Abstract
Background: The West Malagasy dry forests support numerous endemic species and experience heavy anthropogenic pressures, yet remain very poorly studied. Further research is required to understand species distributions and overall diversity in these threatened forest ecosystems.
Objectives: We aimed to provide a description of the avifaunal community of a particular dry forest, Mariarano forest, north-western Madagascar, as well as other habitats that are heavily integrated with these Forests. The study site possesses a highly endemic bird community and is under severe environmental pressure, but remains poorly explored.
Method: We compiled all records from a 9-year (2010–2018) bird survey data set (the most extensive compiled from a Madagascan dry forest to date), which yielded data from a combination of point count and mist-netting protocols. This was further supplemented by approximately 4384 h of opportunistic observation effort.
Results: In total, 95 species were detected, including 63 regional endemics (66.3% of all species), 2 local endemics and 7 Threatened or Near-Threatened species.
Conclusion: We highlight the forest mosaic habitats of Mariarano as a potential new Important Bird Area, given the regional importance of its endemic avifauna.
Keywords: birds; endemic; Madagascar; inventory; dry forest.
Introduction
The West Malagasy dry forests Endemic Bird Area (EBA 093) encompasses a large section of western Madagascar and supports seven locally endemic bird species, including Van Dam’s Vanga (Xenopirostris damii) and Coquerel’s Coua (Coua coquereli) (Stattersfield et al. 1998). They also represent the core range of several endangered mammal species, such as Coquerel’s sifaka (Propithecus coquereli) (Salmona et al. 2014). These habitats are also under heavy anthropogenic pressure; over 40% of Madagascar’s dry forests were lost in the second half of the 20th century (Harper et al. 2007), and deforestation rates continued at a rate of between 0.46% and 1.17% per annum between 2000 and 2010 (Grinand et al. 2013). Remaining stands of dry forest are also mostly highly fragmented (Safford & Hawkins 2013). However, despite their ecological value and the degree of conservation urgency they represent, these dry forests remain very poorly known ornithologically, with little research having been produced from them in recent years. This lack of detailed fieldwork data has led to some uncertainty regarding species distributions within this EBA generally, along with the overall richness and conservation value of many specific dry forest ecosystems found within its delimitations.
In this study, we summarise records from the monitoring programme run in Mariarano dry forest and its immediate vicinity by Operation Wallacea – a UK-based expedition company (www.opwall.com) – and its partners. While the biological value of Mariarano is expected to be high (Long et al. 2012), at present, it possesses no legal protection and represents one of the last few patches of unprotected deciduous dry forest larger than 8 km2 remaining in western Madagascar (Moat & Smith 2007; Nicoll 2003). Like most remaining stands of dry forest in western Madagascar, Mariarano is highly threatened by clearance related to expanding agriculture, charcoal production and timber extraction (Long et al. 2012). A previous study by Ackerman (2003) quantified annual deforestation rates of 3%, more than twice the higher estimates of regional average rates (Grinand et al. 2013). The Operation Wallacea programme here has supported one of the longest running ornithological surveys yet completed in western Madagascar. These have encompassed not only the closed-canopy forests of Mariarano, but also other important habitats that occur adjacent to, and are often entwined within, these forest ecosystems. As with many other West Malagasy dry forests (World Wildlife Fund 2019), Mariarano represents a heterogeneous mosaic landscape where stands of forest are integrated with patches of non-forest habitats (Evans et al. 2013). As forests may support different species to the non-forest habitats that occur in close proximity, survey effort was completed in all major habitats in the study area to determine a more comprehensive overview of diversity within the Mariarano landscape as a whole. These records represent the first description of Mariarano’s avifauna, apart from a previous short communication reporting the presence of Van Dam’s Vanga (Xenopirostris damii) in the study area (Rasamison et al. 2018). We report all species records made from these surveys, giving particular attention to the status of threatened species and local endemics. We then examine this information to assess the conservation value of the study area.
Methods
Study area
Mariarano classified forest (15°29′S, 46°41’E) and the immediately adjacent Matsedroy forest fragment (15°29′S, 46°38′E) (collectively referred to henceforth as Mariarano forest) are located in Mahajanga II district, Boeny region, north-western Madagascar, approximately 80 km northeast of Mahajanga city (Figure 1). The study site encompasses approximately 65 km2 and comprises a matrix of low-lying dry deciduous forest interspersed with areas of wetlands (small- to medium-sized lakes), lightly wooded grassland, scrub habitats and agricultural land, reaching a maximum elevation of 80 metres above sea level (Evans et al. 2013; Moat & Smith 2007). Approximately 38 km2 (58.5%) of this area consists of forest, with the remaining area being encompassed by non-forest habitats. The area experiences a tropical savannah climate classified as Am on the Köppen climate classification system (Peel, Finlayson & McMahon 2007) with fairly constant monthly temperatures (mean 27.3 oC). Monthly rainfall is highly variable (1 mm – 360 mm) with a distinct wet season peaking in intensity between December and February followed by a pronounced dry season running between July and September (unpublished Operation Wallacea data).
|
FIGURE 1: Maps showing (a) the location of the Mariarano study area within Madagascar as a whole and within the Mahajanga II district of north-western Madagascar and (b) satellite image of the Mariarano region, with the boundary of the study area demarked with a dashed red line. |
|
Data collection and analysis
Bird surveys in Mariarano forest were completed annually between 2010 and 2018, as a component of the long-term ecological monitoring programme run as a collaborative partnership between Operation Wallacea, the Malagasy NGO ‘Development and Biodiversity Conservation Action for Madagascar’ (DBCAM), the University of Antananarivo and local community forest management groups. All bird surveys were completed in the dry season over a 6–8-week period between June and August, and comprised two main methodologies – point counts and mist-netting. This survey work was conducted by Rachel Hannah Palfrey (RHP), Jack Baddams (JB), Bruno Andriandraotomalaza Raveloson (BAR), Solohery Rasamison (SR), Fionn Ó Marcaigh (FÓM) and Jamie Neaves (JN).
Unlimited-radius circular-plot points (Bibby et al. 2002) were completed along 11 randomly placed transects situated throughout the dry deciduous forest habitats of the study area. Grassland, scrub, agricultural land and wetlands were not covered by the point count surveys. Transects were between 2 km and 4 km long and had between 8 and 10 permanent point count stations located along their length, each separated by a distance of at least 200 m. In total, 92 point count stations were established in the study area. Attempts were made to survey these sites three times per season, although this was not possible in every year because of some sites not being established at the start of formal survey work in 2010. Point counts were completed between 06:00 and 09:30 each morning. Each count lasted 10 min, during which observers recorded all birds seen and heard, with the distance of each contact from the observer noted. Specific observers completing these point count surveys varied between years. Point counts were not completed in heavy rain or other adverse weather conditions, although as all survey work was conducted in the dry season, such conditions almost never occurred.
Mist-netting surveys were completed mostly on an informal basis in forest habitats throughout the study period. Mist-netting was conducted most mornings between 06:00 and 09:00. Four 25 m nets were used for these sessions, and these were checked at 30 min intervals after opening.
Records from point count and mist net surveys were further supplemented by the authors’ opportunistic records. These records included sightings made when travelling to and from sample points, as well as records made during casual exploration of Mariarano forest and the surrounding area. Grassland, scrub, agricultural land (chiefly rice paddies) and wetland habitats in Mariarano, which were not covered by our formal point count surveys, were a particular focus for opportunistic recording.
Once all records from surveys and opportunistic sightings were collated, we compiled a full inventory of all species detected in Mariarano forest following the taxonomy provided by Gill and Donsker (2018). Endemism and conservation status of each species was also noted. Two levels of endemism were recorded – regional endemics restricted to the Madagascan and the Indian Ocean Islands biodiversity hot spot as defined by Myers et al. (2000), and local endemics restricted to the West Malagasy dry forests EBA as defined by Stattersfield et al. (1998). Conservation status followed the International Union for the Conservation of Nature (IUCN 2018). We also identified non-native species using descriptions in Safford and Hawkins (2013), Sinclair and Langrand (2013) and the IUCN (2018). We recorded species as notable range extensions if they were not indicated as occurring in the study area by previous published work from the study area (Rasamison et al. 2018), maps provided in authoritative online resources (BirdLife International & NatureServe 2018; Del Hoyo et al. 2018; IUCN 2018), or in two of the region’s widely used ornithological references (Safford & Hawkins 2013; Sinclair & Langrand 2013). We also assigned categorical abundance estimates for each species in our inventory, based on frequency of observations. Relative abundance categories followed those used in previous ornithological papers by the authors (e.g. Martin et al. 2012, 2014, 2018). These were: abundant (typically recorded multiple times each day in suitable habitat), common (typically recorded at least once per day), fairly common (typically recorded about once per week), uncommon (recorded less than five or six times, on average, per field season) and rare (known from less than five observations within the study area). Species restricted to highly specific habitats only found in limited parts of the study area were also denoted with an additional ‘local’ categorisation (locally abundant, locally common, locally fairly common and local and uncommon). We also noted which species were observed by which authors, and which of the five main habitat types occurring in the heterogeneous matrix of Mariarano forest (agricultural land, forest, grasslands, scrub and wetlands) each species was observed in. Following recommendations from Lees et al. (2012), we also collated photographs taken from within the study area of as many species on our checklist as we possessed, and uploaded them to an online depository (Lynx Edicions 2018), noting the unique voucher numbers of each record to provide evidence of occurrence for as many species on our inventory as possible. Where possible, we also collated sound recordings for species we were unable to obtain a photograph for. These were uploaded to the same online depository. Finally, we tallied the number of species detected by point count surveys in the Mariarano’s forest habitats for each survey year, in order to examine changes in forest bird diversity over time. We restricted this analysis to forest habitats for two reasons. Firstly, because forests represent the most threatened habitat in the Mariarano study area and are the most likely to have been affected by land use pressures since surveys began. Secondly, because applying this analysis to other habitats that were not consistently surveyed by a systematic methodology would strongly influence the amount of opportunistic survey effort directed towards them, which was inconsistent between years.
Ethical considerations
This article followed all ethical standards required for conducting bird survey work in Madagascar, with mist-netting surveys being led by suitably qualified bird ringers.
Results
We reviewed a total of 40 144 individual birds detected in the course of our point count surveys. Opportunistic observations (mist-netting and author personal records generated outside of the point count surveys) equated to approximately 4384 person-hours (based on an average of 8 h per day for each author where they were in the field, but not conducting formal point count surveys). Mist-netting yielded 1192 individual captures (including 204 recaptures). In total, we recorded 95 species in Mariarano forest and its vicinity, including 63 regional endemics (66.3% of overall community richness) and two local endemics restricted to the West Malagasy dry forests EBA (2.1% of overall community richness). Recorded diversity of forest birds was lowest in 2010 (34 species detected); this is in likelihood an artefact of fewer point count stations being established in this year. Recorded diversity was then fairly consistent between 2011 and 2018, fluctuating between 39 and 45 species detected per year (Figure 2). The small variations in species detections per year may be a result of minor annual changes in the diversity of these forest ecosystems, but could also be a result of differential skill levels between observers, who were not always consistent between years. Four species in the study area are considered to be Near-Threatened, two species Endangered, and one species Critically Endangered (IUCN 2018). None of our records represents notable species range extensions following our criteria, although numerous species are only noted to occur in the region by some of our consulted sources, with others indicating them to be absent in this region of Madagascar. The most notable of these is Coquerel’s Coua (C. coquereli) (see details below). We were able to obtain photographic records for 74 species in our inventory (77.9%), and sound recordings for a further one species for which we could not obtain a photograph. These can be accessed here: https://www.hbw.com/ibc/u/36081. Table 1 summarises our findings from the main study area.
|
FIGURE 2: Number of species detected per year by formal point count surveys in the Mariarano study area’s forest habitats. |
|
TABLE 1: Checklist of bird species recorded in Mariarano Forest and its vicinity between 2010 and 2018. |
TABLE 1 (Continues...): Checklist of bird species recorded in Mariarano Forest and its vicinity between 2010 and 2018. |
Discussion
The results presented in this article represent one of the most complete inventories of a western Madagascan dry forest completed to date, casting new light on the biological value of this otherwise poorly studied area. Particularly notable records include:
Malagasy Pond Heron (Ardeola idae) – An Endangered species that is a locally fairly common breeding visitor to the lakes and other wetlands found in Mariarano forest, where it is most often observed singly (although sometimes in loose groups of up to three individuals) within shallow waterbodies with plentiful floating vegetation, which are in close proximity to the forest edge, in concordance with habitat descriptions in Safford and Hawkins (2013) and Del Hoyo et al. (2018).
Madagascan Fish Eagle (Haliaeetus vociferoides) – A Critically Endangered species which is known to breed in the adjacent Ankobohobo Wetlands. Single individuals have also been occasionally seen passing over the wetlands of the Mariarano forest, although no evidence has ever been provided for the species breeding here.
Coquerel’s Coua (C. coquereli) – A local endemic confined to the West Malagasy dry forests EBA. It is an uncommon resident in the Mariarano study area, being cryptic and rarely observed directly. Most detections have been from vocalisations, and it has been detected by camera traps several times (Operation Wallacea unpublished data). Prior to the records reported here, the species’ presence in this part of Madagascar appears to be ambiguous. It is not noted to occur anywhere in the vicinity of Mariarano forest by Del Hoyo et al. (2018), while Sinclair and Langrand (2013), BirdLife International and NatureServe (2018) and IUCN (2018) indicate its status here as ‘possibly extant’ or ‘unconfirmed’. The work of Safford and Hawkins (2013) is the only consulted source to indicate its occurrence within the region of Mariarano forest, albeit at lower densities than in certain other parts of its range. Formal point counts, camera traps and opportunistic observations detected this species 21 times in total, providing confirmation of the species’ occurrence within and the Mariarano study area.
Van Dam’s Vanga (Xenopirostris damii) – An Endangered local endemic confined to the West Malagasy dry forests EBA. It is an uncommon species in our study area, having been recorded once per season in 2011 and 2014, and twice per season in 2015 and 2017. Details of our observations of Van Dam’s Vanga in the Mariarano study area have previously been published elsewhere (Rasamison et al. 2018).
While this study provides important data on the avifauna of Mariarano forest, it should be acknowledged that our results are limited to a degree by the seasonal nature of our fieldwork. Given all our surveys were conducted between June and August, it is likely that certain wintering migrants may have been missed – for example, Sooty Falcon (Falco concolor), Ruddy Turnstone (Arenaria interpres) and Eurasian Curlew (Numenius arquata), all of which could conceivably occur in the study area according to distribution maps provided in other sources (BirdLife International & NatureServe 2018; Sinclair & Langrand 2013). As this period also does not correspond to the main breeding period for Malagasy birds (Safford & Hawkins 2013), our results may have been influenced by decreased detectability in certain species, given breeding species tend to be more localised and more vocal, and hence more detectable (Bibby et al. 2002). It is worth noting that seasonal detectability in dry forest species is not thought to be as pronounced as that found in the rainforest ecosystems of eastern Madagascar (Sinclair & Langrand 2013), although it is still possible that some species may be present in Mariarano in the wet season, but absent during our study period. The vast majority of our records (including all our formal point count surveys) were also made during daylight hours, with no specialist survey effort being dedicated towards the surveying of nocturnal birds. This could also have led to some other species being overlooked, such as White-browed Owl (Athene superciliaris), which we never observed, despite it being indicated to occur in the Mariarano region by multiple sources (BirdLife International & NatureServe 2018; Safford & Hawkins 2013; Sinclair & Langrand 2013). Most non-forest habitats in our main study area, as well as the extensive mangroves in the adjacent Ankobohobo Wetland, would also benefit from more intensive exploration. Further fieldwork effort in these habitats may yield new species for our site inventory, and also allow for more accurate assessments of the relative abundance values presented in Table 1.
In summary, while further work in the area is desirable, our results highlight Mariarano Forest as an area of high conservation importance, which supports at least 57% of the 168 non-vagrant species known to occur in western Madagascar (Safford & Hawkins 2013), including numerous threatened and endemic species. Other recent survey work demonstrates that this high conservation value is also reflected in other taxa, especially mammals and herpetofauna (Evans et al. 2013; Long et al. 2012; Mann et al. 2015). The consistency of richness in forest bird communities over time indicated by our point count surveys also suggests that population of most species probably remains stable at present, and still represents a relatively intact ecosystem. It is also important to note that the conservation value of Mariarano is heavily bolstered by the non-forest habitats that occur within the mosaic landscape of the study area. For example, of the 95 species detected by survey work, 46 (48.4%) were only detected in non-forest habitats, including 20 regional endemics (31.7% of all endemics encountered), and three out of the seven Threatened or Near-Threatened species recorded. We therefore recommend that Mariarano Forest, as well as the non-forest habitats that are heavily integrated within it, be considered for further conservation attention and legislative protection. The site supports sufficient endemic and threatened species to be considered an Important Bird Area (IBA) in its own right, located immediately adjacent to the existing Ankobohobo wetland IBA. The designation of a protected forest status for the Mariarano area would also be an important step in safeguarding the future of its high-value biodiversity.
Acknowledgements
This study was supported by and completed in conjunction with Operation Wallacea. The authors thank all Operation Wallacea staff who provided the logistical support which made this study possible, especially Rachel Daniels. Although all records presented herein were made by one or more of the authors, they thank the other ornithologists who have participated in the bird surveys in Mariarano forest, along with the numerous Operation Wallacea student volunteers whose support made this research possible. The authors also thank two anonymous reviewers for their helpful and constructive comments.
Competing interests
The authors have declared that no competing interests exist.
Authors’ contributions
R.H.P. and T.E.M. conceived the study, analysed the data and wrote the manuscript. P.L. designed the field methodologies and curated the data set. R.H.P., J.B., B.A.R., S.R., F.Ó.M. and J.N. collected the data. All authors read and approved the final manuscript.
Funding
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.
Data availability statement
The data analysed in this study belong to Peter R. Long. Reasonable requests for data sharing can be directed towards the corresponding author.
Disclaimer
The opinions and views expressed in this article are the authors’ own and not an official position of any institution or funder.
References
Ackermann, K., 2003, ‘The role of dry forests in Madagascar as a safety net in rural livelihood system’, The International conference on rural livelihoods, forests and biodiversity 19–23 May 2003, Bonn, Germany.
Bibby, C.J., Burgess, N.D., Hill, D.A. & Mustoe, S.H., 2002, Bird census techniques, 4th edn., Academic Press, London.
Birdlife International & NatureServe, 2018, Bird species distribution maps of the world, BirdLife International & NatureServe, Cambridge, Arlington, TX.
Del Hoyo, J., Elliott, A., Sargatal, J., Christie, D.A. & De Juana, E. (eds.), 2018, Handbook of the birds of the world alive, Lynx Edicions, Barcelona, viewed 07 July 2018, from www.hbw.com.
Evans, B., Rakotondraparany, F., Cole, L., Graham, S., Long, P. & Gandola, R., 2013, ‘The carnivores of Mariarano forest, Madagascar: First insights’, Small Carnivore Conservation 49, 15–19.
Gill, F. & Donsker, D., 2018, IOC world bird list (v 8.2), viewed 07 July 2018, from www.worldbirdnames.org.
Google Earth Pro, 2019, Google Earth Pro, viewed 01 June 2019, from https://www.google.co.uk/intl/en_uk/earth/.
Grinand, C., Rakotomalala, F., Gond, V., Vaudry, R., Bernoux, M. & Vieilledent, G., 2013, ‘Estimating deforestation in tropical humid and dry forests in madagascar from 2000 to 2010 using multi-date landsat satellite images and the random forests classifier’, Remote Sensing of the Environment 139(1), 68–80. https://doi.org/10.1016/j.rse.2013.07.008
Harper, G.J., Steininger, M.K., Tucker, C.J., Juhn, D. & Hawkins, F., 2007, ‘Fifty years of deforestation and forest fragmentation in Madagascar’, Environmental Conservation 34, 325–333. https://doi.org/10.1017/S0376892907004262
IUCN, 2018, The IUCN Red List of threatened species, Version 2018.1., viewed 07 July 2018, from www.iucnredlist.org.
Lees, A.C., Naka, L.N., Aleixo, A., Cohn-Haft, M., Piacentini, V.Q., Santos, M.P.D. et al., 2012, ‘Conducting rigorous avian inventories: Amazonian case studies and a roadmap for improvement’, Revista Brasileira de Ornitologia 22(2), 107–120.
Long, P., Rabenandrasana, M., Rabenoro, M., Darlington, A., McCann, R., Gandola, R. et al., 2012, ‘Biodiversity surveys of Mariarano and Matsedroy tropical dry forests and associated wetlands, Western Madagascar; 2010–2012, status report’, Unpublished report to the Ministry of Forests and the Environment, Antananarivo, Madagascar.
Lynx Edicions, 2018, The Internet bird collection, viewed 12 July 2018, from www.ibc.lynxeds.com.
Mann, G.K.H., Long, P., Rakotondraparany, F., The Seing, S. & Gandola, R., 2015, ‘First record of Fossa Cryptoprocta ferox in Mariarano forest, Madagascar’, Small Carnivore Conservation 52 & 53, 45–55.
Martin, T.E., Blackburn, G.A., Kelly, D.J., Heriyadi, D. & Singer, H., 2012, ‘The avifauna of the Lambusango forest reserve, Buton Island, south-east Sulawesi (with additional sightings from Southern Buton)’, Forktail 28, 107–112.
Martin, T.E., Nivet-Mazerolles, V., Landsmann, C., Guillemin, M., Dubos, J., Vallejo, F. et al., 2014, ‘Bird records from south-central Uzbekistan, 2010–2013’, Sandgrouse 36(1), 34–49.
Martin, T.E., Dubos, J., Fleureau, J., Loubon, M., Riviere, T., Nivet-Mazerolles, V. et al., 2018, ‘Bird records from the arid and semi-arid areas in southern Kazakhstan, 2009–2017’, Sandgrouse 40(1), 53–74.
Moat, J. & Smith, P., 2007, Atlas of the vegetation of Madagascar, Kew Publishing, Royal Botanic Gardens, Kew.
Myers, N., Mittermeier, R.A., Mittermeier, C.G., Da Fonseca, G.A.B. & Kent, J., 2000, ‘Biodiversity hotspots for conservation priorities’, Nature 403(6772), 853–858. https://doi.org/10.1038/35002501
Nicoll, M.E., 2003, ‘Forests outside protected areas’, in S.M. Goodman & J.P. Benstead (eds.), The natural history of Madagascar, pp. 1432–1437, The University of Chicago Press, Chicago and London.
Peel, M.C., Finlayson, B.L. & McMahon, T.A., 2007, ‘Updated world map of the Köppen -Geiger climate classification’, Hydrology and Earth System Sciences 11(5), 1633–1644. https://doi.org/10.5194/hess-11-1633-2007
Rasamison, S., Raveloson, B.A., Palfrey, R.H. & Martin, T.E., 2018, ‘Records of Van Dam’s Vanga Xenopirostris damii in Mariarano forest, north-west Madagascar’, Bulletin of the British Ornithologists’ Club 138(3), 275–277.
Safford, R. & Hawkins, F., 2013, The birds of Africa: Volume VIII: The Malagasy region: Madagascar, Seychelles, Comoros, Mascarenes, Christopher Helm, London.
Salmona, J., Jan, F., Rasolondraibe, E., Besolo, A., Ousseni, D.S., Beck, A. et al., 2014, ‘Extensive survey of the endangered Coquerel’s Sifaka Propithecus coquereli’, Endangered Species Research 25(2), 175–183. https://doi.org/10.3354/esr00622
Sinclair, I. & Langrand, O., 2013, Birds of the Indian Ocean Islands: Madagascar, Mauritius, Réunion, Rodrigues, Seychelles and the Comoros, 2nd edn., New Holland Publishers, London.
Stattersfield, A.J., Crosby, M.J., Long, A.J. & Wege, D.C., 1998, Endemic bird areas of the world: Priorities for biodiversity conservation, BirdLife International, Cambridge.
World Wildlife Fund, 2019, Southern Africa: Northwestern Madagascar, viewed 01 June 2019, from https://www.worldwildlife.org/ecoregions/at0202
|