Diversity and Composition of Tropical Metropolitan ants (Hymenoptera: Formicidae): An Indicator to Urban Biodiversity and Ecosystem Health

Ant communities are highly sensitive to niche changes; therefore, they are useful indicators of ecosystem health. Little to nothing is known about ant communities inhabiting Kuala Lumpur’s metropolitan parks. Therefore, we aimed to survey ant species diversity and composition in 12 parks within the Federal Territory of Kuala Lumpur, Malaysia. Forty-nine species from 29 genera and six subfamilies were identied from 9,677 individuals in 1,008 eld collections, made using a combination of hand-collection and pitfall and yellow pan methods, from February to July 2019. The species accumulation curve reached a plateau after approximately 80 samplings. The difference in ant diversity between parks was nonsignicant. The dominant ant species was Oecophylla smaragdina, followed by Carebara diversa and Anoplolepis gracilipes. This study provides the rst ant checklist, comprising information on eight functional groups, seven invasive and several forest-associated species, and one new locality record, for highly urbanized areas in Peninsular Malaysia. Our results provide a basis for future studies to better understand the potential consequences of urbanization on ant communities and their ecology in urban settings.

Urbanization generates a cascade of secondary effects on habitat quality, including increased fragmentation, exposure to habitat edge effects, and introduction of non-native species (Marshall & Shortle, 2005). Consequently, it poses a signi cant threat to biodiversity (Fattorini, 2016). Therefore, the study of urban biodiversity is crucial not only for the purposes of recreation and wellbeing but also for conservation (Andersson et al., 2015).
Arthropods inhabiting urban green areas experience different outcomes (positive or negative) depending on a range of factors (McIntyre, 2000). Urban parks in central Brazil have been ranked the highest for conservation value compared to other types of urban green spaces as they support the most ant species, primarily native taxa. However, the abundance of specialist insects are reported to decline due to urbanization; herbivorous, parasitic, saprophagous, and ower-visiting insects are much more negatively affected by urbanization than predatory insects (Ye et al., 2013). In contrast, some generalist taxa, including aphids, cockroaches, and termites, may increase in abundance (Santos, 2016).
Ants serve as ideal ecological indicators of changes in the environment (Andersen, 1999;Brown, 1997;Kremen, 1992;Lawes et al., 2005;McGeoch, 1998;Santos, 2016;Uehara-Prado et al., 2009). Ant community patterns re ect the status and quality of their habitats, which range from natural forest to disturbed ecosystems/habitats (Andersen et al., 2003;Andersen & Mayer, 2004;Folgarait, 1998;Koch & Vohland, 2008;Underwood & Fisher, 2006). Ants yield vital information on intensi cation and/or rehabilitation processes in agroecosystems at low cost and within a relatively short time (Underwood & Fisher, 2006). In urban environments, they also serve as food sources for reptiles, birds, and mammals, which are highly dependent on the limited natural food resources in these habitats.
More than 800 ant species are reported from Malaysia. Of these, 23 are classi ed as urban pest species (Na & Lee, 2001 1994). Therefore, this study aims to comprehensively explore the diversity and composition of ants inhabiting metropolitan parks within the Federal Territory of Kuala Lumpur. The results of this study will provide baseline data to help predict the roles and functions of ants in urban development and ensure that urban expansion proceeds sustainably.

Study site
Detail and map of sampling locations across Federal Territory of Kuala Lumpur is presented in Table 1 and Fig. 1. Kuala Lumpur, or o cially known as Federal Territory of Kuala Lumpur, is the country's national capital which is located in the west of Peninsular Malaysia in the middle of the state of Selangor. Covering an area of 243 km 2 , the population is estimated 1.79 million as of 2018 (Dept. of Statistics Malaysia, 2020). The climate is tropical, hot and humid all over the year with minimum and maximum of the temperature estimated around 23°C and 35°C with annual rainfall of 2600 mm.  Andersen (2000) and Brown (2000).

Data analysis
Each species was subjected to frequency of occurrence (FO), presented in percentage, calculated by the total number of a species occurrence divided by the total number of collections (n = 1008), and as well as site occurrence (SO), calculated by the number of sites where a species was taken divided by the total sites (n = 12) were calculated in percentages (Ya'cob et al., 2016). The abundance distribution of ants from each park were tested for their t to the log series model. Sampling e ciency was obtained from species richness estimators and species accumulation curves. The mean of all estimators involved (Chao, Jack 1, Jack 2, Bootstrap) was calculated and compared with the observed species for an easier assessment of sampling e ciency. Species accumulation curves were plotted based on sample-based dataset. Diversity and richness indices were also calculated. For α-diversity, the indices used are Shannon-Weiner diversity index (H'), Simpson's (D), and Evenness J. For β-diversity comparison, Bray-Curtis dissimilarity index was used as a quantitative measure between communities experiencing changing environments. Normality tests were performed using Shapiro-Wilk (W) at 95% con dence level (p≤0.05). Data were not normally distributed and therefore Kruskal-Wallis was used to test the signi cance different between number of species and abundance at all sites. All calculations and plots were done using "vegan" and "ggplot2" packages in software RStudio version 1.2.5033 (R Core Team, 2020). Additionally, the ants collected were classi ed into functional groups and invasiveness status.
In term of total individuals collected, Oecophylla smaragdina, was the most abundant species (20%) while the others were collected at lower than 10%. Based on site occurrence (SO), Camponotus sp.1, Camponotus (Tanaemyrmex) sp.1, Cardiocondyla sp., Monomorium orientale, and Pheidole parva were collected at all urban parks (12 sites). However, thirteen (13) ant species, including Crematogaster quadriruga and Philidris sp., were collected only at one park.  Details of each subfamily recorded in this study are presented in Table 4. The highest number of species and individuals were from the Myrmicinae and Formicinae, while only a single species was recorded for the Ectatomminae. Species richness and estimated richness are shown in Table 5. The overall estimated species richness was between 52 and 60 species, with more than 81% sampling e ciency. The estimated species richness for each urban park ranged from 17 to 31 species, indicating more than 80% sampling e ciency. The species accumulation curve showed that the sample size of ants was su cient to accurately describe ant species diversity, since a plateau was reached after approximately 80 samplings (Fig. 2).   (Fig. 4).

Discussion
Ants are used as indicators to assess landscape conversion due to their rapid response to changes in climate/environmental parameters (Cuautle et al., 2016). Urban parks are some of the most highly disturbed areas within ecosystems, and their ora and fauna are expected to be less diverse than those of natural forests. This assumption is based on many factors, the most important being habitat complexity.
Urban habitats are less complex because of lower diversity of ecological structure compared to natural forests that offer greater habitat heterogeneity, which allows a wider variety of habitat choices for most ant species (Cuautle et al., 2016).
This study provides the rst checklist of ant species for Kuala Lumpur's urban parks. The urban parks in this study showed no signi cant differences in ant diversity and composition. This is probably due to all parks sharing similar ecological characteristics, such as the composition of common tree species and climate. This may also re ect the local councils' excellent parks management and maintenance. Several forest-a liated ant taxa, such as Gnamptogenys sp., Platythyrea parallela, Proatta butteli, and Strumigenys sp., were collected from the parks suggesting that their ecosystem status has improved to the point of preserving biodiversity. These forest-associated species are commonly found in rotten wood and leaf litter in tropical and subtropical mesic forests. Platythyrea parallela is common in logs, tree trunks, and rotten wood, and is deemed to be rare since observations of stable colony nests are few (Ashikin & Hashim, 2015). Proatta butteli is found in leaf litter and rotten logs in tropical rainforests and is important to the ecosystem as it recycles nutrients back into soil (Moffett, 1986). Strumigenys tends to associate with primary and secondary rainforest, but several species within this genus, i.e., S. emmae and S. membranifera, are tramp species found in open disturbed habitats (Tang et al., 2019). Another notable observation of this study is the presence of Dolichoderus thoracicus rufescens in Taman Dusun Bandar (TDB), which has been described to date only in Sarawak, East Malaysia. Our observation is the rst record for Peninsular Malaysia. Although the biology of this subspecies is unknown (Dill, 2002), the presence of this subspecies is most possibly due to the presence of various tropical fruiting trees which dominates the vegetation in this park.
The species accumulation curve reached an asymptote, which indicates that the sampling method applied in this study was effective enough to successfully sample most of the species present. Further sampling of ants and ant-associated environmental parameters at different times is vital to reveal the distributional patterns and ecology of metropolitan species.
Oecophylla smaragdina was the most common species found in this study, occurring in 10 parks, and dominating in seven. This species is well known for its predatory, aggressive, and territorial behavior, which potentially deters other species, resulting in low species diversity and evenness in sites where it occurs (Fig. 5). However, periodical surveillance of this predatory species is necessary to ensure urban biodiversity is in consistent balance. The success of this dominant species is due to its large body size, enormous colonies, extensive foraging range, and the ability to construct nests nearly everywhere from leaves of almost any shape (Bharti & Silla, 2011;Holldobler, 1983;Van Mele & Cuc, 2000). Despite being regarded as a threat, O. smaragdina also provides valuable bene ts for the urban park ecosystem as a biological control agent. Its aggressiveness and predatory habit help in the control of heteropteran, lepidopteran, and leaf-feeding coleopteran pests (Van Mele, 2008). Their wastes may provide useful nutrients for plants, which in turn aids in the nutrient cycle (Barzman et al., 1996).
Based on the Global Invasive Species Database (GISD) of the International Union for Conservation of Nature (IUCN) and Antkey.org, seven (14.3%) of the 49 species recorded in this study were categorized as invasive in this region. The presence of these invasive species was corroborated by a previous study of urban ants in the state of Penang, which demonstrates that they are common inhabitants of highly urbanized areas (Ab-Majid et al., 2016; Na & Lee, 2001). Invasive ants were found in all the parks sampled, regardless of the total number of ant species present. In fact, a few of the invasive species were among the most abundant and dominant in certain parks (Fig. 4). Their presence may be in uenced by other factors, for example, anthropogenic activities, since the urban parks are close to human habitation. However, the actual roles and functions of these invasive ants remain undetermined. Previous studies suggest that the introduction of invasive ants is disruptive to native species abundance and diversity (Dejean et al., 2007;Fournier et al., 2012;Strayer, 2012). Despite their invasive behaviors, this study showed that the percentage abundance of this group was positively correlated with species evenness (Fig. 6), indicating stabilized biodiversity. Therefore, they seem to have little impact on the ecosystem compared to "high-impact" invaders, which alter ecosystem functions (Strayer, 2012). Consequently, periodic surveillance of urban ants is necessary to ensure that biodiversity equilibrium persists in line with urban conservation strategy.
Our study found eight out of the nine ant functional groups proposed by Andersen (2000). Undoubtedly, Opportunists (O) are the most abundant group due to having a wide range of ecological preferences, high tolerance to harsh and highly modi ed environments, less sensitivity to small changes in edaphic factors, poor competitive behavior, and very large habitat distributions (Grime, 1979;Masse et al., 2021). In

Conclusion
Research on urban tropical habitats is crucial as it can increase the knowledge of ant species occurrence, allowing better assessment of the biodiversity and conservation potential of semi-natural habitats. This rst checklist of ants for highly urbanized areas in Peninsular Malaysia will become a foundational framework for future studies. We hope that data will be collected continuously to further improve the checklist and provide concrete information on the ecological relationships between invasive and native ant species.

Con ict of Interest/Competing interests
The authors have no con icts to declare.

Availability of data and material
Material is available at Universiti Malaya's Museum of Zoology (MZUM), Faculty of Science, Universiti Malaya.
Code availability Not applicable.
Authors' contributions