Content uploaded by Frederick H Sheldon
Author content
All content in this area was uploaded by Frederick H Sheldon on Dec 19, 2018
Content may be subject to copyright.
A preview of the PDF is not available
Ornithology of Sabah: History, Gazetteer, Annotated Checklist, and Bibliography
Abstract
Sabah is the Malaysian state representing former British North Borneo. It is well known for its rich forests, impressive mountains (including Mt. Kinabalu), and diverse avifauna. Ornithologists have explored and published on the birds of Sabah since the 1850s. This monograph reviews the work of these ornithologists by providing an historical overview of their explorations and studies; a gazetteer describing the sites at which they worked; an annotated checklist summarizing what they learned about birds, particularly in regard to distribution, habitat preferences, and breeding; and a bibliography of ornithological and related publications. The purpose of this report is to set the stage for future research on the birds of Sabah. This goal is accomplished in two ways. First, by reviewing what has been done, the history and gazetteer serve as a guide for expedition planners, wildlife managers, conservationists, and other scientists as they seek information on study sites. Certain parts of Sabah have been well studied (e.g., the East and West coasts), but other areas are poorly explored (e.g., the northcentral mountains and the south-central uplands). Second, by summarizing what is known about the birds of Sabah, the annotated checklist and bibliography provide a guide to research needs. Although a great deal is known about the distribution of birds in Sabah, remarkably little work has been done on bird ecology and evolution. Only a handful of modern, quantitative studies have been conducted on the population genetics, phylogeny, biogeography, migration, community ecology, or autecology of Sabah's birds. This dearth of ornithological information is a burden for conservationists, as they struggle to preserve the last bits of forest in Sabah, but it presents a great opportunity for bird researchers.
... On Borneo, Storm's Stork occurs widely in the island's lowlands but the majority of recent records are concentrated in Sabah (Malaysia), probably as a result of the easy access to sites such as the Kinabatangan floodplains (Sheldon et al. 2001, Mann 2008, Phillipps & Phillipps 2014. The species also occurs at several peat and freshwater swamp-forests in Indonesian Borneo (Silvius & Verheugt 1989, Danielsen et al. 1997, Cheyne et al. 2014). ...
Storm’s Stork (Ciconia stormi) is the rarest and most narrowly distributed stork species in Asia, confined to the forested lowlands of the Thai-Malay Peninsula, Borneo, Sumatra, and two satellite islands of Sumatra, North Pagai, and Siberut. Sumatra and Borneo are the core regions of its range, with extensive recent documentation highlighting their importance. BirdLife International (2021) estimates the current population at approximately 260−330 individuals, classifying the species as Endangered.
In Borneo, Storm’s Stork is widely distributed across the lowlands, with most recent records from Sabah, Malaysia, due to easier access to sites like the Kinabatangan floodplains. The stork also inhabits several peat and freshwater swamp forests in Indonesian Borneo, with recent sightings from Danau Sentarum in West Kalimantan, Sebangau, Tanjung Puting National Park, the Mawas conservation area in Central Kalimantan, and Gunung Lumut in East Kalimantan.
Danau Mesangat in East Kalimantan, covering over 180 km², is a significant site for Storm’s Stork. The area, characterized by waterlogged low-lying topography and a patchwork of forest remnants, scrub, and swamps, has been heavily logged and degraded for agriculture. Despite the primary research focus on threatened crocodile species, surveys have identified over 100 bird species in Mesangat. This article reports direct observations and camera trap documentation of Storm’s Stork in Mesangat during 2020 and the first half of 2021, discussing the significance of the wetlands for the species and suggesting key conservation actions.
... The species is largely confined to dipterocarp and freshwater swamp forests, as well as tall secondary and logged forests (Lekagul & Rou n d 19 91, We l l s 20 07, Tr e e s uc on & Limparungpatthanakij 2018) in the Peninsula's lowlands. On Borneo, it is known to extend into forest at submontane elevations to well over 750 m on Mount Kinabalu (Sheldon et al. 2001, Harris et al. 2012also A.J. Boyce. in litt. 2022), while in the Lesser Sundas, the species occurs to nearly 1,000 m on Flores (J. ...
In Malaysia, all 10 hornbill species have been recorded in Peninsular Malaysia whereas Sabah and Sarawak have eight species. Knowledge on Sabah’s hornbill distribution status, ecology, and biology remains limited. This rapid assessment covered the population status of the Helmeted Hornbills in the Imbak Canyon Conservation Area, its potential threats, and the level of hornbill awareness and local knowledge among the protected area’s rangers. The rapid assessment showed that Helmeted Hornbills can be regularly heard from the several trails surveys although direct sightings were rare. These hornbills were found to prefer the conservation area’s closed canopy forests that are dominated by large dipterocarp trees at Tampoi Research Station, Bukit Beruang, Big Belian Tree, Nepenthes Trail, Sungai Kangkawat Research Station, Mount Kuli Research Station, and Southern Rim. Culturally, the Murut indigenous peoples of this area believe that the Helmeted Hornbill is a sacred bird and hunting this species can lead to bad luck. The regular (call) encounters with the Helmeted Hornbills in the area surveyed is attributed to the protectionist beliefs of the Murut local communities. Hence, the Helmeted Hornbills in the conservation area are not hunted. However, Helmeted Hornbills are also threatened by habitat loss in its distribution range countries due to land use conversion for agriculture. For Sabah, further habitat loss or conversion need to be prevented and/or mitigated to ensure that Helmeted Hornbill populations continue to thrive in its remaining old-growth forests.
Recent studies on the cover, phytomass and turnover of bryophytes in tropical rain forests along altitudinal transects in the high mountains of Columbia, Peru, Borneo and Papua New Guinea reveal the existence of similar altitudinal zonations in different parts of the humid Tropics. As bryophytes are excellent climate indicators, are rather few in terms of species number, have very wide geographical ranges, and are a characteristic component of the structure of the tropical rain forest, they may be useful tools for the construction of a general scheme of the altitudinal zonation of tropical rain forests. Five altitudinal belts are distinguished: the lowland forest, the submontane forest, the upper lower montane forest, the montane forest, and the subalpine forest. A comparison is made with other zonation terminologies and the elevational variations of the belts, due to local climate, latitude, and other factors, are discussed briefly.
Studies at Pasoh Forest Reserve (Negri Sembilan, Peninsular Malaysia) established that flowers and fruits were consistently rare as food for birds. Food resources (flowers, fruits, arthropods) were less abundant in regenerating than in virgin forest; bird species richness and individual abundance were also lower in the regenerating forest understory. However, the two forests did not differ significantly in the relative importance of the various foraging guilds, suggesting that similar types of resources were present in similar proportions. None of the birds was a specialized consumer of understory flowers or fruit, whereas birds feeding mainly on foliage-dwelling arthropods were abundant and were represented by many species. This trophic organization is contrary to that reported for rain forests in other tropical regions but may simple reflect an allocation of harvestable productivity that is different rather than lower.-from Author
1. A vegetation map of Malesia, the festoon of islands from Sumatra to the Bismarcks, is presented on a single sheet, at scale 1:5 million and using Mercator's projection. 2. Ten different kinds of natural forest, either single Formations or groups of related Formations are depicted. The natural ranges of Pinus spp. are indicated. 3. Deforested areas and the natural swamp grasslands of the Fly and Sepik river valleys are left blank. Thus the extent of deforestation by Man is vividly depicted. 4. Legally constituted conservations areas of 200 km2 and larger are shown. 5. The sources for the map are described and inadequacies discussed. 6. It is concluded that a perfectly accurate map is impossible to make, partly because of inadequate information, partly because of the small scale, but that this flawed effort might stimulate investigations to fill in gaps.
Small litterfall (leaves, wood = or <2cm diameter, reproductive parts, and trash) was compared on sites of 1 ha in each of 4 types of lowland forest. The total small litterfall was (t ha-1yr-1 + or - 95% confidence limits): for alluvial forest (AF), 11.5 + or - 1.9; dipterocarp forest (DF), 8.8 + or - 0.6; heath forest (HF), 9.2 + or - 1.2; and limestone-site forest (LF), 12.0 + or - 1.3. A peak of litterfall occurred on all sites during a May-June period of high rainfall before the driest time of the year. The low Ca concentration in the DF litterfall and low nitrogen in the KF litterfall were notable. Total polyphenol concentrations in the leaf litterfall were ranked HF>LF>DF>AF. Virtually no wood >10cm diameter fell in the collection quadrats. Results for the >2cm-= or <10cm diameter wood were ranked AF>HF>LF>DF. Small-litter standing crop lowest value was 4.7 t ha-1 in the AF in March 1978, the highest 7.5 t ha-1 in the LF in February 1978. There were large differences between the 4 forests in the annual addition of elements in litterfall and in the element contents of small-litter standing crop and soils. There were several examples of high litterfall content/soil pool quotients. The highest quotient (2.8) was for Ca in the DF, an extreme example of rapid cycling of a deficient element in a rain forest.-from Authors
1. Relationships between lowland rain-forest structure, dynamics and site conditions were examined by the establishment of plots and profile diagrams within mature-phase forest at 13 sites and with permanent plots at three of these sites where trees have been enumerated over 20 years. 2. Most forest structural measures were intercorrelated but forest stature was uncorrelated with HCl-extractable soil nutrient concentrations and was apparently related to topography, soil depth and soil water. The relationship between diameter and height varied between forests and was correlated with HCl-extractable P and Mg for dry-land sites. 3. Tall forests generally had more slender canopy trees for a given height than short forests. Vertical stratification was associated with the presence and stand density of emergent trees, which comprise an architecturally and floristically distinct guild. Where emergents are scattered or absent, the main canopy and understorey were dense and no vertical stratification of crowns was perceptible; where emergents formed a continuous canopy, the main canopy was sparse, the understorey less dense and tree crowns were more separated into two horizontal strata. This type of forest appears to be confined to continuously moist soils on lower slopes, flat or undulating land. 4. Leaf size in the subcanopy was positively correlated with measured soil nutrients, and with the relationship between diameter and height. Among canopy trees it was not correlated with soil nutrients, nor with other aspects of forest structure. 5. Plots were originally established within mature phase, avoiding canopy gaps. Since establishment, gaps have accumulated within the permanent plots. Tree mortality was significantly clustered on clay soils, but not on sandy soils where windthrow was apparently less frequent and trees more often died standing. 6. Mean proportional diameter increments of large trees were not correlated with measured soil nutrients; but mean proportional diameter increments of recruits were correlated with measured soil nutrients. 7. A guild of tree species with the seed and/or seedling characteristics and the fast growth of pioneers, but with plagiotropic branching and crowns which reach the forest canopy or even emerge above it, was an important component of the building phase on soils with high mineral concentrations, but was sparsely distributed on other soils. 8. Standing volume and net volume increment were dominated by the mature phase; but soil nutrients probably influenced volume increment in the building phase. 9. Two inland forests, Bukit Mersing and Lambir, exhibited no net change in stand volume over 20 years. At Bako, a coastal forest on shallow freely draining soils, stand volume increased by 15.9% of the initial measurement. This increment occurred among trees of the largest diameter class while there was a net loss of volume in the small classes.
Extensive tracts of tropical rain forest were burnt in Borneo during the El Nino drought of 1983. Severe droughts have occurred previously but without causing such extensive fires. This extensive burning is a result of forests becoming more fire-prone after disturbance by logging. Rates of tree mortality after drought and fire ranged from 38 to 94 percent in logged forests and from 19 to 71 percent in unlogged forests, while for saplings rates of mortality exceeded 80 percent in both forest types. Secondary succession after logging was truncated by fire with the result that the post-fire condition of a forest logged six years before fire was similar to that of a forest logged two years before fire. The impact of fire differs from other natural or man-made gap forming processes in that most pre-existing seedlings and saplings are killed by fire. This left the regrowth in burnt forest depauperate in species diversity and in regeneration of upper-canopy species. Logged and burnt forests suffered severe canopy loss and the ground cover was dominated by grasses (e.g., Imperata cylindrica) or woody creepers (e.g., Eupatorium odoratum). In burnt primary forest, however, canopy loss was less severe and there was a low density of grasses. The prospects for recovery of forest structure appear to be good in burnt primary forest although species composition may be permanently altered. In forests logged prior to the fire, however, prospects for recovery of forest structure are not good, especially if further burning occurs. The recovery of these forests depends heavily on the ability of the secondary tree species to shade out the vigorous grasses, the continued presence of which may herald the conversion of the forest to unproductive grassland as has occurred widely elsewhere in the tropics after over-intensive shifting cultivation.
Herbivory (percentage area missing from large numbers of leaf samples of several species) in four plots was: 7.6% (330 m), 6.8% (610 m), 6.9% (790 m) and 7.1% (870 m). Annual mortality of trees (≥10 cm dbh) was 0.025 (280 m), 0.012 (330 m), 0.015 (410 m), 0.011 (480 m), 0.008 (540 m), 0.011 (610 m), 0.008 (700 m), 0.008 (790 m), 0.010 (790 m) and 0.016 (870 m). Median diameter increment of tree boles was (mm year-1): 1.56 (280 m), 1.13 (330 m), 1.31 (480 m), 0.93 (610 m), 0.91 (790 m) and 0.77 (870 m). Small litter fall was (t ha-1 year-1): 6.51 (280 m), 7.37 (330 m), 5.22 (480 m), 5.60 (610 m), 5.53 (790 m) and 4.80 (870 m). Chemical analyses of a small litter fall showed a number of important features, including an excess of Ca concentrations over Ma concentrations and relatively high concentrations of Ni. Three species rooted in ultrabasic soil have an ability to grow at low foliar nutrient concentrations. Shorea tenuiramulosa (Dipterocarpaceae) appeared to accumulate Ni and had up to 1.0 mg-1 Ni in its leaf dry matter. An unidentified Eugenia species (Myrtaceae) had 13.7 mg g-1 foliar Mn. -from Authors
During fieldwork in an area of wet-rice production in Selangor State, West Malaysia, calls and displays of the Cinnamon Bittern (Ixobrychus cinnamomeus) were observed. These are described and an attempt is made to relate them to the ecology of the species. Three call types were noted, Type I, which may serve to advertise the presence and availability of the male, was produced repeatedly by males perched in an exposed position. Type II was recorded in conjunction with the snap display during all intraspecific interactions and as an alarm. Type III, given by birds flighing to and from the feeding grounds, is regarded as a contact call. Two types of advertisement display are described. The first is given by males perched in an exposed position calling (Type II). This display involves exaggeration of the contrasting colors of the bill and coloration on the neck and head. It is likely that powder applied to the crown from the powder down tracts at the base of the neck also serves to enhance the bright coloration. The second involves 1-5 males circling while performing a slow wing-flapping display. Aggressive interactions were noted on the feeding grounds involving frequent chases and supplanting of birds arriving on site by birds already present. On each occasion the Snap Display and calling (Type II) was noted.
Eight species of Phylloscopus warblers, and the closely related goldcrest, Regulus regulus, breed along a restricted altitudinal gradient in Kashmir. None of the species breeds over the entire altitudinal range. Many altitudinal distributions are related to the presence of specific tree species, although most species will readily forage in all trees available on their territory. Arthropod composition was similar among the habitats. There are striking associations of morphology with ecology. Three orthogonal axes are used to summarize morphological variation among the species. The first principal component (PC1, essentially body size) is correlated with prey size. PC2, a shape measure relating beak size to tarsus length, is correlated with habitat occupied and mean elevation. PC3, a measure of beak width relative to its length, is correlated with feeding method. In total the morphological variation explains 95% of the measured ecological variation. There are obvious adaptive explanations for the PC1 and PC3 associations, but the adaptive significance of the variation in PC2 is unclear. However, the association of PC2 with elevation is also observed within the one species which is common and widely distributed along the altitudinal gradient. The breeding season distribution of species among habitats may be partly a result of inherent differences in foraging ability, but other factors, probably including current competition, must contribute to limiting distributions. -from Author