NHPI develops in natural systems. Here, we tested for NHPI in juvenile common loons (Gavia immer) that foraged on lakes in the vicinity of their natal lake after fledging. Juveniles visited lakes similar in pH to their natal lakes, and this significant effect persisted after controlling for spatial autocorrelation. On the other hand, juveniles showed no preference for foraging lakes of similar size to their natal one. When lakes were assigned to discrete classes based on size, depth, visibility, and trophic complexity, both juveniles from large lakes and small lakes preferred to visit large, trophically diverse lakes, which contained abundant food. Our results contrast with earlier findings, which show strict preference for lakes similar in size to the natal lake among young adults seeking to settle on a breeding lake. We suggest that NHPI is relaxed for juveniles, presumably because they select lakes that optimize short-term survival and growth. By characterizing NHPI during a poorly-studied life-stage, this study illustrates that NHPI can take different forms at different life stages.

Master_NHPI_Loon_Juveniles.csv: Dataset represents the 85 instances (in rows) of juvenile loon visitations to other lakes recorded between 2012 and 2015 that were used in the study. Column variables consist of chick(s) IDs, natal lake territory, the destination lake territory that chicks were observed visiting, the observer, date and time of day for the observation.  For each lake type (Natal vs. Destination), we include column variables for: Latitude; Longitude; pH replicate measurements; secchi replicates; lake depth (m); lake size (ha); lake roundness, and the class of the lake taken from the Trophic Lakes classificaiton dataset available from: Rypel et al. (2019)

Rypel AL, Simonson TD, Oele DL, et al (2019) Flexible Classification of Wisconsin Lakes for Improved Fisheries Conservation and Management. Fisheries 44:225–238. https://doi.org/10.1002/fsh.10228

Lake_location_size_chemistry.csv: Dataset represents lake variables and associated information recorded for all lakes within the general study area.  Lake names and ID numbers are listed, followed by spatial coordinate data in a veriety of formats and physical lake variables measured in a number of replicates for depth, pH, size, roundness, and clarity (secchi). 

For analysis, replicates in the Master_NHPI_Juvnile_Loon dataset are averaged into a mean unit measurement, and basic statistics (mean and SE) for Destination lake variables are computed.  This dataset is then paired with the larger lake dataset containing physical variables for ALL lakes (i.e. Basic Features dataset) to construct the randomization analyses used in the study.

Replicate measurements for Secchi, pH, and roundness are also present in the Basic_Features dataset.  These values are averaged to provide a mean value for each lake when used in analysis. 

Prior to randomization analysis, we filter the lakes in the Basic Features dataset to remove lakes smaller than 10ha as such small lakes are typically unsuitable for loons. In randomizaiton analyses, lakes with missing values are ignored. For example, if a lake in the Basic Features dataset has pH values but not Secchi measurements, it would be used in the randomzation analysis for pH but not for Secchi.