analyses 1898-2014(written: 2015, 2017)


The past and present status of migrating Aquatic Warblers (Acrocephalus paludicola) in the Netherlands

Vincent van der Spek, Albert de Jong & Erik van Winden

Figure 1. Aquatic Warbler, first calendar-year, Kwade Hoek, Zuid-Holland, the Netherlands, 16 Augustus 2014

This is a rudimental web version published in 2018. The full paper (in Dutch, with English summary) was published in Limosa in 2020. The pdf was shared on ResearchGate and it can be viewed there.  

1. Introduction

Temminck (1815, 1820) was the first to mention the occurrence of Aquatic Warbler (Acrocephalus paludicola) in the Netherlands. In his work on the birds of Europe he qualifies the species as rare and accidental, without further details. Based on several clutches and two chicks present in Dutch museums, Aquatic Warbler must have been a rare or scarce breeder in the Netherlands until the late nineteen forties. Currently it’s a scarce passage migrant in varying numbers, almost exclusively in autumn (van den Berg & Bosman 2001; Boele & van Winden 2006). This article summarizes the past and present status of Aquatic Warbler as a migrant, based on museum specimens and ringed birds and field sightings. Though historical data is scarce, there's evidence that the migration pattern seems to have changed over time.


As the most endangered passerine in Europe, Aquatic Warbler is classified as Vulnerable by Birdlife International (2014). Its current breeding range is restricted to six breeding areas in Eastern-Europe, with main breeding sites in marshlands in Poland, Belarus and Ukraine. The population size is an estimated 10.200 – 14.200 singing males (Flade & Lachmann 2008). Only recently two wintering sites were discovered in West-Africa: in Djoujd, Senegal (Salewski et al 2009) and in the Inner Niger Delta (Poluda et al 2012). Autumn migration of Aquatic Warblers takes place along the Atlantic coast of the European mainland, where birds pass through Germany, The Netherlands, Belgium, France, Portugal and Spain. During migration, they prefer stopovers in coastal area’s with reed and sedge vegetation near open water and estuaries (De By 1990). First calendar year birds show a strong western orientation and cover most records in The Netherlands.

Figure 2. Aquatic Warbler, first calendar-year, Kwade Hoek, Zuid-Holland, the Netherlands, 16 Augustus 2014.

2. Methods

We analysed data from museum specimens in Naturalis Biodiversity Centre (the Dutch museum for natural history, Leiden). Apart from four insufficiently labelled specimens (no date or locality) in the Natuurmuseum Fryslân, most likely no other Aquatic Warbler items are present in other Dutch museums (pers comm J. J. F. Jansen). Since we focussed on migration we only used the historical lighthouse victims in the collection. These birds are all believed to be migrants (and not former breeding birds).  For information on former breeding records see Jansen & Roselaar (2017).


A spreadsheet with all ringing data from the Netherlands up to 2014 was received by the Dutch ringing centre (Vogeltrekstation, centrum voor vogeltrek –en demografie). All digitally available data is stored in the online database GRIEL (link). The calculations in this article (timing, age ratio) are based on the 1989-2014 period when for each individual the ringing number, date, place and age are known. Data prior to1989 is digitally incomplete, though we do know the total number of Aquatic Warblers ringed per year.

Aquatic Warbler was reviewed by the Dutch rarity committee CDNA between 1977 and 1992 (van den Berg & Bosman 2001; For this period, only accepted records were used for our analysis of the field sightings (link). Field sightings from 1992 up to and including 2013 were derived from the online database and Bijzondere Soorten Project, the scarce species project from Sovon (the Dutch centre for field ornithology). All these records were added to the ringing data for an overall analysis.

3. Results


3.1 Historical records: museum specimens

A lighthouse victim collected at IJmuiden, Noord-Holland on 17 September 1887, a first calendar-year female, is the first proof of an Aquatic Warbler in The Netherlands (Figure 3). The vast majority of the birds in the Naturalis collection are lighthouse (or lightship) victims (n=205), collected between 1887 and 1971, but mostly until 1958 (n=191). Not a single bird was collected during spring. According to Verheijen (1981) bad weather conditions determine the magnitude of death among migrating birds. Intense lights can be an attractive force for nocturnal migrants, especially during nights with a very low cloud ceiling, fog, drizzle or rain and a lack of moonlight. Mainly passerines suffer from man-made light sources such as TV-towers and lighthouses during autumn migration, when their navigation becomes confused by the light source and so they collide with buildings. Keepers of the Dutch lighthouse at Goeree, Zuid-Holland found out that the risk is highest during nights around dark moon and lowest in nights with a full moon (Brouwer 1929). One remarkable night deserves to be mentioned: on 13 August 1912 a staggering 29 Aquatic Warblers were collected at the lighthouse at Haamstede, Zeeland (Figure 4).

Figure 3. First proven record of Aquatic Warbler in The Netherlands, collected at the lighthouse of IJmuiden, Noord-Holland on 17 September 1887. Specimen currently stored in Naturalis Biodiversity Centre, Leiden, Zuid-Holland, the Netherlands (Albert de Jong)

Figure 4. A staggering 29 Aquatic Warblers collided with the lighthouse of Haamstede, Zeeland, The Netherlands during the night of 12 to 13 August 1912. Specimens currently stored in Naturalis Biodiversity Centre, Leiden, Zuid-Holland
(Albert de Jong)

3.1.1 Temporal distribution

Based on the lighthouse and lightship victims in the Naturalis collection, two migration waves are visible (Figure 5). There’s a sudden and steep peak in the first two decades of August, that quickly drops. There’s a second peak in the second decade of September, after which the numbers quickly drop again. The median date of autumn passage is 6 September, right before the second peak. Most adults passed by during the first peak. Date extremes for all birds are 27 July and 15 October. The bird from 27 July 1930, an adult collected at Schiermonnikoog, Friesland is the only one collected during this month. 14 specimens were collected in October. The bird picked up dead at the lighthouse at Texel, Noord-Holland on 15 October 1958 is the latest date for any Aquatic Warbler ever recorded in the Netherlands.


3.1.2 Age and sex ratio

Most museum specimens were aged (n=195). With 10 adults and 185 first calendar-year birds, the adult - immature ratio is 1:18.5 (or 5,1% and 94,8%). 162 birds are sexed. With 59,2% males are slightly more common (n=96).

Figure 5. Temporal distribution of Aquatic Warbler lighthouse victims 1887-1971 (n=205). Green is all birds, red is first calendar-year birds and blue is adults. Note the two migration waves: one in early August and one mid-September.

3.2 Modern records: ringed birds and field sightings

Between 1943 and 2014 641 Aquatic Warblers were trapped and ringed. With the exception of 1980 birds have been trapped every year since 1961 (Appendix 1). The number of birds trapped has, on average, increased. 1995 (n=26), 2002 (n=33), 2008 (n=25) and 2009 (n=25) were good years, while 2003 (n=56) was exceptional. Between 1997 and 2014 nearly half (n= 162; 46%) of all birds ringed (n= 354) were caught in Castricum, Noord-Holland. In both ringed birds and field sightings, spring records are very rare.  The date extremes in spring for ringed birds and field sightings combined (n=8) during 1977-2014 are 16 April and 5 May.


3.2.1 Temporal distribution

In the ringing and field sighting data, only one clear migration peak is visible (Figure 4). In autumn, the first birds are usually trapped at the end of July or the beginning of August. The first birds seem to arrive earlier than in the past, especially since the second part of the 1990s, though at the end of the period the first birds started to arrive slightly later again. Date extremes are 19 July and 10 October. After the second decade of September Aquatic Warblers become rare. The mean date for trapped birds is 10 August.

Figure 4. Temporal distribution of field sightings and ringed Aquatic Warblers in the Netherlands 1978-2014. Note that i) only one clear migration wave is visible, ii) July records are not unusual, while iii) October records are (nearly) absent.

Figure 6. Aquatic Warbler, adult, Meijendel, Zuid-Holland, the Netherlands, 7 August 2013. Only 4,9% of all trapped birds are adults.

3.2.2 Age ratio in autumn

448 out of 452 birds ringed in autumn in 1989–2014 were aged. 95,1% birds are identified as first calendar-year birds (n=426). Hence, adults (n=22) represent 4,9% of the aged birds.


3.2.3 Recoveries

Birds in the Netherlands don’t make long stop-overs at the sites where they are ringed: out of 613 birds caught between 1943 and 2013, only three individuals were re-trapped at the same site. Eight ringed birds have been reported abroad. Five out of eight foreign recoveries are of birds that were re-trapped during the same autumn they were ringed: two from Belgium and three from France. One bird was recovered on Lanzarote, Spain two years later (found as prey item in an Eleonora's Falcon nest!) and one was re-trapped the following year in Germany. Finally, one bird ringed in the Netherlands was photographed in Poland on the breeding grounds. Two foreign birds were recovered in the Netherlands: a bird from Belarus was re-trapped and a colour-ringed adult from Poland was seen in the field (Appendix 1). The Polish bird is mentioned in Boele & van Winden (2006), but the details were apparently not reported to either the Dutch or the Polish ringing schemes and the original observer has lost the bird’s life history.


3.2.4 Moult
A first calendar year bird trapped in Castricum on 20 August 2005 was re-trapped on 3 October, when it was actively moulting its primaries. On 10 October, when it was re-trapped once again, the moult was completed (Levering & Keijl, 2008). No other birds with wing moult have been reported. This bird represents the only October record amongst the ringed birds.

4. Conclusions and discussion

The datasets of museum specimens, ringed birds and the field sightings have very different backgrounds, so any comparison needs to be done with care. We however found some striking differences between ‘older’ (up till 1971) and ‘newer’ (1978-2014) data.


4.1 Temporal distribution

Autumn passage over the course of the 20th century seems to have changed. The difference in mean dates is for instance remarkably high: 6 September in museum specimens vs. 10 August in ringed birds. Three apparent changes influence the mean dates.


i) First of all, the data of the field sightings and ringed birds combined show a shift towards an earlier arrival of the first ten birds of the season. Out of a total of 205 birds there’s only one July record among the museum specimens, while since the second part of the 1990s, July records occur almost every year.


ii) The data of the museum specimens shows two clear migration peaks: one in early August and one mid-September. The September peak is (almost) invisible in the more recent data of the ringed birds and field sightings.


iii) The number of museum specimens collected in October (n= 14 on 205 birds) is remarkably high compared to the single ringing record and the complete lack of field sightings during this month.


One can assume that the length of the nights had an impact on the number of victims at lighthouses. Hence, during the longer September and October nights the number of victims is probably relatively higher than earlier in the season and this might effect the number of birds collected later in the season. Furthermore, the ringing activities in The Netherlands are not standardised and rather opportunistic: ringers usually focus on peak times of a species’ migration. Sound playback of Aquatic Warblers in the second part of September and in October is hardly used – if at all. However, the relatively few field sightings in September and the lack of any in October in a time when the number of active birdwatchers has grown, their knowledge of bird identification has increased, their equipment has improved and their sightings are shared on the internet, does strongly suggest that both the September peak and the seemingly regular passage in October truly are something of the past. Aquatic Warblers have two breeding periods and this clarifies a double peak in migration. Only part of the females breed in both periods, but females tend to be flexible and shift the timing depending on unfavourable conditions like water levels and change in vegetation (Vergeichik & Kozulin 2006). If this flexibility relates in any way to the shift in the migration period in the Netherlands is not investigated, but there could be a correlation.


4.2 Age

The age ratio of migrating Aquatic Warblers has not changed over time: the percentage of adult Aquatic Warblers amongst lighthouse victims between 1887 and 1971 is virtually identical to the ratio in ringed birds between 1989 and 2012. For ringed birds the low number of adults was already mentioned in other publications with smaller data sets (de By 1990; van Eerde 1998; Levering & Keijl 2008). De By (1990) stated that in western Europe, the number of adults increases towards the south. In France the percentage of trapped adults is higher, for instance 13,2% in 2009 and 17,8% in 2010 (Jiguet et al. 2011). In Portugal adults and first calendar-year birds are even equally common: 51 % of the birds in Neto et al (2010) were adults. Miguélez et al (2009) mention 17 adults out of a total of 28 trapped birds in Léon, northwest Spain in 2004-2006. It suggests that adults and first calender-year birds have a different migration strategy, most probably by taking another, more easterly route, or maybe by making less, shorter or different stop-overs


4.3 Increased number of ringed birds
That the number of ringed birds has on average increased over the years, cannot be explained by an increase of Aquatic Warblers: their numbers in the breeding areas have significantly dropped over the years (Flade & Lachmann 2008). Most likely the increase relates to an increased effort and improved trapping techniques. For instance, between 1960 and 1996 not a single Aquatic Warbler was trapped in Castricum, Noord-Holland. In 1997 playback of Aquatic Warbler was introduced (Levering & Keijl 2008). Ever since the species has been trapped at the site every year, and in larger numbers than anywhere else in the country. Influxes however do seem to occur. For instance in 1953, 20 birds were reported at De Beer, Zuid-Holland at 9 August and no less than 40 on 12 August (van den Berg & Bosman, 2001).Twelve were ringed at Makkumer Zuidwaard, Friesland on 12 August 1986 alone. In 2003 a total of 56 Aquatic Warblers were trapped, including 29 at Castricum.


4.4 Recoveries and stop-over time

The relatively few foreign ringing recoveries yield no surprises: they’re either from known breeding areas east of the Netherlands, or along known migration routes in the south. Of more interest is that there are hardly any re-traps on Dutch ringing sites of locally ringed birds. This suggests that the Netherlands is not of significant importance for stop-overs for this species. However, please note that the vast majority of the ringing sites in Holland is not in prime habitat for the species. Hardly any ringing activities take place on the few sites that look suitable and where birds are seen in the field (almost) every autumn. Further investigation on at least some of these sites is recommended. After all, Aquatic Warbler is a declining, globally threatened species that requires protection, also along its flyway.


4.5 Primary moult
That the bird caught in Castricum in October 2005 was actively moulting its primaries is remarkable, as Aquatic Warblers normally moult their wing feathers in the wintering areas in Africa (Kennerley & Pearsson 2010).



Henk van der Jeugd and Murad Maas (Vogeltrekstation, centrum voor vogeltrek – en demografie) provided the spreadsheet with all ringing data from the Netherlands. Hisko de Vries kindly provided all data of field sightings of the online database Kees (C J) Roselaar at NBC Naturalis helped us a great deal with the birds in the collection. Arjan Boele kindly provided information and commented the first draft of this article. Justin J F Jansen provided data on specimens from Natuurmuseum Fryslân. Adrzej Dyrcz was consulted several times. Michał Polakowski shared information about the Dutch ringed Aquatic Warbler photographed in Biebrza, Poland.




van den Berg, A.B. & C. W. Bosman (2001). Zeldzame vogels van Nederland. Avifauna van Nederland deel 1.  GMB Uitgeverij, Haarlem / Stichting Uitgeverij van de KNNV, Utrecht.

Boele A., & E. van Winden (2006). Waterrietzanger: de meest bedreigde BSP-soort. Sovon nieuws 19: 3-4.


de By, R. A. (1990) Migration of Aquatic Warbler in western Europe. Dutch Birding 12: 165-181.

Brouwer, G.A. (1929): Overzicht van de vogels aangevlogen tegen den vuurtoren 'Het Westhoofd’ op Goeree gedurende de jaren 1924 t/m 1928. Ardea 18:140-161.

Flade M. & Lachmann, L. (2008) International Species Action Plan for the Aquatic Warbler Acrocephalus paludicola. Brussels: European Commission.

Jansen, J.J.F.J. & C.J. Roselaar. 2017. Bijzondere legsels uit Nederland in collectie van Naturalis Biodiversity Center [Rare Clutches From The Netherlands In Collection Of Naturalis Biodiversity Center. Ditch Birding 39:2, 73-80

Jiguet, F. F. Chiron , O. Dehorter , H. Dugué , P. Provost , R. Musseau , G. Guyot , F. Latraube , P. Fontanilles , E. Séchet , J. Laignel , X. Gruwier & A. Le Nevé (2011). How Many Aquatic Warblers Acrocephalus paludicola Stop Over in France during the Autumn Migration? Acta Ornithologica, 46(2):135-142.


Levering, H. P. & Keijl, G. O. (2008) Vinkenbaan Castricum 1960-2006 - een halve eeuw vogels ringen. VRS Castricum, Castricum.


Miguélez, D., Zumalacárregui C., Fuertes B., Astiárraga H., González-Jáñez R., Roa S. & de la Calzada F. (2009) Habitat, phenology and biometrics of the Aquatic Warbler during autumn migration through a riverine wetland in Iberia. Ringing & Migration 24, 277-279.


Neto, J.M., Encarnaçao V. & Fearon, P. (2010) Distribution, phenology and condition of Aquatic Warblers Acrocephalus paludicola migrating through Portugal. Ardeola 57(1), 181-189


Pearson D. L. & Kennerley, P. R. (2010) Reed and Bush Warblers. Helm Identification Guides, an imprint of A & C Black, London.

Poluda, A., M. Flade, J. Foucher, K. Grzegorz, C. Tegetmeyer & V. Salewski (2012). First confirmed connectivity between breeding sites and wintering areas of the globally threatened Aquatic Warbler Acrocephalus paludicola. Ringing & Migration (2012) 27, 57–59

Salewski, V., Bargain, B., Diop, I. & Flade, M. (2009) Quest for a phantom – the search for the winter quarters of the Aquatic Warbler Acrocephalus paludicola. African Bird Club Bulletin. 16, 61–66


Vergeichik, L. & Kozulin, A. (2006) Changing nesting dates and nest placement as adaptations of Aquatic Warbler Acrocephalus paludicola to unstable nesting conditions on fen mires in Belarus. Vogelwelt 127:145-155.


Verheijen F.J. (1981) Bird kills at tall lighted structures in the USA in the period 1935-1973 and kills at a Dutch lighthouse in the period 1924-1928 show similar lunar periodicity. Ardea 69 (2) : 199 - 203.


Temminck, C.J. (1815) Manuel d’Ornithologie. ou Tableau systématique des oiseaux qui se trouvent en Europe : précédé d'une analyse du système général d'ornithologie, et suivi d'une table alphabétique des espèces


Little Bunting
The most interesting thing about this chapter is the sexing of birds. This could be done on the basis of lateral crown stripe: a new feature for me. Because of overlap between adult females and young males, birds should first be aged. The authors emphasize that this is their vision, and not necessarily the final answer to this matter. I found one in the field in October (Figure 4-5) of which I took reasonable pictures. The very rounded tail feathers look adult, as do the tertials and large coverts. The lateral crown stripe appears to be jet black and is separated from other feathers, such as the greyer back of the head. The orange-red crown stripe really stands out. Conclusion: according to this book this must be an adult male! The deep brown-red cheek - a variable feature according to this book - also fits that. So it works then? Well… how useful this really is, remains the question: based on this book I would label two "females" in Shirihai & Svensson (p546 both top right and bottom right) as males! The caption of one of these even mentions "safely identified as a female" because it has a relatively dull plumage for an adult bird. But is also has a jet black, strongly separated lateral crown stripe! The bird at the bottom right even has a pretty deeply coloured cheek, which I would be happy to call a male. Is this due to different insights, and if so, who should I believe? Or am I not interpreting this correctly? In any case, it is fascinating!

© 2018 by Vincent van der Spek