of animals relative to oceanographic variables. State-space models (SSMs) have been developed to estimate foraging effort (by distinguishing foraging time from transiting time) and likewise, drift rate changes provide a proxy for foraging success. While these metrics are commonly used to study at-sea behavior, they are rarely compared to each other or to oceanographic variables. Our objectives were twofold: 1) To determine how drift rate changes and proportion of time spent foraging relate to body condition gains; and 2) To determine which oceanographic metrics are correlated with condition gains. We analyzed ARGOS-linked satellite tag data from 88 northern elephant seals Mirounga angustirostris across ten years. For daily latitude and longitude positions, we used a state-space model to estimate behavioral states, a custom code to measure seven-day sum of drift rate, and the R package Xtractomatic to obtain oceanographic metrics Chlorophyll a (Chl a; productivity), sea surface temperature (SST; water masses and fronts) and sea surface height (eddies). Of the 14,600 elephant seal positions, positive drift rate changes and foraging states occurred most often near the gyre-gyre boundary. Percent time foraging and drift rate changes were positively correlated to body composition changes but varied clearly by year. Mean Chl a showed the most pronounced differences between foraging and transiting states and was significantly higher in foraging areas than transit areas for all five years (paired t-test, p<0.05 for all). Mean SST was not a good metric for distinguishing foraging from transiting state, but it was the metric with the highest correlation to drift rate change (R2=0.1). Understanding the oceanographic drivers of animal foraging effort and success will help to highlight important spatial areas.

Using agent-based models to predict behavioral and physiological responses of top predators to environmental change: a case study with Weddell seals
2016 International Statistical Ecology Conference, USA
One of the crucial scientific challenges of this century is characterizing the vulnerability of ecosystems to global change. Bioenergetic models can be used to estimate total energy requirements by extrapolating simple physiological calculations to population-level metrics; however, these models often fail to link energy deficiencies with reproductive consequences or consider behavioral plasticity, and thus cannot be used to predict population-level consequences. An alternative approach is to use agent-based models which permit unique individuals to interact with variable extrinsic conditions (e.g. weather, prey), link energy deficiencies to reproductive and survival consequences, and allow individuals to adapt their behaviors. Here, we present an agent-based, ecophysiological model that simulates the energy balance of adult, female Weddell seals (Leptonychotes weddellii). The inputs include physiological parameters and population-wide ranges for the duration and phenology of life history events. Energy intake depends on foraging effort and stochastic prey availability at each timestep, whereas energy expenditure is calculated from time- and behavior-specific demands. The simulated animals select their activities (forage, nurse pup, molt, rest) based on body condition and life history constraints (i.e. dependent pup). Following model development and validation with empirical data, we ran simulations and compared the responses of individuals under baseline conditions to scenarios with reduced prey availability. A 10% reduction in prey availability resulted in seals foraging more and resting less (from 52.2%±6.2% resting to 40.3±8.4% resting). At the end of the year-long simulations, animals in the baseline simulation were in significantly better condition than animals with reduced prey availability (T-test, t28=5.6, p<.0001). Our model successfully explored decision-based energy allocation strategies that occur under energetic stressors and elucidated how extrinsic conditions may impact individual fitness. Predicting the behavioral and physiological responses of predators is valuable for the study of global change biology and can be used to inform management decisions in polar regions.

Whisker growth dynamics:  a validated approach for assigning timescales to stable isotope analyses
2013 Society for Marine Mammalogy conference, New Zealand
The extensive foraging migrations of many pinniped species discourage the use of traditional methodologies (e.g. scat analysis) for dietary reconstruction. Stable isotope analysis (SIA) of serially sub-sampled vibrissae (whiskers) is a common method to investigate pinniped foraging ecology; however, knowledge of tissue synthesis is required to assign accurate timelines to past foraging activity. In some species, whisker synthesis rates slow as the length asymptotes, so equally-sized subsamples for SIA represent differing time-scales.  Applying linear growth values to tissues exhibiting non-linear growth would lead to severe misinterpretations of temporal scales represented by serial isotope data. Photogrammetric analysis allows for non-invasive documentation of vibrissae growth and molting patterns in living animals. In this study, we used photogrammetric methods to obtain length measurements of 93 vibrissae over 18 months in a trained, captive northern elephant seal (Mirounga angustirostris). Vibrissae exhibited consistent asymptotic growth that was regulated by three von Bertalanffy growth function parameters: (1) initial time of growth, (2) asymptotic length and (3) a species-specific curvature constant. Unfortunately, photogrammetry does not account for the portion of vibrissae contained within the follicle. To correct for this photogrammetric underestimation, we constructed a linear correction model by correlating photogrammetric estimates to direct vibrissae measurements in three deceased northern elephant seals. Lastly, we quantified δ 13C and δ 15N ratios in archived blood, vibrissae, and prey samples from the captive seal. δ 13C and δ 15N ratios fluctuated along the length of each analyzed vibrissae, but exhibited similar values when matched to appropriate time scales based on length-specific curvature values. The vibrissae growth rates calculated from this captive seal are a key component in placing SIA data from vibrissae of wild pinnipeds within appropriate time frames. This study is the first to use vibrissae growth dynamics for appropriate interpretation of isotopic ratios in the northern elephant seal.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s