The geographic distribution of a species is fundamental to biogeography and is essential for studying patterns of population change (Brown et al. Reference Brown, Stevens and Kaufman1996). Knowledge of a species’ geographic distribution is also essential for assessing its conservation status (Gaston Reference Gaston2003). Researchers have increasingly used long-term datasets of bird species’ occurrence to determine population changes or shifts in geographic distribution driven by climate and land-use changes (Gómez et al. Reference Gómez, Hobson, Bayly, Rosenberg, Morales-Rozo and Cardozo2021; Gula et al. Reference Gula, Green, Fritts, Dean and Sundar2022; Nussbaumer et al. Reference Nussbaumer, Nussbaumer, Guchu, Hatfield, M. Kanga and Kung’u2025; Gicquel et al. Reference Gicquel, Nilsson, Alonso, Low, Alonso and Boiko2026). In Africa, citizen science datasets have played a critical role in this research in several countries (e.g., South Africa and Kenya; Lee et al. Reference Lee, Brooks and Underhill2022; Nussbaumer et al. Reference Nussbaumer, Nussbaumer, Guchu, Hatfield, M. Kanga and Kung’u2025), and national bird atlases provide important baseline information on species’ distributions (e.g., Dowsett et al. Reference Dowsett, Aspinwall and Dowsett-Lemaire2008; Dowsett-Lemaire and Dowsett Reference Dowsett-Lemaire and Dowsett2014; Languy Reference Languy2020). However, because most national bird atlases and citizen science efforts primarily map species presence, details such as breeding occurrence often go unrecorded. This lack of information highlights an important data gap for Africa (Gula and Downs Reference Gula and Downs2026), especially given examples in which breeding records have helped document the loss of local breeding populations (Gula et al. Reference Gula, Green, Fritts, Dean and Sundar2022; Gula and Barlow Reference Gula and Barlow2023).

The Goliath Heron Ardea goliath occurs across sub-Saharan Africa, the Middle East, and South Asia (Kushlan and Hancock Reference Kushlan and Hancock2005). Despite its wide range, the population status of the world’s largest heron is poorly understood across most of its range. In particular, breeding has been poorly documented in most countries outside South Africa (Dean Reference Dean1988; Kushlan and Hancock Reference Kushlan and Hancock2005), probably because it often nests singly and can be easily overlooked. We therefore assessed the breeding distribution of this widespread heron to identify geographic information gaps and examine evidence of changes in its breeding range.

We surveyed the literature, citizen science databases, and other open-access data sources for records of Goliath Herons breeding throughout the species’ range up to 30 April 2026. We considered records of nest-building, active nests, recently fledged juveniles (easily identified by brownish wing coverts), or fledged juveniles being fed as evidence of breeding. We searched publications on the Web of Science® and Google Scholar®, as well as eBird records, using combinations of the common or scientific name with ‘nest’, ‘nesting’, ‘breeding’, ‘chicks’, ‘young’, and ‘fledgling’. We also searched publications available through the Biodiversity Heritage Library (https://www.biodiversitylibrary.org). Additionally, we searched country-specific journals (e.g., Babbler and Lanioturdus), published bird atlases, and other literature known to us that did not appear in the formal literature searches. We also followed references cited in the sources we located. We searched citizen science databases, including eBird (eBird 2026), iNaturalist® (all uploaded images), BirdLasser®, Observation.org, and the Namibian (Environmental Information Services 2026) and South African (University of Cape Town, unpub. data) nest record schemes, the latter of which ended in the late 1990s. In addition to searching images on iNaturalist®, we searched all images uploaded to the Cornell Lab of Ornithology’s Macaulay Library (http://www.macaulaylibrary.org) because eBird users may upload images that show evidence of nesting without adding breeding codes or notes that include our search terms in their checklists. We located egg records in international museum collections using the Global Biodiversity Information Facility (GBIF.org 2026), requested egg records directly from several major natural history museum collections in South Africa, and used records of leg-ringed nestlings from the SAFRING bird ringing scheme (unpub. data) as evidence of breeding. Finally, we solicited breeding observations from colleagues who work in the field and included our own field observations. If breeding was recorded multiple times at a site in the same season, we retained one record so that each entry represented a distinct breeding event. We retained records from the same site in different years.

We mapped Goliath Heron breeding records using approximate locations when place names were imprecise. These locations may therefore have errors of a few kilometres. If a place name was too general or a historical name could not be determined after an online search, we did not map the record. Where national bird atlases reported breeding only by half-degree square, we mapped the square rather than a more precise location. To assess changes in reporting over time, we summarised breeding records by decade, grouping records from before 1970 because some sources dated records only as ‘before 1970’. We assigned records to regions (Middle East, East Africa, Central/West Africa, and southern Africa) and summarised breeding months per region for records that included this information. Finally, we classified nests as solitary, in single-species colonies or in mixed-species colonies, and summarised the number of Goliath Heron nests in each colony record.

We found 216 records of distinct Goliath Heron breeding events between 1881 and 2026 at 150 sites across 22 countries, in addition to atlas squares with evidence of breeding (Figure 1). Just over half (50.5%) of the records were from South Africa. Most records came from citizen science databases (40.3%) and the literature (31.5%), while nest record schemes (15.7%), personal observations (7.4%), ringing data (3.2%), and museum collections (1.9%) contributed the remaining records. Over half (52.6%) of the records with year of observation were from 2000 to 2026 (Figure 2). The longest recorded period of activity at a nest was eight years in Kasanka National Park, Zambia (F. Willems pers. comm.).

Figure 1.

Recorded breeding occurrences of Goliath Herons across the global range (grey shading). Half-degree atlas squares were derived from four national atlas projects (Lewis and Pomeroy Reference Lewis and Pomeroy1989; Ash and Miskell Reference Ash and Miskell1998; Dowsett et al. Reference Dowsett, Aspinwall and Dowsett-Lemaire2008; Ash and Atkins Reference Ash and Atkins2009). Note: breeding months have been suggested for Upemba National Park, D.R. Congo, without any specific records (Mackworth-Praed and Grant Reference Mackworth-Praed and Grant1970). Goliath Heron illustration generated by ChatGPT 5.4 (1 April 2026, provided personal photograph and requested it in illustration form).

Figure 1. Description longue

Une carte géographique mondiale centrée sur l’Afrique et l’Asie du Sud-Ouest montre l’aire de répartition du Héron goliath en gris foncé. L’axe des ordonnées indique la latitude de 30 degrés Nord à 30 degrés Sud et l’axe des abscisses la longitude de 20 degrés Ouest à 100 degrés Est.

La zone grise couvre la majeure partie de l’Afrique subsaharienne, s’étendant du Sénégal à l’Ouest jusqu’à l’Éthiopie à l’Est, et descendant jusqu’en Afrique du Sud. Des poches isolées sont visibles le long des côtes de la péninsule Arabique et dans le Sud de l’Asie.

Les données de reproduction sont superposées à la cart:

- Des points orange (Breeding record) sont concentrés principalement dans l’Est de l’Afrique (du Soudan du Sud à l’Afrique du Sud) et de manière éparse en Afrique de l’Ouest.

- Des carrés blancs (Atlas breeding square) sont localisés en Éthiopie, au Kenya, au Zimbabwe et en Afrique du Sud.

- Un point d’interrogation orange est situé en R D C (République Démocratique du Congo) au niveau du parc national de l’Upemba.

Dans le coin inférieur droit, une illustration montre un Héron goliath adulte au plumage gris bleu et à la tête rousse, debout sur un nid de branches contenant trois œufs blancs.

Figure 2.

Frequency distribution of Goliath Heron breeding records per decade.

Breeding was recorded in every month across the species’ range. No clear regional pattern in breeding timing emerged, although there was a clear peak in September–October in South Africa (Table 1). For records with sufficient information (n = 155), most (78.1%) were of solitary nests, 11.6% were of single-species colonies, and 10.3% were of nests in mixed-species colonies. Five of the 16 records (31.3%) from mixed-species colonies contained only one Goliath Heron nest. For records that reported multiple nests in colonies (discrete and mixed), the mean ± SD number of nests was 6.4 ± 3.5 nests (range: 2–20, n = 23).

Table 1.

Numbers of Goliath Heron breeding records by month and country. Note: December and January have been suggested as breeding months for Upemba National Park, D.R. Congo, without any specific records (Mackworth-Praed and Grant Reference Mackworth-Praed and Grant1970)

Table 1. Description longue

Le tableau présente les données de reproduction mensuelles (de janvier à décembre) pour le Héron goliath.

* Moyen-Orient:

- Égypte: 1 en avril.

- Iran: 1 en mai, juin et juillet.

- Arabie Saoudite: 1 en mars, octobre et novembre.

- Yémen: 2 en mars.

* Afrique de l’Est:

- Érythrée: 1 en février.

- Éthiopie: 1 en février, mai et août.

- Kenya: Relevés répartis sur l’année avec des pics à 3 en mars et 4 en août.

- Somalie: 2 en septembre.

- Tanzanie: 1 en février, mars, avril, juillet, août et octobre.

- Ouganda: Présence notable avec 2 en février et 3 en juillet et août.

* Afrique australe:

- Angola: 1 en février et décembre.

- Botswana: 1 en mai, juin, septembre et décembre.

- Malawi: 1 en août, octobre et novembre.

- Mozambique: 1 en février.

- Namibie: 1 en mars, 2 en juillet, 3 en septembre et 1 en décembre.

- Afrique du Sud: Région la plus active avec des pics majeurs de 20 en septembre et 24 en octobre.

- Zambie: 1 relevé en janvier, mars, mai, juillet, août, septembre et novembre.

- Zimbabwe: Activité constante avec 3 relevés en février, août, septembre et octobre.

* Afrique centrale et occidentale:

- Côte d’Ivoire: 1 en février.

- R D Congo: Présence incertaine (point d’interrogation) en janvier et décembre, 1 relevé en juillet.

- Gabon: 1 en janvier et mars.

- Guinée-Bissau: 1 en octobre.

Our global assessment of Goliath Heron breeding distribution found relatively few breeding records for a species with such a large geographic distribution, with substantial gaps in records from Central to West Africa and on the Indian subcontinent. In addition to limited ornithological coverage in many parts of Africa (Gula et al. Reference Gula, Green, Fritts, Dean and Sundar2022), we may have found few records because Goliath Herons predominantly nest solitarily and their nests can be concealed in secluded areas, particularly along rivers or in dense vegetation on islands (Dean Reference Dean1988; Kushlan and Hancock Reference Kushlan and Hancock2005; pers. obs.). The minority of records from breeding colonies contrasts with a previous report of the majority of Goliath Herons nesting colonially in South Africa (Geldenhuys Reference Geldenhuys and Ledger1984). The variation in this breeding behaviour, as well as the range in colony size, may be unusual among herons worldwide.

A large proportion of Goliath Heron breeding records came from South Africa, where there is a large birding community that has increasingly adopted the use of citizen science platforms to report observations (Govender et al. Reference Govender, Maseko and Downs2026). There were few breeding records from Central/West Africa, including none from the Sahel region, where the species probably bred historically in the Inner Niger Delta, Mali (Zwarts et al. Reference Zwarts, Bijlsma, van der Kamp and Wymenga2009). A large population was found in the Sudd of South Sudan in 1979–1980 (Howell et al. Reference Howell, Lock and Cobb1988), but breeding there may have gone unrecorded because of limited survey coverage. Mackworth-Praed and Grant (Reference Mackworth-Praed and Grant1970) reported breeding months for Upemba National Park, Democratic Republic of Congo (D.R. Congo), without any specific records. Therefore, the eBird record we found on the Fimi River in western D.R. Congo appears to be the first site-specific breeding record from the country. Previously, no breeding records had been reported from Angola (Dean Reference Dean2000), but we found eBird records of a nest site on the Kunene River on the Angolan side of the border with Namibia.

Goliath Herons historically bred in the Mesopotamian Marshes of Iraq, where they were once common (Ticehurst et al. Reference Ticehurst, Buxton and Cheesman1922; Ticehurst et al. Reference Ticehurst, Cox and Cheesman1926; Allouse Reference Allouse1953). However, we could not locate any specific breeding records in the country, although it was reported about two decades ago to persist as a breeder on the Iraq side of the Hawizeh Marshes based on local conversations (Salim et al. Reference Salim, Porter and Rubec2009). A 2021 eBird record we found showed newly fledged juveniles on the Iran side of the Hawizeh Marshes, suggesting that breeding persists in at least this part of the marshes. This was also the first evidence we found of breeding in Iran in over 40 years, where nests were last reported from the Hara Mangrove Forest in Hormozgan Province in 1976 and 1977 (Scott Reference Scott2007, Reference Scott2008). The Goliath Heron’s breeding status in this province is currently unknown, but observations of single adults around Jask have been reported over the years (eBird 2026), including one that refers to the species as a common migrant to the area. The Iranian breeding population may be small and fragmented if breeding also persists in Hormozgan, and it faces substantial threats. The breeding population on the Red Sea islands of Eritrea, Saudi Arabia, and Yemen may also be small, with fewer than 50 breeding pairs estimated (PERSGA/GEF 2003; Semere et al. Reference Semere, Hagos, Seleba, Gebrezgabhier, Haile and Chiozzi2008).

Over a century ago, Blanford (Reference Blanford1898) described the occurrence of the Goliath Heron in South Asia as “somewhat mysterious.” Little has changed in over 120 years regarding its status, because we could locate no breeding records in the region. However, there have been occasional observations over many years in the Indian Sundarbans (eBird 2026), as well as several older observations in coastal Bangladesh (Thompson et al. Reference Thompson, Harvey, Johnson, Millin, Rashid and Scott1993), where it was reportedly common historically (Prater Reference Prater1926). Repeated observations over many years in India suggest that an undetected breeding population is possible, although the species appears to be uncommon there at present. Limited survey coverage or population decline, or both, could explain the rarity of recent observations.

The apparent decline of Goliath Herons in the Mesopotamian Marshes may reflect the effects of wetland degradation and possibly hunting (Salim et al. Reference Salim, Porter and Rubec2009). The remoteness of the Sundarbans in India and Bangladesh may have helped the population there persist, although increasing habitat threats here may affect Goliath Herons (Sumon et al. Reference Sumon, Kanok, Sadat, Mainuddin, Wahid and Karim2025). These populations are isolated from the African core and may be especially vulnerable. Their persistence also raises questions about the species’ historical distribution in Asia. This unusual biogeography warrants further research.

Although we found evidence of recent breeding in Iran, substantial data gaps in Central/West Africa remain to be addressed. We advocate that researchers and birdwatchers report observations in detail to help document breeding. No citizen science project currently focuses specifically on documenting bird breeding across Africa, but incidental reporting remains valuable, as demonstrated in this study and several recent studies (Muller et al. Reference Muller, Nwaogu, Helm, Tieleman and Amar2025; Gula and Downs Reference Gula and Downs2026). Detailed observations submitted to widely used databases such as eBird can support more robust assessments in the future.