Journal of Tropical Forest Research 1 (2) : 1-20 (2017)

Original article

Comparison of Small Communities in Logged and Unlogged Areas in the Proposed Mahamyaing Wildlife Sanctuary,

Khine Thazin Wai Yongyut Trisurat* Ronglarp Sukmasuang

Department of Forest Biology, Kasetsart University, Bangkok, 10900 *Corresponding Author, E-mail: [email protected]

Received: Aug 1, 2017 Accepted: Dec 28, 2017

ABSTRACT The forest area in Myanmar has been gradually decreasing due to commercial logging and the expansion of agricultural area. The objective of this study was to investigate the response of small in logged and unlogged areas of the proposed Mahamyaing Wildlife Sanctuary, Myanmar. We used 180x180 m plot which consisted of 9 rows and 9 columns, 20-m apart, with a total of 81 traps. In addition, 3 replications representing habitat characteristics or disturbance levels in logged and unlogged areas were established and surveys were conducted during the rainy and cold season. Quantitative Sorensen Index and Shannon Weiner diversity Index (H’) were used to estimate species similarity and diversity, respectively, while the capture-mark-recapture methods were used to estimate the population size and density. The results indicated that 201 individuals were captured and 10 small mammal species were identified. The three most common species were Tupaia belangeri, rattus, and Callosciurus finlaysonii. Species richness was higher (H’ = 1.94) in the unlogged forests due to increased habitat heterogeneities as a result of small scale extraction. In contrast, the abundance of individual species was greater in the logged areas. The estimated population size of Rattus rattus and Tupaia belangeri in human-dominated landscape was 17.0 individuals.ha-1 and 23.4 individuals.ha-1 in the logged areas, respectively, while there were 9.6 individuals.ha-1 and 16.0 individuals.ha-1 in the unlogged areas. The models were not able to calculate the population size and density of the remaining eight species due to a limited sample size (n<20). It is concluded that the maintenance of mosaic and heterogeneous habitats is essential to enhance the small mammal diversity. คลังความรู้ดิจิทัล มหาวิทยาลัยเกษตรศาสตร์ Keywords: Habitat disturbance, logging concession, species diversity, species abundance. หมดอายุวันที่ 27-10-2564 INTRODUCTION diversity in terrestrial ecosystems and the Deforestation and habitat fragmentation important factors that lead to altering of are considered as the major threats to biological wildlife communities (Krishna et al., 2013; 2 Journal of Tropical Forest Research 1 (2) : 1-20 (2017)

Singer and Roger, 2014). These processes can urban green spaces, carried out in Ghana, also lead to a reduction in the available biomass led to a similar conclusion, that moderately and substantially modify the structure and disturbed habitats enhance species diversity composition of the original ecosystems, or and create an abundance of small mammals per se, create heterogeneous ecosystems and (Connell, 1978; Gbogbo et al., 2017). However, may result in an increase or decrease in species this is not the case for complete disturbance and diversity (Sodhi et al., 2004). Intermediate contrasting ecosystems. For example, oil palm disturbance hypothesis which is a prominent plantations are not suitable for or are of little theory regarding species diversity in ecosystems, importance to the non-volant small mammal observes that diversity is usually maximized fauna and they act as an effective barrier to the when moderate levels of disturbance (e.g., dispersal of these mammals (Bernard et al., frequency, spatial extent, intensity, interspecific 2009). In addition, Soontornpitakkool (1996) interactions) are practiced (Connell, 1978). This revealed that the population size and density theory has been tested and validated both in of rats and mice in the grasslands was less plant (Molino and Sabatier, 2001) and wildlife than that in the evergreen forest of Khao Yai communities (Conde and Rocha, 2006; Bendix National Park, Thailand, because a grassland et al., 2017; Gbogbo et al., 2017). contains a higher habitat homogeneity and Wildlife scientists usually have lower plant diversity. low interest in small mammals compared to Besides, the variations in micro-scales the larger ones such as elephants and tigers of a vegetation structure as a result of forest (Soontornpitakkool, 1996). However, small fragmentation also influences small mammal mammals play an important role in supporting distributions (Püttker et al., 2008). In addition, the food chain as food sources for small and native species are the most diverse and abundant medium carnivores, as they have short generation in relatively undisturbed habitats, but many periods and high turnover rate (Lekagul and can exist in an disturbed forest system and McNeely, 1988). Additionally, they can be used some can persist in second-growth. In contrast, as proxies for impact assessment as a result non-native species are predominantly found of disturbance in the habitat. For example, an in severely disturbed habitats (Eric et al., estimation of species richness, diversity, and 2007). They usually use remnant vegetation abundance of small mammals in three forest between fragmented landscapes as a mode of areas in Southern Brazil indicated that the two movement depending on seasonal availability most conservedคลังความรู้ดิจิทัล forests มหาวิทยาลัยเกษตรศาสตร์had a lower species of food and habitat preferences (Passamani richness and smaller mammal diversity at a and Ribeiro, 2009). Therefore, protection moderate disturbance, while in the anthropic of the remainingหมดอายุวันที่ forests in conjunction 27-10-2564 with area, wild species were absent (Conde and restoration of degraded habitats can be an Rocha, 2006). An investigation of small effective conservation strategy for native small mammals along disturbance gradients in an mammals. Subsequently, this will increase the Journal of Tropical Forest Research 1 (2) : 1-20 (2017) 3 overall health of the ecosystem as they would Mahamyaing forest reserve, located play an important link in the food chain (Lima in the western part of Myanmar, has been et al., 2002). proposed as a new wildlife sanctuary since According to FAO (2016), the forest 2002 (BNCA, 2010). In the last few decades, area in Myanmar declined from 392,180 km2 extraction of timber by logging companies, or 58% of the country’s land area in 1990 to from the Mahamyaing timber extraction site, 290,410 km2 or 43% in 2015, indicating a situated inside the proposed Mahamyaing loss of 4,070 km2 annually. This could be Wildlife Sanctuary (PMWS), was stopped largely attributed to commercial logging and two years ago due to expired concessions. the expansion of area under agricultural. In Although most large trees were removed, addition, wildlife poaching for bush meat can small− and medium−sized trees and shrubs, also lead to decline in numbers of rare species as well as small mammals can still be found. that inhabit both areas inside and outside the Furthermore, Buddhist monks living in the protected areas. The Ministry of Environmental area do not allow illegal cutting (Aung, 2007). Conservation and Forestry, Myanmar (2014) Hence, some biodiversity still exists, with is developing certification schemes to ensure the population of small mammals tending to that all forest products be derived under a increase after the stopping of logging activity sustainable forest management framework and (Adler and Levins, 1994). However, the buffer conform to a systematic legal step, starting zones of PMWS are heavily used by local from logging to consumption by the end users. people for their livelihoods. Thus, the semi- Therefore, some timber concessions are not evergreen forest is now characterized by a renewed after expiration and the government mosaic of different disturbance levels . puts extra efforts to reiterates the importance The objectives of the study were to of protected areas for biodiversity. The first estimate the species abundance, diversity, and wildlife sanctuary of Pidaung was established population of small mammals in the logged in 1918, and there are 43 gazetted and 43 and unlogged areas inside PMWS. The results proposed protected areas as of now, covering should be able to offer preliminary insights 7.3% of the total land area. The Ministry of into how different disturbance levels affect Forestry is aiming at increasing protected the small mammal communities and future areas to 10% of the total country area by recommendations for a sustainable management. 2030 (Istituto Oikos and Biodiversity and Nature Conservationคลังความรู้ดิจิทัล Association มหาวิทยาลัยเกษตรศาสตร์ or BNCA, MATERIALS AND METHODS 2011). However, only a limited number of Study area studies have been undertaken to analyze PMWSหมดอายุวันที่ is located in Kalay 27-10-2564 and Mawelik the biodiversity (e.g., wildlife and plants) to Townships of Sagaing Division, covering an support the establishment of protected areas area of 1,181 km2. It lies between latitudes in Myanmar. N22° 57ʹ and 23° 42ʹ and longitudes E94° 28ʹ 4 Journal of Tropical Forest Research 1 (2) : 1-20 (2017)

30″ and E94° 50ʹ 30″. The study area is a part Research Forest, covering an area of 183 km² of PMWS. It is located in the Mahamyaing (Figure 1).

คลังความรู้ดิจิทัล มหาวิทยาลัยเกษตรศาสตร์ หมดอายุวันที่ 27-10-2564

Figure 1 Map of Mahamyaing Reserve Area and the proposed Mahamyaing Wildlife Sanctuary, Saging Division, Myanmar. Figure 1 Map of Mahamyaing Reserve Area and the proposed Mahamyaing Wildlife Sanctuary, Saging Division, Myanmar. Journal of Tropical Forest Research 1 (2) : 1-20 (2017) 5

PMWS has been designated as an the altitude ranging between 200 to 500 m Important Bird Area (IBA) in 2010, due to the above sea level (Lynam, 2003). presence of Green Peafowl. In addition, the mixed deciduous forests in PMWS are a priority Small mammal sampling area for biodiversity conservation in central A preliminary survey was conducted Myanmar due to the occurrence of several during February to March in 2015 before the actual globally threatened species, such as Hoolock field work began (July 2016 to January 2018), Gibbon (Bunipithecus hoolock), Capped Leaf in order to ascertain the general characteristics Monkey (Trachypithecus pileatus), Asian of study area. In addition, interviews of the Elephant (Elephas maximus), and Banteng villagers were also conducted in 10 out of (Bos javanicus) (Tordoff et al., 2005; Lwin et the total 24 villages, who live both inside al., 2011). To date, more than 86 bird species and in the buffer zones. The questionnaire have also been reported. PMWS is dominated aimed at gathering information on the local by semi-evergreen forests, while oher forest livelihood and the current conditions of types include lower mixed deciduous in the the forest and wildlife in PMWS. Based valleys and Indaing forest (dry dipterocarp on a preliminary vegetation survey by Wai forest) on the ridges and flat plains. (2018), each treatment (logged and unlogged) There are 24 villages situated inside the consisted of 3 replications (sampling plots), sanctuary and in the buffer zones of PMWS. randomly situated, to reflect the gradient of Local people depend on the natural resources disturbance and habitat characteristics (Figure (edible plants and bush meat) available in the 2). The logged areas include agriculture wildlife sanctuary for their livelihoods and dominated areas (grass cover and herb cover), sustainance. The Minister of Forestry has succession stage (growing avegetation structure therefore employed forest rangers since 2002 or secondary forest, mainly consisting of to protect the remaining forests from further saplings), and disturbed forest, consisting of human disturbance and to ensure the long- only small standing trees (e.g. <10 cm dbh), term survival and well-being of the remaining tree stumps, and woody debris. The unlogged biodiversity. areas, dominated by bigger standing tress (tree The climate of the study area is height >10 m), with a higher density and dense characterized by tropical monsoons. Rainy crown cover, consisted of one open forest area season starts in June and ends in August (dwarf forest) and two closed forest areas. with a meanคลังความรู้ดิจิทัล annual rainfall มหาวิทยาลัยเกษตรศาสตร์ of around 1,473 Open forest includes trees and shrubs that do mm. Soil is very sandy and its water holding not have any visual interlocking and from a capacity is low, and as such, fresh water distance, they หมดอายุวันที่appear to grow 27-10-2564continuously, is nearly absent during the dry season. The while the closed forest areas include trees or southern side has a lower elevation and the shrubs that have a crown interlocking, or are north is hilly with numerous steep ridges with touching, or are very slightly separated. 6 Journal of Tropical Forest Research 1 (2) : 1-20 (2017)

Figure 2 Small mammal sampling design, Mahamyaing Reserve Area, PMWS. Figure 2Small mammal sampling design, Mahamyaing Reserve Area, PMWS. We used the quadrat design to survey was set per plot of each habitat type (logged small mammals in the logged and unlogged and unlogged forest areas). Thus, altogether forest areas situated in semi-evergreen forest. there were 4,374 trap nights for both the logged In each plot, 9 rows and 9 columns, at 20-m and unlogged forest areas during the 6 month intervals, were placed which resulted in 81 study period. quadrats per plot, covering 3.24 ha. At the Peanuts were used as bait in all the traps lower left corner of each quadrat, a single door to attract the small mammals (Gunther et al., live-trap (15 × 15 × 30 cm) was placed either 1983). The traps were inspected every morning on the ground, or on fallen logs, or on low (06.00-08.00 am.) and species identification branches, whichever was more appropriate. was done following the procedure outlined Small mammals were captured with live−traps in Lekagul and McNeely (1988) and Francis for threeคลังความรู้ดิจิทัล consecutive nights มหาวิทยาลัยเกษตรศาสตร์ per month between (2008). Body length (cm), weight (g) and sex July and September, 2016 (rainy season) and were determined and recorded, after which the between November to December, 2016 and captured animalsหมดอายุวันที่ were tagged 27-10-2564and released January, 2017 (cold season), at both sites. There in the grid where they were trapped and the were a total of 1,458 nights (3 trap nights × 6 traps were reset again in the afternoon. months × 81 traps) during which the trapping

Figure 3 Species accumulation curves found in logged and unlogged areas, Mahamyaing Reserve Area, PMWS. Journal of Tropical Forest Research 1 (2) : 1-20 (2017) 7

Data analyses H’ = -Σ pi ln (pi), where pi = the proportion Trap success of the ith species in the total sample, and ln Trap success (TS) was used to determine = natural logarithm with base e (Magurran, the relative abundance of the caught species 2005). Additionally, similarities in pairs of (Stanley et al., 1996, Barnett et al., 2002). It small mammal communities were assessed was calculated by dividing the total number of with quantitative Sorensen’s index: Cn = 2jN/ caught (Tc) with the number of traps (aN + bN), where jN = minimum number of used and trap nights (Tn; 4,374 for the entire individual in the ith species captured in both survey and both forest area types). A trap communities, aN = total individuals captured night is defined as a trap in use for a 24-hour in habitat a, and bN = total individuals captured period from sunrise to sunset. In addition, the in habitat b (Magurran, 1988). relative abundance is a qualitative analysis of the abundance of animals (Pettingill, 1950). Population size It was derived by dividing the number of Population size (N) and density (D) species caught with the number of time surveys are important variables in wildlife ecology conducted. studies. We used two commonly used methods, i.e., capture-mark-recapture method, the Sex ratio Lincoln-Peterson Estimator (White et al., 1982; Sex ratio was calculated based on Menkens and Anderson, 1988) and the program the proportion of the size of captured male CAPTURE (Cooch and White, 2013), to estimate and female population during the six month the population size of small mammals in the period. We determined the gender of captured PMWS. The first method assumes that the small mammals in all ages by visually examining proportion of marked individuals recaptured the anogenital distance. The distance between in the second sample represents the proportion the anus and the urogenital region is relatively of marked individuals in the population as a short in female compared to that whole. In this study we assumed a demographic in males (Manno 3rd, 2008). A Chi-square closure in which there are no births, deaths, (χ²) goodness of fit was used to determine immigration, and emigration during the study the difference between the genders from a period. 1:1 sex ratio in logged and unlogged forests CAPTURE is a comprehensive computer during theคลังความรู้ดิจิทัล two seasons (Terman มหาวิทยาลัยเกษตรศาสตร์ and Sassaman, program to calculate population size based on 1967). a constant capture probability model, with

Mo being the หมดอายุวันที่time variation model, 27-10-2564 Mt; the

Species diversity and similarity behavioral response model, Mb; the individual Species diversity of small mammals heterogeneity model, and the combination of was also computed by using Shannon’s Index: Mb and Mh or Mbh. More details about the 8 Journal of Tropical Forest Research 1 (2) : 1-20 (2017) software are mentioned in Cooch and White RESULTS AND DISCUSSION (2013). Diversity of small mammal communities In order to assess individual A total of 201 individual small heterogeneity, we calculated a coefficient of mammals belonging to 2 orders (Rodentia and variation (CV= standard error (SE)/average Scandentia), 3 families and 10 species were number of individuals) in capture probabilities captured. There were five species Berylmys( bowersi, Rattus rattus, R. norvegicus, R. nitidus, of all small mammals (Chao et al., 1992). CV and indica) belonging to values of 0 indicate that the investigated species family; four (Callosciurus pygerythrus, C. is equally captured and CV values less than erythraeus, C. finlaysonii, and Dremommys 20% are generally acceptable among wildlife rufigenis) belonging to Sciuridae, and only scientists (White et al., 1982). one species, Tupaia belangeri, belonging to Tupaiidae family (Table 1).

Table 1 Small mammal species caught in Mahamyaing Reserve Area in the proposed Mahamyaing Wildlife Sanctuary (2015-2016).

Number of individuals capture (%) Order Family Species Unlogged Logged Total Areas areas Rodentia Muridae Berylmys bowersi 3 (3%) 0 (0%) 3 (1%) Rodentia Muridae Rattus nitidus 4 (5%) 2 (2%) 6 (3%) Rodentia Muridae R. norvegicus 4 (5%) 5 (4%) 9 (4%) Rodentia Muridae R. rattus 21 (24%) 29 (26%) 50 (25%) Rodentia Muridae Bandicota indica 2 (2%) 14 (12%) 16 (8%) Rodentia Sciuridae Callosciurus 4 (5%) 5 (4%) 9 (4%) pygerythrus Rodentia Sciuridae C. erythraeus 5 (6%) 2 (2%) 7 (3%) Rodentia Sciuridae C. finlaysonii 8 (9%) 16 (14%) 24 (12%) Rodentia Sciuridae Dremommys rufigenis 9 (10%) 2 (2%) 11 (5%) Scandentia Tupaiidae Tupaia belangeri 28 (32%) 38 (34%) 66 (33%) Total 88 (100%) 113 (100%) 201 (100%)

คลังความรู้ดิจิทัล มหาวิทยาลัยเกษตรศาสตร์ หมดอายุวันที่ 27-10-2564 Journal of Tropical Forest Research 1 (2) : 1-20 (2017) 9

Table 2 Species richness and species diversity at different disturbance levels in Mahamyaing Reserve Area in the proposed Mahamyaing Wildlife Sanctuary.

Unlogged areas Logged areas (no. of captured individuals) (no. of captured individuals) Open Closed Closed Disturbed Species Total Agriculture Succession Total forest forest 1 forest 2 forest Berylmys bowersi 2 0 1 3 0 0 0 0 Rattus norvegicus 3 1 0 4 2 3 0 5 R. nitidus 3 1 0 4 1 0 1 2 R. rattus 9 5 7 21 10 10 9 29 Bandicota indica 0 1 1 2 5 5 4 14 Callosciurus pygerythrus 2 1 1 4 2 1 2 5 C. erythraeus 1 3 1 5 1 0 1 2 C. finlaysonii 2 2 4 8 7 6 3 16 Dremomys rufigenis 1 3 5 9 1 0 1 2 Tupaia belangeri 7 15 6 28 19 8 11 38 Number of individual 30 32 26 88 48 33 32 113 % trap success 3.29 2.26 2.19 2.01 2.06 2.19 1.78 2.58 Number of species 9 9 8 10 9 6 8 9 Species diversity (H’) 1.93 1.70 1.89 1.94 1.72 1.62 1.7 1.74

We observed 113 individuals of 9 small Although, this species is a generalist species mammal species in the logged areas, resulting and can be found in a wide variety of habitats in a 2.58% capture success. In the unlogged from disturbed primary and secondary forests areas, we captured 88 individual from the 10 to abandoned shifting agriculture, scrub, moist species, resulting in a 2.01% trap success (Table deciduous, and evergreen patches (Bernard 2). The overall trap success was 2.31%, which et al., 2009), but its preferred habitat is that was lower than the mean capture success in which has a continuous tree cover (Molur et a study conducted in the rainforests of Brazil al., 2005). (4.57%) (Conde and Rocha, 2006) and in Himalayan Field Rat (R. nitidus) urban green spaces in Ghana (7%) (Gbogbo and Northern Tree shrew (T. belangeri) were et al., 2017). This was largely due to hunting common to both forest management types and pressure from local people living inside and contributed 25% and 33% individuals in the in the bufferคลังความรู้ดิจิทัล zones of the มหาวิทยาลัยเกษตรศาสตร์ PMWS. It should be total captured population, respectively. They are noted that there were more species but fewer recognized as very adaptable species and can individuals in the unlogged areas (10 species, occur in variousหมดอายุวันที่ forest types includingcropland 27-10-2564 88 individuals) than in the logged areas (9 and human settlements (Molur et al., 2005; species, 113 individuals). White-toothed rat, B. Aplin et al., 2016). In contrast, R. nitidus, bowersi, was not recorded in the logged areas. Brown Rat (R. norvegicus), Hoary-bellied 10 Journal of Tropical Forest Research 1 (2) : 1-20 (2017)

Squirrel (C. pygerythrus), and Pallas's Squirrel al., 2005). In this study, three individuals were (C. erythraeus) were rare in both unlogged and observed in the unlogged areas and none were logged areas. In addition, only 2% of the total found in the logged areas (Table 2) because number of caught small mammals consisted of lack of interlocking crown covers (Wai, of D. rufigenis in logged areas. 2018). Greater rat and Finlayson’s The number of captured species was Squirrel were more frequently found in the less than that resulting from the preliminary logged areas (Table 2). In Lao PDR, these survey and discussion with the 10 villagers two species are commonly found in fields and living both inside and in the buffer zones. It villages and are very tolerant to degradation should be noted that the survey was conducted and fragmentation (Duckworth et al., 2008; during February to March in 2015, before the Aplin et al., 2016). Himalayan squirrel, Asian actual field work (July 2016−January 2017) house mouse, and Asiatic long-tail climbing in order to familiarize with the study area and mouse can be found living alongside humans used methodology. There were 17 species in fruit tree and coconut plantations (Lekagul found during the preliminary survey. The and McNeely, 1988; Francis, 2008; Smith most common species found in both the sites and Xie 2008). The remaining species were were R. rattus, B. indica, C. finlaysonii, and categorized as arboreal (e.g., Black giant T. belangeri. One of the more rarer species squirrel, Perny's long-nosed squirrel). Lunde caught in unlogged areas was B. bowersi. and Molur, (2016); Smith and Xie, (2008) Missing species were the black giant squirrel reported that in tropical and subtropical montane (Ratufa bicolor), Himalayan squirrel (Tamiops forests or nocturnal species (Particoloured macclellandii), Swinhoe's striped squirrel flying squirrel) (Krishna et al., 2013). Our (T. swinhoei), Perny's long-nose squirrel (D. traps were placed on the ground, fallen logs pernyi), Particolored flying squirrel Hylopetes( and low branches, and they may, therefore, alboniger), Polynesian rat (R. exulans), Asian not have been able to capture these species. house mouse (Mus musculus), Asiatic long- In contrast, the three additional species were tail climbing mouse (Vandeleuria oleracea), widespread and locally abundant, at least lesser bamboo rat (Cannomys badius), and in areas with suitable tree cover (e.g., open Asian house shrew (Suncus murinus). woods, plantations, degraded forests, and Additionally, three species were recorded dense forest (Lekagul and McNeely, 1988). which were not listed in the preliminary list, Further long-term studies are recommended in includingคลังความรู้ดิจิทัล white-toothed มหาวิทยาลัยเกษตรศาสตร์rat, greater bandicoot order to monitor the small mammal population rat and Finlayson’s squirrel. White-toothed dynamics. rat is generally present in a wide variety of หมดอายุวันที่ 27-10-2564 habitats where there is a suitable continuous Species richness and diversity tree cover, which include forest (intact or Species accumulation curves indicate disturbed conditions) to plantations (Molur et that there were smaller mammal assemblages Journal of Tropical Forest Research 1 (2) : 1-20 (2017) 11 captured in the logged areas during the early areas remained constant during November or part of the rainy season (July-August) compared early cold season (Figure 3). These patterns o that in the unlogged areas. However, the indicated that species richness of both areas species accumulation curve of the unlogged was greater in the cold season. This is mainly areas indicated to a rapid upward trend during due to an increased source of food as many theFigure late rainy 2Small season mammal and the sampling curves design,of both Mahamyaingtrees have Reserve fruits Area,that ripen PMWS. in cold weather.

Figure 3 Species accumulation curves found in logged and unlogged areas, Mahamyaing FigureReserve 3 Species Area,accumulation PMWS. curves found in logged and unlogged areas, Mahamyaing Reserve Area, PMWS. The H' diversity index of small mammals individuals, in each species, captured in open in unlogged areas was 1.94, while it was 1.70 and closed forest (2) were not a dominant in logged areas. Although we captured 9 small species (Table 3). Our results differ with the mammal species and 48 individuals in the previous studies (Mitchell et al., 1997; Conde agriculture areas (severely disturbed area), and Rocha, 2006; Gbogbo et al., 2017), who which was greater than the two individuals indicated that a higher species richness was captured in open and closed forest (relatively observed in a moderate disturbance. This is undisturbed),คลังความรู้ดิจิทัล its H’ diversity มหาวิทยาลัยเกษตรศาสตร์ index was only due to the fact that plots located in open and 1.72. This is largely due to an abundance of closed forest have been relatively disturbed by T. belangeri (28 individuals), and R. rattus (10 humans throughหมดอายุวันที่ activities such as 27-10-2564 illegal cutting individuals). In addition, the Shannon’s Index of big trees. As a result, these forests provide determines the neutral weight for all species a more heterogeneous habitat than heavily (Magurran, 2005). In contrast, the number of disturbed forest areas and human-dominated 12 Journal of Tropical Forest Research 1 (2) : 1-20 (2017) habitats (agriculture and early succession) in grasslands (Soontornpitakkool, 1996), both the logged forest areas, as well as man-made of which are relatively more homogeneous. Table 3 Species similarity between the different disturbance levels with the associated Quantitative Sorenson’s Index (number of shared species observed in the two habitats). Unlogged areas Logged areas Habitat Open Closed Closed Agriculture Succession Disturbed forest forest forest 1 forest 2 Unlogged forests 0.61 0.68 0.64 0.70 0.74 Open forest 1 (8) (7) (8) (5) (8) 0.66 0.70 0.55 0.72 Closed forest 1 1 (7) (9) (6) (8) 0.57 0.64 0.69 Closed forest 2 1 (7) (5) (7) Logged forests 0.79 0.80 Agriculture 1 (6) (8) 0.77 Succession 1 (5) Disturbed forest 1 Note: Similarity index between unlogged and logged forests is 0.73 within the 9 shared species.

คลังความรู้ดิจิทัล มหาวิทยาลัยเกษตรศาสตร์ หมดอายุวันที่ 27-10-2564

Figure 4 Monthly species diversity index of small mammals in the logged and unlogged areas, Figure 4Monthly species diversity index of small mammals in the logged and unlogged areas, proposed Mahamyaing Reserve Area, PMWS. proposed Mahamyaing Reserve Area, PMWS.

Journal of Tropical Forest Research 1 (2) : 1-20 (2017) 13

Meanwhile, the monthly diversity indices with similarity indices rank ranging between calculated for the cold season (November− 0.6−0.8. However, the composition of small January) were greater than that during the mammal community (2) between closed forest rainy season (July−September). The lowest and agriculture habitat (value of 0.6), and H’ index of 0.96 was observed in August in between closed forest habitat (1) and succession unlogged areas, while an H’ index of 1.35 was (value of 0.5) are significantly different. This observed in September in the logged areas is because the two closed forest types have (Figure 4). In both areas, the highest monthly dense vegetation and high percentage of crown species diversity was obtained in November. cover, while agriculture and the early succession During the survey, we noticed surface run-off stage have more open spaces and few standing in both the unlogged and logged areas, and trees. These characteristics result in different sometimes, the vegetation structure of shrubs ecological niches for small mammals (Conde and understory was destroyed and that trees and Rocha, 2006). fell as a result of the strong wind and heavy rain during the rainy season. In November, Sex ratio small mammals feed on fruits and nuts and Table 4 indicates that there were 39 store paddy and other grains in various places male and 49 female small mammals (sex ratio (Sims, 1992). For example, Mangifera indica of 1:1.26; male: female) that were captured in is ripe in the cold season and attracts many the unlogged areas compared to 58 males and small mammals to the areas. In addition, the 55 females in the logged areas (sex ratio of paired means t-test indicated that the monthly 1:0.95, male: female), during the study period. diversity indices between the unlogged and In addition, the results from the Chi-square logged areas were significantly different at test indicated that the male: female sex ratio 95% confidence level . in the unlogged (χ2 = 40.00, p < 0.01) and In addition, the quantitative Sorensen the logged areas ((χ2 = 101.75, p < 0.01) was index values for the pairs of six disturbance statistically different from 1:1 ratio. Similarly, levels (three for unlogged forests and three the proportion of males (97 individuals) and for logged forests) are shown in Table 4 and females (104 individuals) was significantly indicate that the small mammal composition different (χ2 = 143.25, p < 0.01). The sex ratios among the 3 habitat types of unlogged forests of two small mammal species, B. indica, and and among the 3 logged forests are similar R. nitidus, were in balance. คลังความรู้ดิจิทัล มหาวิทยาลัยเกษตรศาสตร์ หมดอายุวันที่ 27-10-2564 14 Journal of Tropical Forest Research 1 (2) : 1-20 (2017)

Table 4 Gender of small mammals caught in Mahamyaing Reserve Area, the proposed Mahamyaing Wildlife Sanctuary (2015-2016)

Unlogged areas Logged areas Total Species Male Female Male Female Male Female Berylmys bowersi 1 2 0 0 2 2 Rattus nitidus 1 3 2 0 3 3 R. norvegicus 2 2 2 3 4 5 R. rattus 6 15 11 18 16 33 Bandicota indica 1 1 7 7 8 8 Callosciurus 1 3 2 3 3 6 pygerythrus C. erythraeus 2 3 1 1 3 4 C. finlaysonii 3 5 11 5 14 10 Dremommys 4 5 1 1 5 6 rufigenis Tupaia belangeri 18 10 20 18 38 28 Total 39 49 58 55 97 104 ratio 1.00 1.26 1.00 0.95 1.00 1.07 χ2 , p 40.00 <0.01 101.75 <0.01 143.25 <0.01

The survey results showed that the male 113 individuals were captured in the logged population was dominant in the rainy season forest areas. Out of these, 23 individuals were but female population was dominant in the recaptured in unlogged and 27 in logged areas cold season. Many factors affect this variation (Table 5). Among all 10 species, the number as reported by O’Connell (1989), who studied of individuals captured from the R. rattus the seasonal changes in Neotropical small and T. belangeri species was more than 20 mammal population due to seasonal habitat individuals in both areas. According to White variation. Moreover, reproductively mature et al. (1982), closed capture-recapture models males in many species usually have a lower can be used for estimating populations of survival rate than the females (Gaillard et al., only these species . In addition, the M model 1993) and there are certain costs for engaging h was selected to estimate population size of in reproduction (Yoccoz et al., 2002) . Long the two species, in either the unlogged or term monitoring of such factors needs to be logged areas. The other remaining eight species done in the area . คลังความรู้ดิจิทัล มหาวิทยาลัยเกษตรศาสตร์ had a lesser number of individuals and Population size and density higher CV values and so, it was not possible Eighty-eight individual small mammals to calculate theirหมดอายุวันที่ population size27-10-2564 using the were captured in unlogged forest areas and CAPTURE model. Journal of Tropical Forest Research 1 (2) : 1-20 (2017) 15 na 3-10 6-22 6-22 3-10 3-10 42-83 18-64 26-62 59-109 95% CI

4 8 8 4 4 N na 55 28 36 76 na SE 1.73 3.18 3.18 1.73 8.85 1.73 10.06 10.18 12.52 p na 0.11 0.11 0.11 0.08 0.13 0.13 0.08 0.07 0.08 na 18 36 25 16 43 40 40 43 43 CV Logged areas h h (1.0) (1.0) (1.0) (1.0) (1.0) na M M (0.88) (0.88) o h h h h h h Model M M M M M M M 8 4 3 0 0 1 1 0 0 10 27 RC 0 2 5 5 2 2 C 29 14 16 38 113 3 4-4 4-4 4-4 3-10 8-25 9-24 25-52 40-78 14-161 95% CI = the constant capture probability model; CV = coefficient of variation; p = probability = p variation; of coefficient = CV model; probability capture constant the = o 3 4 4 4 4 N 11 31 13 35 52 SE 2.13 2.47 2.47 6.34 1.73 2.47 4.01 3.19 9.55 29.05 p 0.18 0.17 0.17 0.17 0.08 0.17 0.06 0.19 0.05 0.12 71 62 62 20 43 62 31 29 83 18 CV Unlogged areas h h h h h h (1.0) (1.0) (1.0) M M M M M M (0.84) o o h h Model M M M M = the individual heterogeneity model; M model; heterogeneity = the individual h 0 0 0 0 0 0 4 1 8 10 23 RC 3 4 4 2 4 5 8 9 C 21 28 88 คลังความรู้ดิจิทัล มหาวิทยาลัยเกษตรศาสตร์ หมดอายุวันที่ 27-10-2564 Small mammal population estimation in Mahamyaing Reserve Area, the proposed Mahamyaing Wildlife Sanctuary. Wildlife Area, the proposed Mahamyaing Reserve population estimation in Mahamyaing Small mammal Species of capture; SE = standard error; N estimated population C = Capture; RC = recaptured; M C = Capture; RC recaptured;

Table 5 Table Berylmys bowersi Rattus nitidus R. norvegicus R. rattus Bandicota indica Callosciurus pygerythrus C. erythraeus C. finlaysonii rufigenis Dremommys belangeri Tupaia Total Notes: 16 Journal of Tropical Forest Research 1 (2) : 1-20 (2017)

The program CAPTURE estimates the and a total of 88 individuals were caught in the population size using the Jackknife estimator unlogged areas. There were only nine species based on the Mh model and it revealed that the caught in the logged areas but the number of estimated population size (N) for R. rattus in individuals was greater (113 individuals). the unlogged areas was 31, and its standard Among the identified species, the three most error was 6.34. The coefficient of variation (CV) common were T. belangeri, R. rattus, and C. of this species was 20%, which is acceptable finlaysonii. In addition, there was a higher species among wildlife scientists (White et al., 1982). richness and diversity in the unlogged areas The estimated population for R. rattus in the compared to the logged areas. For example, logged areas was 55 individuals/ha (CV = B. bowersi was not detected in the logged 18%), indicating to a higher precision for areas. However, there was an abundance and the population size estimate. The estimated higher density in the unlogged areas as a result population of T. belangeri was considerable of increase in ground vegetative cover and greater than that of R. rattus in both the areas habitat heterogeneities. Species composition and the CV values were relatively less than among the 3 disturbance levels within the same that of R. rattus. Using the above figures and management (unlogged and logged) was similar sampling plot size, the densities for R. rattus (as indicated by the quantitative Sorenson’s in unlogged and logged areas were estimated index values ranging between 0.7−0.8), but the at 9.56 individuals/ha and 16 individuals/ha, species similarity across the various habitat respectively. Comparatively, the density for types (e.g., between closed forest 2 habitat and T. belangeri in unlogged and logged areas was agriculture habitat and between closed forest estimated at 16.97 individuals/ha and 23.4 1 habitat and succession) was substantially individuals/ha, respectively. This is because different (values between 0.5-0.6). The results these two species are highly associated with indicate that the logged forest has a higher cropland and human settlements (Molur et level of significance than the unlogged forest al. 2005; Aplin et al., 2016). In addition, the in terms of the small mammal community. logged forests not only have a different floral Although there were less native species, but composition and distribution of components the small mammals that inhabit the logged (Cannon et al., 1998), but also differ in their forest are more abundant and very tolerant of proportion of -dispersed and mammal- degradation and fragmentation. They play an pollinatedคลังความรู้ดิจิทัล trees (Chazdon, มหาวิทยาลัยเกษตรศาสตร์ 2003). important role in supporting food to small− and medium−sized carnivores living higher CONCLUSION up in the foodหมดอายุวันที่ chain. In addition, 27-10-2564 a higher Two hundred and one small mammals proportion of native species are expected, if were captured and 10 small mammal species protection of remaining forests together with were identified. Ten small mammal species restoration of degraded habitats is effectively Journal of Tropical Forest Research 1 (2) : 1-20 (2017) 17 implemented and such measures would enhance groves as forest refugia. J. Trop. Ecol. the ecological value of the PMWS. 41 (2): 127- 142. Bendix, J., Jr. Wiley, J. John and M.G. Commons. ACKNOWLEDGEMENTS 2017. Intermediate disturbance and patterns of species richness. J. Physical The authors would like to express Geography. 38 (5): 393-403, doi: their gratitude to H.R.H Princess Maha Chakri 10.1080/02723646.2017.1327269. Sirindhorn and The Graduate School of Bernard, H., J. Fjeldså and M. Mohamed. Kasetsart University for providing scholarship 2009. A case study on the effects to Khine Thazin Wai for her master degree at of disturbance and conversion of Kasetsart University. The authors also thank tropical lowland rain forest on the the Faculty of Forestry, Kasetsart University non-volant small mammals in north for waiving her tuition fees. Special thanks Borneo: management implications. to Palle Havmoller for the technical and J. Mammal. 34 (2): 85-96. English editing and to the reviewers for their Cannon, C.H., D.R. Peart and M. Leighton. 1998. constructive comments that helped to improve Tree species diversity of commercially this manuscript. logged Bornean rainforest. J. Sci. 281: 1366-1368. REFERENCES Chao, A., S-M. Lee and L. Jeng. 1992. Estimating Adler, G. H. and R. Levins. 1994. The island Population Size for Capture-Recapture syndrome in populations. J. Data When Capture Probabilities The Quart. Review of Biol. 6 (4): Vary by Time and Individual Animal. Biometrics. 48. 201-16. . 473-490. Chazdon, R.L. 2003. Tropical forest recovery: Aplin, K., D. Lunde and S. Molur. 2016. Rattus legacies of human impact and natural nitidus. The IUCN Red List of disturbances. Perspectives in Plant Threatened Species 2016. http://dx.doi. Ecol., Evol. Sys. 6 (1): 51-71. org/10.2305/IUCN.UK.20163.RLTS. Conde, C.F. and V.F.D. Rocha. 2006. Habitat T19352A22443559.en (Accessed 04 Disturbance and Small Mammal December 2017) Richness and Diversity in an Atlantic Aung, M. 2007. Policy and practice in Myanmar's Rainforest Area in Southeastern Brazil. คลังความรู้ดิจิทัลprotected area system.มหาวิทยาลัยเกษตรศาสตร์ J. Environ. J. Biol. 66 (4): 983-990. Manage. 84 (2007): 188-203. Connell, J.H. 1978. Diversity in tropical rain Barnett, A.A., N. Read, J. Scurlock, C. Low, H. forestsหมดอายุวันที่ and coral reefs. 27-10-2564 J. Sci. 199 Noris and R. Shapley. 2002. Ecology (4335): 1302-1310. of Rodent communities in agricultural Cooch, E. and G.C. White. 2013. Program habitats in eastern Sierra Leone: Cocoa Mark: A Gentle Introduction, eighth 18 Journal of Tropical Forest Research 1 (2) : 1-20 (2017)

ed., 659 pp. http://www.phidot.org/ (UK) Ltd. and Asia Books Co., Ltd, software/mark/docs/book/. (Accessed Bangkok, pp. 392. 29 July 2017). Istituto Oikos and Biodiversity and Nature Duckworth, J.W., R. Timmins and M. Parr. Conservation Association (BANCA). 2008. Callosciurus finlaysonii. The 2011. Myanmar Protected Areas: IUCN Red List of Threatened Species Context, Current Status and 2008: e.T3596A9964363. http://dx.doi. Challenges. Milano, Italy: Ancora org/10.2305/IUCN.UK.2008.RLTS. Libri. T3596A9964363.en Krishna, M., A. Kumar, P. Chandra, K. Sarma, Eric, A. Ricart, Danilo S. Balete and Lawrence A. Devi and M.L. Khan. 2013. Impact R. Heaney. 2007. Habitat Disturbance of road widening on wildlife in and the Ecology of Small Mammals Namdapha National Park, Arunachal Pradesh, : a conservation issue. in the Philippines. J. Environ. Sc. Asia. J. Con. Bio. 291: 76-78. and Manage 10 (1): 34–41 Lekagul, B. and J.A. McNeely. 1988. Mammals Gaillard, J.M., D. Delorme and J.M. Jullien. of Thailand. 2nd ed. Darnsutha Press. 1993. Effects of cohort, sex, and birth Bangkok. date on body development of roe Lima, M., N.C. Stenseth and F.M. Jasic. 2002. deer (Capreolus capreolus) fawns. Food web structure and climate effects J. Oecologia 94 (1): 57-61. on the dynamics of small mammals Gbogbo, F., K. Tabiri and M. Yahaya. 2017. and owls in semi-arid Chile. Ecol. Diversity and abundance of small Letters 5 (2): 273-284. mammals along a disturbance gradient Lunde, D. and S. Molur. 2016. Dremomys on a university campus in Ghana. Int. pernyi. The IUCN Red List of J. Ecol and Devel. 32 (1): 0973-7308. Threatened Species 2016. http:// Gunther, P. M. and G.D. Horn and G.D. Babb. www.iucnredlist.org/details/6822/0 1983. Small mammal populations and (Accessed 02 December 2017) food selection in relation to timber Lwin, N., T. Geissmann, S.S. Aung, T.N. Aung, harvest practices in the Western Cascade H.Z. Myo, G. T. Htin, M. Mark and Mountain. Northwest Sci. 7(1). F. Momberg. 2011. The Myanamr FAO. 2016. Global Forest Resources hoolock gibbon conservation staus คลังความรู้ดิจิทัลAssessment 2015 มหาวิทยาลัยเกษตรศาสตร์. Rome, Food and review: first results. J. Gibbon. 6: Agriculture Organization of United 18-21. Nations, Rome. Lynam, A.J. 2003.หมดอายุวันที่ A national 27-10-2564 tiger action Francis, C. M. 2008. A field guide to the plan for the Union of Myanmar. New mammals of Thailand and South- York: Myanmar Forest Department East Asia. New Holland Publishers and the Wildlife Conservation Society. Journal of Tropical Forest Research 1 (2) : 1-20 (2017) 19

Magurran, A.E. 2005. Species abundance Passamani, M. and D. Ribeiro. 2009. Small distributions: pattern or process? J. mammals in a fragment and adjacent Fun. Ecol. 19 (1): 177-181. matrix in southeastern Brazil. Braz J. Manno 3rd, F.A. 2008. Measurement of the digit Biol. 69 (2): 305-309. lengths and the anogenital distance in Püttker, T., R. Pardini, Y. Meyer-Lucht and mice. J. Physiol. Behav. 93 (1-2): S. Sommer. 2008. Responses of five 364-8 small mammal species to micro-scale Menkens, G.E. and S.H. Anderson. 1988. variations in vegetation structure in Estimation of Small‐Mammal Population secondary Atlantic Forest remnants, Size. J. Ecol. 69 (6): 1952-1959. Brazil. BMC Ecol. 8:9 doi:10.1186/1472- Ministry of Environmental Conservation and 6785-8-9. Forestry. 2014. Fifth National Report Sims, C.A. 1992. Interpreting the macroeconomic to the United Nations Convention on time series facts: The effects of monetary Biological Diversity. Nay Pyi Taw, policy. J. European econ. review. 36 Republic of the Union of Myanmar. (5): 975-1000. Mitchell, J.C., S.C. Rinehart, J.F. Pagels, K.A. Singer, N.J.D. and G.M. Rongers. 2014. A Buhlmann and C.A. Pague. 1997. classification of New Zealand’s Factors influencing amphibian and terrestrial ecosystems. Department of small mammal assemblages in central conservation, The Terrace, Wellington, Appalachian forests. J. For. Ecol. and New Zealand. Manage. 96 (1-2): 65-76. Sodhi, N.S., L.P. Koh, B.W. Brooks and P.K.L. Molino, J. and D. Sabatier. 2001. Tree Diversity Ng. 2004. Southeast Asian biodiversity: in Tropical Rain Forests: A Validation an impending disaster. J. Trends in of the Intermediate Disturbance Ecol. and Evol. 19: 654-660. Hypothesis. J. Sci. 294: (5547). Soontornpitakkool, S.1996. Population Molur, S., C. Srinivasulu, B. Srinivasulu, S. characteristics of rats and mice Walker, P.O. Nameer and L. Ravikumar. in evergreen forest and grassland 2005. Status of south Asian non- at Khao Yai National Park. M.Sc. volant small mammals: conservation thesis, Faculty of Forestry, Kasetsart assessment and management plan University, Thailand. (CAMP). Workshop report. Zoo Smith, A.T. and Y. Xie. 2008. A Guide to คลังความรู้ดิจิทัลOutreach Organisation/CBSG-South มหาวิทยาลัยเกษตรศาสตร์ the Mammals of . Princeton Asia, Coimbatore, India. University Press, Princeton, New Jersey. O'Connell, M.A. 1989. Population dynamics Stanley, W.T., หมดอายุวันที่S.M. Goodman and 27-10-2564 R. Hutterer. of Neotropical small mammals in 1996. Notes on the insectivores and seasonal habitats. J. Mammalogy 70 elephant shrews of Chome Forest, South (3): 532-548. Pare Mountains. Tanzania. (Mammalian: 20 Journal of Tropical Forest Research 1 (2) : 1-20 (2017)

Insectivora et. Macroscelididae). J. logged and unlogged areas, proposed Zool. Abh. Tierkd. 49 (8): 131-147. Mahamyaing Wildlife Sanctuary, Terman, C. and J. Sassaman. 1967. Sex Myanmar. M.S. Thesis. Graduate ratio in deer mouse populations. J. School, Kasetsart University, Bangkok. Mammal. 48(4): 589-597. http://www. White, G.C., D.R. Anderson, K.P. Burnham and jstor.org/stable/1377582. (Accessed D.L. Otis. 1982. Capture-recapture 12 May 2017) and removal methods for sampling Tordoff, A.W., J.C. Eames, K. Eberhardt, M.C. closed populations. Los Alamos Baltzer, P. Davidson, J. Leimgruber, U. National Laboratory, Los Alamos. Uga and U.A. Than. 2005. Myanmar: Yoccoz, N.G., K.E. Erikstad, J.O. Bustnes, Investment Opportunities in S.A. Hanssen and T. Tveraa. 2002. Biodiversity Conservation. Birdlife Costs of reproduction in common International. http://birdlifeindochina. eiders (Somateria mollissima): an org/content/myanmar-investment- assessment of relationships between opportunities-biodiversity-conservation. reproductive effort and future survival (Accessed 15 July 2017) and reproduction based on observational Wai, K.T. 2018. Detecting changes in small and experimental studies. J. Applied mammal community structure in Statis. 29 (1-4): 57-64.

คลังความรู้ดิจิทัล มหาวิทยาลัยเกษตรศาสตร์ หมดอายุวันที่ 27-10-2564