Talk2NatureBrown-throated parakeet (Eupsittula pertinax) / Parrot contact-call dataset review
Island parakeet contact calls vary across locations
Dataset publication year 2021 · Note reviewed 2026-10-25
The study analyzes one call type, loud contact calls, rather than the species’ full vocal repertoire. It reports population-level differences in call duration, frequency measures and spectrogram structure/variability across five island populations and mainland Venezuela. The authors explicitly say receiver experiments are needed to test whether these acoustic patterns correspond to receiver responses. This is a useful parrot-specific acoustic-variation study, not a call-meaning or translation result.
What we reviewed · Targeted publisher-indexed primary-article passages and the official Dryad v4 record/file manifest reviewed; no article or dataset file acquired.
Reviewed the publisher-indexed abstract, recording approach, Table 1 location statement, frequency/SPCC-PCO methods, duration results and selected discussion passages; this is targeted review, not a complete article audit. The official Dryad API v2 record identifies dataset 77790, version 4 / version ID 147704, dates, dataset-level CC0-1.0 and eight file records with declared SHA-256 digests. The direct publisher fetch returned HTTP 403. The official README API download returned HTTP 401 and its record-page download link returned HTTP 403. No article PDF, README, spreadsheet, audio, supplement or code was acquired; upstream hashes were not locally checked.
Not independently reviewed by a qualified parrot-bioacoustics or animal-communication specialist.
Study at a glance
- Study question
- How do Brown-throated Parakeet contact-call characteristics vary within individuals and among island and mainland locations?
- Vocal scope
- The analysis focuses on loud contact calls only. The paper says the species produces other call types, many context-dependent, which this study did not analyze.
- Sampling and methods
- The article reports 1,319 calls from 92 units for duration analysis; a separate 80-call clean-spectrogram subset supports frequency comparisons, and 150 calls from five locations support SPCC-PCO comparisons. Mainland recordings can represent groups rather than known single callers.
- Reported pattern
- Island calls were generally longer and lower on selected frequency measures, while SPCC-PCO analyses found geographic structure and variability differences. These are acoustic associations, not semantic labels.
- Sample-count reconciliation
- The duration-results prose reports 1,319 calls from 92 units, but the 10 Table 5 rows sum to 1,313 calls from 91 units. The article page does not explain the difference; reconcile against the README and files before replication.
- Meaning boundary
- The authors state that experimental receiver tests are required to determine whether call patterns correspond to response patterns. The study does not establish message meaning.
- Dataset provenance
- Dryad dataset v4 (record ID 77790, version ID 147704) is listed at 57,091,189 bytes with dataset-level CC0-1.0. The eight-file SHA-256 manifest is stored locally in the repository; no payload or code was acquired and the hashes remain unverified.
- Rights and reproduction
- The dataset record points to CC0-1.0, but the journal article’s specific reuse license was not verified. The archive lists no code file; a custom MATLAB routine is described in the article. Sample grouping and split suitability need data-level review.
- Atlas
- Five island study-area centers from Table 1 are rounded to 0.1 degrees; mainland coverage uses only a country-scale Venezuela reference. Individual recording sites in the Dryad workbook were not accessed or mapped.
Reading location: Official Dryad dataset page and v4 API file manifest; targeted publisher-indexed article passages and official Ornithology author instructions; GeoNames Venezuela country feature.
What this does — and doesn’t — establish
Only selected article sections and public metadata were reviewed. The duration-results prose reports 1,319 calls from 92 units, while the 10 rows in Table 5 sum to 1,313 calls from 91 units; the article page does not reconcile this six-call/one-unit difference. Verify against the README and data before replication. The Dryad manifest contains no executable code file although the paper describes a custom MATLAB routine; no separate code repository was audited. Caller/group/session linkage, label provenance, recording conditions, exact mainland site rows, site sensitivity, split structure and receiver-response labels remain unverified. The article-specific reuse license was not established; the dataset-level CC0 record does not license the paper or a derived model. No qualified independent specialist review has occurred.
What we can build on
The most defensible first parrot task suggested by this source is a descriptive, group-aware replication of acoustic variation—not a system that translates a parrot. Keep call structure, population/site association, independently observed behavior and receiver response as different evidence levels. Even a reproducible classifier of island or caller labels would not establish what a call means.
See the related tools and reading paths →
The next reading task
If the dataset is selected after qualified parrot-bioacoustics review, first obtain the README and tabular schema through an authorized accessible route; reconcile the 1,319-calls/92-units duration prose with the 1,313-calls/91-units summed from Table 5; inspect caller/group/session fields; review RecordingSites for spatial sensitivity; and locate/version any analysis code. Preserve the official v4 file manifest and compare local hashes only after an authorized acquisition decision. Then define a descriptive target, grouped/site-aware split and baseline. No files are admitted and no model has been run.
Go to the sources
- Brown-throated parakeet vocalization dataset, Dryad ↗
Primary Dryad record for Brown-throated Parakeet contact-call data. The current dataset-level record reports version 4, a CC0-1.0 license and eight files totaling 57,091,189 bytes; its v4 API manifest includes repository-declared SHA-256 values.
Official Dryad v4 dataset metadata and API version history ↗
- Dryad publication policy, updated April 8, 2026 ↗
Current Dryad policy says accepted datasets are published under a CC0 instrument and notes that Dryad does not verify scientific validity, veracity, accuracy or completeness.
- Eberhard et al. (2022), island and mainland Brown-throated Parakeet contact calls ↗
Primary study comparing acoustic character and variability in loud contact calls of Brown-throated Parakeets across five Caribbean islands and mainland Venezuela.
- Ornithology author instructions on licensing and article sharing ↗
Official American Ornithological Society journal guidance distinguishes standard and open-access article licensing and restricts mass redistribution of subscription-based article links/content.
Official Ornithology author instructions: licensing and sharing ↗
- GeoNames country reference point for Venezuela ↗
Official GeoNames country feature used only as a one-decimal country-scale atlas reference for the mainland study regions in Venezuela.
Original editorial note by Talk2Nature contributors. AI-assisted, source-grounded synthesis; not a peer review or an independent replication. Linked works retain their own rights.