Many dramatic range extensions in Holarctic birds in recent decades have been well documented (e.g. Fisher 1953, Bonham & Robertson 1975, Hengeveld & van den Bosch 1991, Wehtje 2003, Choi et al. 2011), benefiting from the region's rich historical data, the relatively high levels of bird status monitoring in many of its countries and an upsurge in interest reflecting changing climate (Rosenzweig et al. 2007, Møller et al. 2010). Across most of the tropics, however, bird recording levels are so low that whilst even quite major extensions of known range are commonplace, it is usually difficult to determine which are genuine expansions and which are simply the belated finding of species previously overlooked. Across the tropics, recent decades are seeing unprecedented human-caused direct changes to habitats, particularly in South-east Asia, simultaneously with global climate change (e.g. Sanderson et al. 2002, Sodhi et al. 2004, 2011, Harris et al. 2011, Wormworth & Sekercioglu 2011), so major range shifts of tropical birds can be expected to become more common. The effects of climate change on tropical forest-dependent species status seem likely to be profoundly negative (Sekercioglu et al. 2008, Harris et al. 2011), but there has been little speculation on what will happen to non-forest species, many of which, particularly in South-east Asia, have greatly increased habitat available to them.

The present note documents the ongoing dramatic northward colonisation of a non-forest tropical Southeast Asian passerine, the Yellow-vented Bulbul Pycnonotus goiavier, into one of the region's least well recorded countries. It exemplifies the difficulties of inferring causation in such changes: northward spread is consistent with climate change, but such major expansion of a non-forest species may simply result from habitat change. The Yellow-vented Bulbul occurs in mainland South-east Asia, the Greater and some Lesser Sundas and the Philippines, and has been intraduced to Sulawesi (Fishpool & Tobias 2005). Its distribution is mainly Sundaic. As of 1970, the northernmost records in the Mekong basin or Vietnam were from Komphong Thorn and Siem Reap provinces, Cambodia (Thomas & Poole 2003) at roughly 13°30′N; records came from a little further north in non-Mekong Thailand and even further so in the Philippine archipelago. The first record for Lao PDR was in May 1996 (Evans et al. 2000), and there were two further records in 1997 (Cunningham 1998, Evans 2001), the northernmost being 250 km north of Siem Reap town. These authors did not speculate whether the species was an overlooked, scarce, resident of long standing in Lao PDR, or a new arrival.

Figure 1.

Lao PDR, showing records of Yellow-vented Bulbul Pycnonotus goiavier, and locations mentioned in the text. Shaded land is at or above 200 m altitude. Records are numbered chronologically, oldest first.

METHODS

RECORDS

All Lao records traced are given in Table 1 and portrayed in Figure 1. All 1990s records were from the south, with the first record in the north not until 2004 (Ban Nongpen); this was a further northward extension of c. 220 km. All early records were in or within a few kilometers of the Mekong channel; in 2005 birds started to be found on the plains of major Mekong tributaries. All records come from within 1 km of either riverbanks or wetland complexes, and most are in areas with highly anthropogenic habitat, often towns. All records are from plains level, the highest in altitude therefore being furthest upstream along the Mekong, at Louangphabang (300 m). The first evidence of general occupation of an area was not until 2010 and involved, surprisingly, the northernmost two areas so far, Paklay and Louangphabang, where birds were recorded from, respectively, three and two adjacent towns/villages along the Mekong. In 2010 there were also multiple sightings from the country's southernmost large Mekong-side town, Pakxe, including the first proof of Lao breeding: a recently fledged juvenile. The 16 records (or clusters of records) all come from February—August, from formal bird survey (four) and, mostly, recreational bird-watching (12).

DISCUSSION

That Lao records come mainly from sub- and peri-urban orchards, scrub and wetlands is consistent with the species's association with non-forest habitats, including towns, elsewhere (Fishpool & Tobias 2005). The monthly distribution of records (14 in March—July with singles in February and August) could suggest the species is a wet-season breeding visitor to Lao PDR, but in the light of its evidently resident status in its main range (Fishpool & Tobias 2005), this seems unlikely. Pending more records for definitive assessment, this pattern is assumed to reflect the reduced observation during the late wet season (August—October) and early dry season (November—January).

That the apparent colonisation might simply reflect a chance pattern from overall low observation effort needs serious consideration, particularly because this bulbul is “of decidedly patchy distribution in Bangkok's suburbs” (Round 2008: 137). If this is so elsewhere in its northern range (it is not applicable to its core Sundaic range, either in Indonesia (J.B.C. Harris in litt. 2012) or Malaysia (JWD, unpubl. data)), it could readily be overlooked in any given region by merely superficial coverage. The pulses of records in 1996–1997 (three), 2004–2009 (five) and, particularly, 2010 (seven) relative to the lack of records in 1998–2003 probably reflects patterns of bird observation rather than of bulbul numbers: little survey or bird-watching was undertaken in the country in 2001–2003, and much of that in 1998–2000 was in the northern half of the country, where the species was not then known. And, in general, 1990s surveys focussed on forest habitats inimical to this species (and were undertaken by observers intent on seeking and documenting threatened and/or restricted-range species, rather than those with no global conservation needs); observation in this species's habitats was much higher from 2004 onwards.

Overall, it is unlikely that the species was present, but overlooked, in Lao PDR much before 1996. Don Pong, Keng Khan-Gneng, Ban Phonsaat, Nong Souy and the Paklay area have been too poorly surveyed to determine this bulbul's abundance there. At four sites with high relevant survey effort this bulbul was seen only once: Don Son lies adjacent to Don Khon, which was surveyed in every month in 1997 without finding the bulbul (Cunningham 1998); Ban Nongpen was surveyed during 2003–2005 at least once in each of the 12 calendar months (Duckworth & Evans 2007); Nong Thaleo lies on the lower Nam Ngum plain where many wetlands were visited several times each during 2009, with Nong Thaleo itself receiving six visits (Duckworth in press); and the Pakxan wetlands have been visited intermittently since 1994, with particularly intense bird-watching during 2000–2002 (C. Wood in litt. 2004), two pulses in 2005 (C. Wood and JWD), and three in 2010 (JWD et al., unpubl. data). A genuine scarcity, perhaps only intermittent presence at these sites, is apparent around time of sighting, consistent with natural rarity at the edge of the species's range or, equally, the early stages of colonisation. Supported by the 2008 record from Don Khon on a short visit a decade after year-round bird-watching failed to find the species, the 2010 observations in Pakxe almost prove recent colonisation: the town was well covered historically by Engelbach (1932), and was a site of much casual bird-watching from 1992 to 2003, including numerous observations by M.K. Poulsen (in litt. 2008) during 2000–2001 and, of particular relevance to this bulbul, several mornings by the author in August 2003 counting urban and suburban birds. Similarly, the author sought urban birds in Louangphabang, including the exact sites of 2010 observations, in January and April 2000 and in November 2004, without finding the species.

Table 1.

Lao¹ records of Yellow-vented Bulbul Pycnonotus goiavier. Records are listed in chronological order. Lao-language place name elements: Ban = village of; don = island; keng = rapids; nam = river; nong = pool; xe = river.

The historical collection record from Lao PDR is so patchy and, overall, limited (Thewlis et al. 1998) that most of the several dozen bird species first found in the country only in the 1990s–2000s, are likely simply to have been overlooked historically. That Yellow-vented Bulbul is a genuine recent colonist is, however, supported by patterns in Lao PDR's western neighbour, Thailand, which has a fuller documentation of bird distribution and status over the last century. Deignan (1963) recorded the species in Thailand no further north than the coastal provinces (up to 13°30′N, perhaps a little further north), and Lekagul & Round (1991) only to Bangkok (13°45′N), with a disjunct population in Bung Boraphet (15°42′N, 100°14′E). Now, it has been found in all recording regions of the country, “just spreading to NW, NE” (Robson 2008: 479). This colonisation northwards through Thailand has resulted in recent records from as far north as Chiang Rai, almost 20°N (P.D. Round in litt. 2010). Colonisation timing is thus consistent with northward spread through Lao PDR, as further indicated by observations at Mukdahan (on the Mekong; 16°33′N 104°44′E): in 2007 W. Sanguansombat found it commonly, where none had been seen in April 2000 (P.D. Round in litt. 2010). That in Bangkok's suburbs, “it is generally scarcer than formerly” (Round 2008) presumably reflects recent major decline in urban mature trees and wetlands there, rather than any regionally applicable trend.

Surprisingly, no expansion has been detected in Lao PDR's eastern neighbour, Vietnam. Here, Robson's (2008) compilation recorded it only from Cochinchina (the southernmost part of the country), where it was known historically (Delacour & Jabouille 1931). Whether this represents a genuine difference between Lao/Thailand and Vietnam, or simply the generally low levels of bird recording in Vietnam (Pilgrim et al. 2009), is unclear.

Directly south of Lao PDR, the species was already common in Cambodia in 1924, when it was found as far north as Komphong Thorn at 12°42'N (Delacour & Jabouille 1925). It seems unlikely that it has been expanding steadily northward since then: rather, a rapid northward movement seems to have started in the 1990s; had it been any earlier, establishment such as around Paklay in 2010 would surely have been found in the southern Lao Mekong in the 1990s, when there were, instead, just odd records. Supporting a long-term presence in southern Indochina, the populations along the Thai coast from Samut Songkran province on the peninsula to Trat province on the Cambodian border (and, by extension, into Cambodia and Vietnam) were distinguished by Deignan (1955, 1963) as P. g. jambu, whereas Sundaic Thailand is occupied by P. g. personatus (considered a synonym of wideranging Sundaic P. g. analis by Mees (1986) and subsequent authors). This recognisable morphological divergence suggests that expansion out of the Sundaic region is not recent. The Lao records are presumably of P. g. jambu, but no specimens exist to confirm this.

For a species presently tied to the plains and associated with damp areas (Figure 2), colonisation of a generally rugged region such as Lao PDR is, unsurprisingly, occurring along river valleys. Whatever the proximate trigger of this bulbul's sudden northward expansion that the Mekong happens to run roughly north—south doubtless aids it. The expansion's cause remains unclear. Two obvious candidates, for a nonforest bird from the south, are deforestation (in Lao PDR and neighbours) and climatic warming; but other factors cannot be excluded.

Over the last few decades, forest cover in Lao PDR has declined in total area, and become much more fragmented and degraded (Table 2). Whatever the precise definition of ‘forest’ used (and that of Department of Forestry (2005) does not discriminate forest as habitat for forest-dependent birds particularly well; Thewlis et al. 1998), these changes are conducive to expansion by non-forest birds. However, non-forest habitat already covered half the country by 1990, such a large area as to suggest simple forest loss is not the explanation. Moreover, large non-forest areas have existed for well over a century around the various Mekong-plain towns and large villages (Garnier 1885), even though their number and size has expanded visibly over the last two decades. Indeed, the area of what appears to be this bulbul's favoured habitat in Lao PDR, mature fruit orchards near wetlands, has probably decreased in the same period, with the recent rapid conversion of many such areas to housing, commerce and light industry (author's observations in most major Mekong plain towns). Thus, perhaps connectivity of degraded land (which plausibly aids Yellow-vented Bulbul dispersal) is more relevant than overall deforestation. The country's main road, route 13 (connecting most of the main Mekong-side towns), runs effectively north—south. In 1992, when the author first travelled it, long stretches still ran through forest, albeit heavily degraded. Now, no substantial lengths of road between at least Louangphabang and Pakxe are flanked both sides by native forest. The average of forest loss in the country masks distinct regional differences that at first sight argue against habitat-triggered colonisation from the south: in south Lao PDR, the access route for the bulbul, forest cover declined by only 1.8% between 1992 and 2002, compared with an 8.4% loss during that period in the north (Department of Forestry 2005); but, this is probably not relevant assuming that deforestation corridors are more important in aiding dispersal than are reductions in overall forest cover. In support of this, in interior Borneo, this bulbul seems less prone to appear in fresh islands of clearance (from mining or logging) amid forest than do two other species of such habitats, Yellow-bellied Prinia Prinia flaviventris and Eurasian Tree Sparrow Passer montanus (I.A. Woxvold in litt. 2012). A challenge to the degradation corridor suggestion is that route 13 south of Louangphabang runs for over 200 km through rugged highlands; perhaps here, the Mekong allowed dispersal from Paklay, itself colonised direct from Thailand a little to its south, and explaining the surprising lack of settlement in Vientiane by at least 2010. The Mekong from Louangphabang south to the Thai border had, at least by 2012, no forest tracts blocking dispersal by non-forest birds (author's boat-based survey).

Table 2.

Changes in forest cover in Lao PDR, 1992–2002¹.

With regard to climate change, the species is tied in Lao PDR to the lowlands, which have the warmest climate; but this is probably an effect of well-watered plains (the species's main biome) being largely lowland. Indeed, the average annual mean and low winter temperatures in the northern Lao areas occupied by this bulbul (Louangphabang and Paklay) are much colder than the southern Lao hills above occupied altitudes (Lefroy et al. 2010, notably their Figs 2.4, 2.6). Calculated temperature change rates for Lao PDR (Lefroy et al. 2010: 7–8), which reflect generally slower change in temperature with latitude in the tropics than at higher latitudes (IPCC 2007, Colwell et al. 2008), the northern occupied areas are today still considerably colder than were the southern areas 30 years ago, before they were colonised by Yellow-vented Bulbuls. However, without detailed knowledge of bulbul dispersal patterns and microgeographic variation in climate, the coarsely averaged pictures of climate change analyses may simply not show the temperature patterns relevant to this bird.

Notwithstanding the difficulties of distinguishing genuine expansions of range in Lao PDR from results of fuller survey, several other non-forest bird species have clearly colonised the country since 1992. Since each species's first record, most have shown, and continue to show, rapid range expansion: Peaceful Dove Geopelia striata, Pied Fantail Rhipidura javanica, House Sparrow Passer domesticus and White-browed Crake Porzana cinerea; additionally, Brown-throated Sunbird Anthreptes malacensis has shown massive population increase and range expansion in the same period (e.g. Duckworth & Evans 2007, Duckworth in press; records of these species will be detailed elsewhere). Including the bulbul, and recognising that the dove derives from non-native populations in Thailand (Round 2008), five of these six species otherwise occur to the south of Lao PDR, although the sparrow is from the west (e.g. Robson 2008). In their main range, cold seasons are far less marked than in northern, and even southern, Lao PDR. All these species have also colonised or greatly expanded their known range in the adjacent countries of Thailand, Cambodia and/or Vietnam (e.g. Duckworth & Evans 2007, Robson 2008, Mahood et al. 2011, F. Goes pers. comm. 2012; and comparisons of Round 2008 with Deignan 1963), consistent with attributing their changes in status in Lao PDR to genuine expansion.

Figure 2.

Nong Souy, Savannakhet province, Lao PDR, 3 March 2007, showing the bush-studded wetland typical of non-urban sites with Yellow-vented Bulbul records in Lao PDR. The foreground bush is the invasive American Mimosa pigra, which is also spreading rapidly in Lao (Photo D. Van Gansberghe).

It is more difficult to make a case that any forest bird species first recorded in Lao PDR since 1990 is a genuine colonist; this does not necessarily mean that no such colonisations are likely to have occurred, but the information to work with is even more fragmentary. The resident ornithologists of the first half of the twentieth century lived in towns, that is, non-forest habitat, so although much of their specific collecting time was spent in forest, their overall familiarity with non-forest birds' status was better. Of the many forest birds first found in Lao PDR after 1990, there are only four species lacking historical records from parts of adjacent countries that suggest strongly that they were overlooked in Lao PDR historically. Of these, only one, the montane Red-tailed Minla Minla ignotincta, is common and conspicuous. Even for this species, colonisation of Lao PDR is less likely than is expansion from areas of Lao PDR not historically covered (to be discussed elsewhere). If a recent colonist, it would have come from the north.

Thus, excepting possibly the minla, there does not seem to be any recent bird colonist of Lao PDR from the north. This is consistent with climatic change being the driver of these species' range extensions. The very little information on the extent to which recent climate change may be causing range shifts (geographic or altitudinal) in South-east Asian birds reflects in part the limitations of incomplete knowledge of past conditions (Clark 2007). However, altitudinal changes in bird distributions consistent with climate change have been found in South-east Asia (Peh 2007, Round & Gale 2008) and more may be in process, unremarked.

Because Yellow-vented Bulbul is a conspicuous and distinctive species, and occupies human-modified environments, even with the current low levels of bird recording in Lao PDR it should be possible to track the further expansion of range that is likely to occur. Indeed, at its current rate of spread it might well soon to turn up in well-watched Hong Kong, assuming that the lack of records in Vietnam north of Cochinchina simply reflects limited recording there. At the larger scale, the factors behind changing bird distributions in tropical areas will remain opaque as long as so little formal monitoring occurs. While expansions such as this bulbul are of no conservation concern for the species itself, they are doubtless paralleled by the less detectable range contractions, which, if extinctions are to be minimised, need to be detected and understood in time to mitigate them. Moreover, range expansions by common species may bring them into contact with species of direct conservation concern, which latter may consequentially decline (Harris et al. 2011). The need for understanding species range changes is arguably more acute in South-east Asia than in any other similarly sized area of the world, given the concentration of other threats driving declines of such high magnitude in so many habitats and species there (Sodhi et al. 2004).