Palmer Penguins

The Palmer penguins dataset by Allison Horst, Alison Hill, and Kristen Gorman was first made publicly available as an R package. The goal of the Palmer Penguins dataset is to replace the highly overused Iris dataset for data exploration & visualization. However, now you can use Palmer penguins on huggingface!

  License

Data are available by CC-0 license in accordance with the Palmer Station LTER Data Policy and the LTER Data Access Policy for Type I data.… See the full description on the dataset page: https://huggingface.co/datasets/SIH/palmer-penguins.

The version of the Palmer Penguins dataset used in the ggEDA manuscript.Also accessible via the palmerpenguins R package.If you use please cite:Horst AM, Hill AP, Gorman KB (2020). palmerpenguins: Palmer Archipelago (Antarctica) penguin data. R package version 0.1.0. https://allisonhorst.github.io/palmerpenguins/. doi:10.5281/zenodo.3960218.Gorman KB, Williams TD, Fraser WR (2014) Ecological Sexual Dimorphism and Environmental Variability within a Community of Antarctic Penguins (Genus Pygoscelis). PLoS ONE 9(3): e90081. doi:10.1371/journal.pone.0090081

The Antarctic Penguin Biogeography Project is an effort to collate all known information about the distribution and abundance of Antarctic penguins through time and to make such data available to the scientific and management community. The core data product involves a series of structured tables with information on known breeding sites and surveys conducted at those sites from the earliest days of Antarctic exploration through to the present. This database, which is continuously updated as new information becomes available, provides a unified and comprehensive repository of information on Antarctic penguin biogeography that contributes to a growing suite of applications of value to the Antarctic community. One such application is the Mapping Application for Antarctic Penguins and Projected Dynamics (MAPPPD; www.penguinmap.com) - a browser-based search and visualization tool designed primarily for policymakers and other non-specialists (Humphries et al., 2017), and ‘mapppdr’, an R package developed to assist the Antarctic science community. The Antarctic Penguin Biogeography Project has been funded by the National Aeronautics and Space Administration (NASA), the Pew Fellowship for Marine Conservation, and the Institute for Advanced Computational Sciences at Stony Brook University.

Antarctic Penguin Biogeography Project: Database of abundance and distribution for the Adélie, chinstrap, gentoo, emperor, macaroni, and king penguin south of 60 S is an occurrence and sampling event type dataset published by SCAR-AntBIOS.

This dataset contains records of Pygoscelis adeliae, Pygoscelis antarctica, Pygoscelis papua, Eudyptes chrysolophus, Aptenodytes patagonicus, and Aptenodytes forsteri annual nest, adult, and/or chick counts conducted during field expeditions or collected using remote sensing imagery, that were subsequently gathered by the Antarctic Penguin Biogeography Project from published and unpublished sources, at all known Antarctic penguin breeding colonies south of 60 S from 1892-11-01 to 2022-02-12.

The data is published as a standardized Darwin Core Archive and includes an event core and occurrence and eMoF extensions. This dataset is published by SCAR-AntOBIS under the license CC-BY 4.0. Please follow the guidelines from the SCAR Data Policy (SCAR, 2023) when using the data. If you have any questions regarding this dataset, please contact us via the contact information provided in the metadata or via data-biodiversity-aq@naturalsciences.be. Issues with dataset can be reported at https://github.com/biodiversity-aq/data-publication/

This dataset is part of the Antarctic Penguin Biogeography Project project funded by National Aeronautics and Space Administration (NASA), the Pew Fellowship for Marine Conservation, and the Institute for Advanced Computational Sciences at Stony Brook University.

The Palmer penguins dataset by Allison Horst, Alison Hill, and Kristen Gorman was first made publicly available as an R package. The goal of the Palmer Penguins dataset is to replace the highly overused Iris dataset for data exploration & visualization. Using this python package you can easily load the Palmer penguins into your python environment.

Population reference: CHI (Chiloé), PUP (Pupuya), ALG (Algarrobo), CAC (Cachagua), TIL (Tilgo), PAJ (Pajaros), CHO (Choros), CHA (Chañaral), GRA (Isla Grande), AZU (Pan de Azucar), and PSJ (Punta San Juan).

This dataset includes Adelie penguin colonies and coastline digitised from Eric J. Woehler, G.W. Johnstone and Harry R. Burton, 'ANARE Research Notes 71, The distribution and abundance of Adelie …Show full descriptionThis dataset includes Adelie penguin colonies and coastline digitised from Eric J. Woehler, G.W. Johnstone and Harry R. Burton, 'ANARE Research Notes 71, The distribution and abundance of Adelie penguins, Pygoscelis adeliae, in the Mawson area and at the Rookery Islands (Specially Protected Area 2), 1981 and 1988'.

Datasets and R script wrokflow for Penguins Exploit Tidal Currents for Efficient Navigation and Opportunistic Foraging.The two main dataframes from which all analysis was conducted are:Ocean_Currents.csv The matlab equivalent (mat_uv_peng_extarea1.mat), utilised in 'Workflow script.R' is also supplied. This is vertically-averaged ocean current U/V components (m/s) from a ROMS simulation of the San Matías Gulf, hourly from 22 Nov 2019 for 240 hours.and Penguins.csv. This is High-resolution movement and behavioral data from GPS- and sensor-equipped Magellanic penguins (Spheniscus magellanicus) foraging in the San Matías Gulf (Peninsula Valdés, Argentina), 22 Nov – 1 Dec 2019.Other data files correspond to data underlying various figures/analysis. More information can be found in 'Workflow script.R'.Ocean_Currents.csv --> This dataset contains simulated ocean current data for the San Lorenzo Gulf, derived from the Regional Ocean Modeling System (ROMS). The dataset provides vertically averaged current velocities in the U (East-West) and V (North-South) components over a discretized spatial grid. Time series data are included at hourly intervals, spanning November 22, 2019, across 240 time steps. The dataset is structured to support analyses of current dynamics in relation to movement strategies of tagged Magellanic penguins, foraging in this area during this time.Data Collection & Model Description:Model Used: Regional Ocean Modeling System (ROMS)Grid & Coordinates:Horizontally: Orthogonal curvilinear grid (Arakawa C-grid)Vertically: Terrain-following stretched coordinatesForcing & Boundary Conditions:Tidal amplitudes and phases interpolated from the TPXO6 global tidal modelEight principal tidal constituents (M₂, S₂, N₂, K₂, K₁, O₁, P₁, Q₁), plus additional long-term and higher harmonicsOpen boundaries: South, West, and NorthVertical Mixing Scheme: Mellor-Yamada parameterizationBathymetry: Derived from digitized nautical chartsHarmonic Analysis & Data Processing:Instantaneous (3D) current outputs underwent harmonic decomposition to extract dominant tidal constituents (M₂, S₂, N₂).Phase adjustments applied using the XTide model (https://flaterco.com).The U (zonal) and V (meridional) velocity components were vertically averaged, as no significant depth-dependent variation was found within the study area.Spatial & Temporal Coverage:Study Region: San Lorenzo Gulf, Patagonia, ArgentinaTime Span: November 22, 2019- December 02, 2019Grid Resolution: Matches ROMS model spatial resolutionDepth Consideration: Vertically averaged currentsThe model setup follows Tonini & Palma (2017) for grid, forcing, and validation details.[Tonini, M.H. and E.D. Palma, Tidal dynamics on the North Patagonian Argentinean Gulfs. Estuarine, Coastal and Shelf Science, 2017. 189: p. 115-130.].Penguins.csv --> This dataset contains high-resolution movement and behavioral data from Magellanic penguins (Spheniscus magellanicus) tracked in the San Lorenzo Gulf, Patagonia, Argentina, during 2019. The data were processed from bio-logging devices, which recorded GPS, acceleration, magnetometry, and depth information, enabling reconstruction of the penguins' kinematics, behavior, and estimated GPS-corrected dead-reckoned (DR) positions.Fieldwork was conducted between 22 November and 1 December 2019 at the San Lorenzo Magellanic penguin colony, Peninsula Valdés, Argentina (42.08° S, 63.86° W). We selected 27 adult Magellanic penguins (Spheniscus magellanicus) brooding small chicks for this study. Each penguin was equipped with a GPS logger (AxyTrek, Technosmart, Italy) and a Daily Diary (DD) logger. The GPS units recorded positions at 1 Hz, while the DD loggers recorded tri-axial acceleration at 40 Hz and tri-axial magnetometry at 13 Hz, and pressure (indicating depth) at 4 Hz. Both devices were housed in hydrodynamic casings designed to minimize drag. Devices were secured on the dorsal midline with Tesa® tape. Penguins completed a single foraging trip before recapture.Ethical approval for the research was granted by Swansea University's Ethics Committee (SU-Ethics-Student: 260919/1894) and the Animal Welfare and Ethical Review Body (AWERB approval: IP-1819-30). Fieldwork permits were authorized by the Conservation Agency of Chubut Province (Disp N° 047/19-SsCyAP; No. 060/19-DFyFS-MP). All procedures involving penguin handling were reviewed and approved by the Dirección de Fauna y Flora Silvestre and the Ministerio de Turismo y Áreas Protegidas de la Provincia de Chubut.This dataset is sub-sampled to 1 Hz per penguin. All smoothed variables ('_sm') had 2 s centre-aligned rolling means applied (circular mean for heading).The dataset includes timestamped movement variables, behavioral classifications, and GPS-corrected dead-reckoned locations for the tracked individuals. These data were used to analyze penguin foraging behavior and navigation strategies, in relation to ocean currents (see associated dataset "Ocean_Currents.csv"). Dataset Contents & Column Descriptions [Penguins.csv]:The dataset is stored as a CSV file (Penguins.csv), with the following columns:Column Name: DescriptionDateTime UTC timestamp of each record (YYYY-MM-DD HH:MM:SS)ID Unique penguin identifierVeDBA.smoothed Smoothed Vectorial Dynamic Body Acceleration (VeDBA)Mag.heading.smoothed Smoothed magnetic (compass) heading in degrees, corrected for sensor offsetssmth.depth Smoothed dive depth (m), adjusted for sensor driftBehaviour.occurrence Cumulated unique Occurence of each behaviour ("dive", "subsurface", "surface")Behaviour.Group 'Behaviour' and 'Behaviour.occurrence' values pasted togetherBehaviour.duration Duration (s) of the current behavioral boutBehaviour General behavior classification: "dive", "subsurface", "surface"corrected.pitch Smoothed body pitch angle (degrees), corrected for sensor biasDive.profile Sub-phase of dive (“Descent”, “Bottom phase”, “Ascent”), otherwise, 'Surface_Sub-surface'speed Estimated swim speed estimates (m/s)CS Ocean Current Speed (m/s) at the penguin’s location and timeCH Ocean Current Heading (direction of flow 'towards' in degrees)DR.longitude.corr GPS-corrected Dead-reckoned longitude (decimal degrees)DR.latitude.corr GPS-corrected Dead-reckoned latitude (decimal degrees)DR.straightline.distance.from.start.2D Planar distance (m) from the trip’s starting point, calculated in 2DRow.number Original row index in the processing pipelinePrey.persuit Occurence of prey-pursuit eventDR.longitude.corr.m GPS-corrected Dead-reckoned longitude converted to a planar metric projection (meters)DR.latitude.corr.m GPS-corrected Dead-reckoned latitude converted to a planar metric projection (meters)trip_status Trip phase indicator (e.g., “Outbound”, “Inbound”)turns Occurence of 'significant' turnsHour Hour of day (0–23)Time Time (YYYY-MM-DD HH) extracted from DateTimeU_comp Zonal (East-West) current velocity component (m/s)V_comp Meridional (North-South) current velocity component (m/s)index (same as 'Row.number')Distance 2D Distance moved (estimated from GPS-corrected dead-reckoned coords)Prop.distance.travelled Proportion of distance moved (scaled 0–1)Prop.time.travelled Proportion of time into round trip (scaled 0–1)Dataset Contents & Column Descriptions [Ocean_Currents.csv]:The dataset is stored as a CSV file (Ocean_Currents.csv), with the following columns:Column NameDescriptionLatLatitude (decimal degrees, WGS84)LonLongitude (decimal degrees, WGS84)HoursTime index in hours from model startU_compZonal (East-West) current velocity component (m/s)V_compMeridional (North-South) current velocity component (m/s)Time/TSTimestamp (YYYY-MM-DD HH:MM:SS, GMT)############################Files pertaining to bathymtery data:gebco_08_-67_-45_-62_-40_pertaining_to_Fig.S8a sgdf_bathy_pertaining_to_Fig.S8aWorkflow script.R --> R script showing step-by-step data processing, analysis, and figure generation. Note Fig 2 and S2, are constructed directly from Penguins.csv and Fig S1 from Ocean_Currents.csv.

This dataset contains information on the distribution of Penguins and their breeding colonies in the Australian Antarctic sector, as of 1983. It forms Australia's contribution to the International Survey of Antarctic Seabirds (ISAS). The results are listed in the documentation. These include counts of chicks, adults and nests, as well as colony distribution maps. The survey includes Emperor Penguins, Adelie Penguins, King Penguins, Gentoo Penguins, Macaroni Penguins, Rockhopper Penguins, Chinstrap Penguins and Royal Penguins.

Original data were taken from ANARE Research Notes 9.

Only data from the Australian Antarctic Territory is described in this metadata record.

Images of rough maps detailing the locations of each of the colonies are available for download from the url given below. Observation and count data have been incorporated into the Australian Antarctic Data Centre's Biodiversity Database.

The data are presented in the format of Croxall and Kirkwood (1979) as recommended by the Report of the Subcommittee on Bird Biology held in Pretoria. In the tables all counts are estimates of the number of breeding pairs except where otherwise indicated. The numerical estimates and counts are of three kinds, indicated by the coded N, C or A:

NESTS (N = count of NESTS or breeding/incubating pairs)
The most accurate count of breeding pairs is that derived from a count of nests. This is usually carried out during incubation, but may also be made while chicks are still in the nest, before creches are formed. Such counts are only underestimates of breeding pairs by the number of breeding failures sustained between egg laying and the date of the count.

CHICKS (C = count of CHICKS)
Late in the breeding season the only counts possible are those of chicks. In general most pygosceild penguins raise one chick per pair per season, so a count of chicks gives a reasonable approximation of the original number of breeding pairs. However, season to season variation in breeding success can often be considerable. For example Yeates (1968) reports breeding success in Adelie Penguins at Cape Royds of twenty-six per cent, forty-seven per cent and sixty-eight per cent ever three seasons. Also, Macaroni Penguins only raise approximately 0.5 chicks per pair per season, so that chick counts of this species may be a considerable underestimate of the true breeding population.

ADULTS (A = count of ADULTS)
Many colony counts and estimates were expressed as total number of birds or adults. These figures are difficult to interpret as they depend on the time during the breeding season at which they were made. For some days prior to and until laying is finished, both birds of a pair will be present at the nest site while during incubation it is more likely that only one bird will be present. A further problem with counts of 'birds' is that they may include individuals who are not breeding and this gives an overestimate of the true breeding population. The counts of 'birds' or 'adults' which appear unqualified in log books have been divided by two to give an estimate of the number of breeding pairs. It must be stressed therefore that these counts are the least accurate.

The degree of accuracy of these counts is inevitably highly variable and it is often difficult to ascertain on what basis a figure was arrived at. For the present survey counts have been allocated to one of five degrees of accuracy.

1. Pairs/nests essentially individually counted. The count is probably accurate to better than + 5 per cent.

2. Numbers of pairs in a known area counted individually and knowing the total area of the colony, the overall total calculated. This technique is useful for very large colonies.

3. Accurate estimates; + 10-15 per cent accuracy.

4. Rough estimate; accurate to 25-50 per cent.

5. Guesstimate; to nearest order of magnitude.

Many references are in the form ANARE (Johnstone) or simply ANARE. These refer to unpublished reports extracted from ANARE station biology logs. Those in the form Budd (1961) refer to published records and are listed in the references at the end of this publication.

The locations of some colonies are indicated on maps. Place names that (as of 1983) have not yet been approved are shown in the tables and on the maps in parentheses, for example: (ROCKERY ISLAND).

In bold, values that were significant for Fs (p < 0.02) and D (p < 0.05). Population reference: CHI (Chiloé), PUP (Pupuya), ALG (Algarrobo), CAC (Cachagua), TIL (Tilgo), PAJ (Pajaros), CHO (Choros), CHA (Chañaral), GRA (Isla Grande), AZU (Pan de Azucar), PSJ (Punta San Juan).

The upwelling hypothesis has been proposed to explain reduced or lack of population structure in seabird species specialized in food resources available at cold-water upwellings. However, population genetic structure may be challenging to detect in species with large population sizes, since variation in allele frequencies are more robust under genetic drift. High gene flow among populations, that can be constant or pulses of migration in a short period, may also decrease power of algorithms to detect genetic structure. Penguin species usually have large population sizes, high migratory ability but philopatric behavior, and recent investigations debate the existence of subtle population structure for some species not detected before. Previous study on Humboldt penguins found lack of population genetic structure for colonies of Punta San Juan and from South Chile. Here, we used mtDNA and nuclear markers (10 microsatellites and RAG1 intron) to evaluate population structure for 11 main breeding colonies of Humboldt penguins, covering the whole spatial distribution of this species. Although mtDNA failed to detect population structure, microsatellite loci and nuclear intron detected population structure along its latitudinal distribution. Microsatellite showed significant Rst values between most of pairwise locations (44 of 56 locations, Rst = 0.003 to 0.081) and 86% of individuals were assigned to their sampled colony, suggesting philopatry. STRUCTURE detected three main genetic clusters according to geographical locations: i) Peru; ii) North of Chile; and iii) Central-South of Chile. The Humboldt penguin shows signal population expansion after the Last Glacial Maximum (LGM), suggesting that the genetic structure of the species is a result of population dynamics and foraging colder water upwelling that favor gene flow and phylopatric rate. Our findings thus highlight that variable markers and wide sampling along the species distribution are crucial to better understand genetic population structure in animals with high dispersal ability.

Models presented are those determined to be most parsimonious, as well as all models receiving ΔAICc values ≤2. Percent of individuals from independent datasets correctly classified for sex based on probability estimates using supported models.Abbreviations: ΔAICc = Akaike's Information Criterion corrected for small sample size, w = Akaike weight, r2mf = pseudo r2 McFadden.

These represent the speed filtered GPS dataset and associated CRAWL models (R package crawl, v2.2.1) for 83 individual Chinstrap penguins instrumented between November 2018 and February 2019 at Deception Island at Bailey Head and Macaroni Point. Each individual penguin has a unique identifier (D_Rxx_Pxx) The data is tidy (one row = 1 observation = 1 location) and clean (no NAs; no duplicated lines, tracks cut off by deployment date + 24h and recovery date; points less than 2 min apart were removed for D_R4_P10). Each track is matched to the individual penguin metadata (ID, instrument type, deployment, recovery dates, original file, breeding status) The data is stored as a tibble in a R environment, contains 84,550 GPS locations and 24 columns. The GPS locations are available unprojected (lat lon " +proj=lonlat +ellps=WGS84 +datum=WGS84"; EPGS 4326) and projected in Polar stereographic (+proj=stere +lat_0=-90 +lat_ts=-71 +lon_0=0 +x_0=0 +y_0=0 +datum=WGS84 +units=m +no_defs; EPSG:3031). Also included are pdf images of individual tracks, and the shapefile (polygon) of Deception Island used as a mask.

Study system This study was conducted on Cape Crozier, Ross Island, Antarctica (77°27’15.00”S, 169°13’45.00”E) during the summers of 2012–13 and 2013–14 (hereafter 2012 and 2013, respectively). Cape Crozier is the largest Adélie Penguin colony in the southern Ross Sea and one of the largest for the species (Lynch and LaRue 2014). It is surrounded by hundreds of nesting South Polar Skuas (S. maccormicki), with most of the colony within skua foraging territories (Wilson et al. 2016). Our study included 43 chicks in 2012 and 69 in 2013 (112 total). Across both years, 84 chicks survived to the crèche stage and could be used to model crèching size and age and survival during that period. The mean crèching age was 21.3 days (SE = 0.46, range 15-26, n = 33) in 2012, and 18.9 days (SE = 0.41, range 10-25, n = 51) in 2013. Across the entire colony, not just study chicks, we observed that substantially more chicks died from apparent starvation in 2013 than in 2012. Although the average amount of...

Three skeletons collected from the late Oligocene Kokoamu Greensand of New Zealand are among the most complete Paleogene penguins known. These specimens, described here as Kairuku waitaki gen. et sp. nov. and Kairuku grebneffi sp. nov., reveal for the first time the unique proportions of a giant fossil penguin and the morphology of many key elements of the stem penguin skeleton associated with underwater flight, including the first reasonably complete sternum, one of only two complete forelimbs and the first described pygostyle. Relative proportions of the trunk, flippers and hindlimbs can now be determined from a single individual, offering insight into the body plan of stem penguins and improved constraints on size estimates for 'giant' taxa. Kairuku is characterized by an elongate, narrow sternum, a short and flared coracoid, an elongate narrow flipper and a robust hindlimb. The pygostyle of Kairuku lacks the derived triangular cross-section seen in extant Spheniscidae, suggesting the rectrices attached in a more typical avian pattern and the tail may have lacked the propping function utilized by living penguins. New materials described here, along with restudy of previously described specimens, resolves several long-standing phylogenetic, biogeographic and taxonomic issues stemming from the inadequate comparative material of several of the first-named fossil penguin species. An array of partial associated skeletons from the Eocene-Oligocene of New Zealand historically referred to Palaeeudyptes antarcticus or Palaeeudyptes sp. are recognized as at least five distinct species: Palaeeudyptes antarcticus, Palaeeudyptes marplesi, Kairuku waitaki, Kairuku grebneffi and an unnamed Burnside Formation species

Beast2, Figtree

Breeding distribution of the Adelie penguin, Pygoscelis adeliae, was surveyed with Landsat-7 Enhanced Thematic Mapper Plus (ETM+) data along the coastline of Antarctica, an area covering approximately 330° of longitude. An algorithm was designed to minimize the radiometric contribution from exogenous sources and to retrieve Adelie penguin colony location and spatial extent from the ETM+ data. In all, 9143 individual pixels were classified as belonging to an Adelie penguin colony class out of the entire dataset of 195 ETM+ scenes, where the dimension of each pixel is 30 m by 30 m, and each scene is approximately 180 km by 180 km. Pixel clustering identified a total of 187 individual Adelie penguin colonies, ranging in size from a single pixel (900 m**2) to a maximum of 875 pixels (0.788 km**2). Colony retrievals have a very low error of commission, on the order of 1 percent or less, and the error of omission was estimated to be 2.9 percent by population based on comparisons with direct observations from surveys across east Antarctica. Thus, the Landsat retrievals can successfully locate Adelie penguin colonies that account for ~97 percent of a regional population. Geographic coordinates and the spatial extent of each colony retrieved from the Landsat data are available publically. Regional analysis found several areas where the Landsat retrievals suggest populations that are significantly larger than published estimates. Six Adelie penguin colonies were found that are believed to be unreported in the literature.

This dataset includes:\r \r (i) a 2 metre resolution digital elevation model (DEM) of Shirley Island, Windmill Islands, Antarctica;\r \r (ii) reliability data for the DEM;\r \r (iii) contours interpolated from the DEM; and \r \r (iv) an orthophoto created using the DEM.\r \r The data are stored in the UTM zone 49 map projection. \r \r The horizontal datum is WGS84. \r \r The data were created by Robert Anders, Centre for Spatial Information Science, University of Tasmania, Australia to support the postgraduate research of Phillipa Bricher into the nesting sites of Adelie Penguins.\r \r See a related URL below for a map showing Shirley island.

This data set contains raw, annotated, and synthesized data used in the analysis by Hinke et al. (2021) titled "Serendipitous observations from animal-borne video loggers reveal synchronous diving and equivalent prey capture rates in chinstrap penguins" (DOI:10.1007/s00227-021-03937-5). The data derive from field work to monitor the diving and predation behaviors of two chinstrap penguins (Pygo...

1. Animal tracking data are indispensable for understanding the ecology, behaviour and physiology of mobile or cryptic species. Meaningful signals in these data can be obscured by noise due to imperfect measurement technologies, requiring rigorous quality control as part of any comprehensive analysis.
2. State-space models are powerful tools that separate signal from noise. These tools are ideal for quality control of error-prone location data and for inferring where animals are and what they are doing when they record or transmit other information. However, these statistical models can be challenging and time-consuming to fit to diverse animal tracking data sets.
3. The R package aniMotum eases the tasks of conducting quality control on and inference of changes in movement from animal tracking data. This is achieved via: 1) a simple but extensible workflow that accommodates both novice and experienced users; 2) automated processes that alleviate complexi...

Growing animals should allocate their limited resources in ways that maximize survival. Seabird chicks must balance the growth of features and fat reserves needed to survive on land with those needed to successfully fledge and survive at sea. We used a large, 34-year dataset to examine energy allocation in Magellanic penguin chicks. Based on the temporal trends in the selective pressures that chicks faced, we developed predictions relating to the timing of skeletal feature growth (Prediction 1), variation in skeletal feature size and shape (Prediction 2), and responses to periods of high energetic constraint (Prediction 3). We tested our predictions using descriptive statistics, generalized additive models, and principal component analysis.

Nearly all of our predictions were supported. Chicks grew their feet first, then their flippers. They continued to grow their bill after fledging (Prediction 1). Variance in feature size increased in young chicks but declined before fledging; this variance was largely driven by overall size rather than by shape (Prediction 2). Chicks that died grew slower and varied more in feature size than those that fledged (Prediction 2). Skeletal features grew rapidly prior to thermoregulation and feet and flippers were 90% grown prior to juvenile feather growth; both thermoregulation and feather growth are energetically expensive (Prediction 3). To avoid starvation, chicks prioritized storing mass during the first 10 days after hatching, then the body condition of chicks began to decline (Prediction 3). In contrast to our prediction of mass prioritization in young chicks, chicks that were relatively light for their age had high skeletal size to mass ratios. Chicks did not show evidence of reaching physiological growth limits (Prediction 3). By examining energy allocation patterns at fine temporal scales and in the context of detailed natural history data, we provide insight into the trade-offs faced by growing animals.

Methods We measured four skeletal features (foot length, flipper length, bill length, bill depth) and mass of 9,491 Magellanic penguin chicks from hatching until death or fledging. Skeletal feature units are cm and mass units are kg. Our study took place at Punta Tombo, Argentina from 1983 to 2017.

Statistics_ByFeatureAge.txt

We calculated descriptive statistics (average, standard deviation, coefficient of variation) for features and mass for each age of chick growth from ages 0 days to 90 days. We used the statistical program R to examine linear and non-linear allometric relationships between skeletal size and mass using linear, polynomial, and generalized additive models in the package mgcv (Version 1.8-31). We report parameters for these models at each age of chick growth for each skeletal feature and for overall body size (the first principal component of skeletal features).

PCAOutputs_ByAge.txt

We ran a principal component analysis on centered and scaled skeletal features using the prcomp function in R. We ran this analysis across all chicks and used it to calculate age-specific body condition by examining the residuals of the relationship between mass and overall body size (principal component analysis). We also ran a principal component analysis for each age of growth to examine how much of the variation in chick size at each age was due to overall size (first principal component) versus body shape (high order principal components).