Introduction

Karst landscapes, i.e., those underlain by soluble rock such as limestone, occupy approximately 20% of the Earth’s ice-free terrestrial surface and in many parts of the world they have been noted for their importance to biodiversity as well as being of high socio-economic value through the provision of a range of ecosystem services (Bystriakova et al. 2019). In Southeast Asia in particular, limestone karsts are associated with high species richness and endemism, associated with the wide range of ecological niches within these complex landscapes and the high degree of adaptation required because of the alkaline conditions (Chin 1977; Clements et al. 2006; Bystriakova et al. 2019). Further, the often steep and rugged terrain of karst outcrops has largely spared them from historic agricultural encroachment, such that these sites may function as biodiversity “arks” (Clements et al. 2006). However, these landscapes are increasingly threatened by exploitation of their natural resources, such that their unique biodiversity is imperilled and conservation action at many karst sites is urgent.

The 445 limestone karsts in Peninsular Malaysia cover only 0.2% of its land area (Price 2014; Liew et al. 2016), but are exceptionally biodiverse, supporting at least 1,216 plant species (Chin 1977) or approximately 14% of the total vascular plant flora of Peninsular Malaysia. Most are typical tower karst formations, are relatively small (occupy about 1 km2 basal area), and are low, rising to less than 700 m elevation from the surrounding plain. They support limestone vegetation, an edaphic subtype of lowland forest (Wyatt-Smith 1963; Saw 2010), notable for the lack of dipterocarps, fruit tree families and calcifuge families like Ericaceae and Nepenthaceae (Chin 1977). In addition to the high plant diversity, many of the constituent species are restricted to growing on limestone, and/or are endemic in the Peninsula, including hyper-endemic species known from a single karst (Chin 1977).

Peninsular Malaysia is divided into 11 states. Land use matters are strictly under the control of the state, not the federal government. Hence, implementing protection for areas of conservation importance is solely a state matter. Ninety-five of the 445 karsts (over 20%) occur within Perak state (Fig. 1); Perak therefore has a high responsibility for the protection of the karst flora of Peninsular Malaysia. Together, these 95 karsts cover only 0.15% (32 km2) of the total 21,035 km2 land area of Perak (Price 2014; Liew et al. 2016; Fig. 1). The largest is Gunung Tempurung Prk 01 (gunung = mountain, henceforth G.) that covers more than 8 km2 and its highest peak rises to 612 m. Perak’s state capital, Ipoh, lies in the Kinta Valley, an area well known for its scenic karst landscape with 45 limestone hills (Cheang & Wong 2009).

Fig. 1.
figure 1

Map of limestone karst distribution in Peninsular Malaysia. Perak state is located in the North West of the Peninsula. Karsts are indicated in red.

Full size image

Besides their landscape beauty and rich biodiversity, karsts have multiple values and uses. These include important geological features such as caves, mirror lakes, hot springs and their associated unique ecosystems (Kiew 1997; Ros-Fatihah & Komoo 2003); cultural significance, notably cave temples that are also a major tourist attraction; recreational activities including rock climbing and ‘caving’; archaeological importance for their neolithic cave paintings and fossils; and, most importantly, economic value for state revenue from quarrying.

All of these activities have associated threats to the fragile karst ecosystems. For example the erection of infrastructure for Buddhist, Taoist and Hindu temples can damage the natural vegetation, and the expansion of agriculture immediately surrounding the isolated karsts eliminates the buffer zone of limestone forest. Particularly significant, however, is the irreversible damage caused by quarrying. In Perak, 52 quarries operate on 16 karsts; G. Lanno Prk 18 alone has 20 quarries. Some hills have been razed to the ground or reduced to a remnant of their former size (Price 2014; Fig. 2). Karsts are fragile because disturbance or destruction is permanent and will eliminate geological features, fossils, plant and animal species. As a result, many karst plant species are threatened and some are already extinct.

Fig. 2.
figure 2

Aerial photograph of G. Kanthan Prk 47 with the buffer zone of forest cleared to the foot of the karst for agriculture on one side and on the other side encroachment by the quarry. The last remaining foothill habitat with limestone forest in Perak is located in the valley between the intact ridge and the quarrying site, indicated by the arrow. photo: p. t. ong.

Full size image

The incompatibility of quarrying and conservation presents the dilemma of balancing economic development with conserving natural heritage. Given the increasing demand for cement in construction, it is clear that not every karst can be conserved, so the challenge is to determine which karsts can be quarried and which should be conserved to protect maximum biodiversity. In the 1970s, quarrying for cement and other limestone products accelerated to a massive scale and continues today. In response, the Malaysian Nature Society (MNS) conducted a one-year assessment of limestone hills in Kinta District, Perak, with the aim of advocating sustainable utilisation of limestone resources without affecting their outstanding natural features (MNS 1991). Of the 45 karsts surveyed, 14 were identified that should be legally protected based on their significant natural and cultural features. Four of these (G. Kanthan Prk 47, G. Datok Prk 36, G. Rapat Prk 36 and G. Tempurung Prk 01) were listed as top priorities for total protection. Besides their other values, botanically they harbour significant plant richness and many species are Perak range-restricted and endangered (Kiew 1991b) and their flora is/was relatively undisturbed. This assessment was followed by the Draft Structure Plan of Ipoh City Council 1997 – 2020 that recognised the need to protect karsts of heritage value and their threatened species, and listed these same four karsts as of paramount importance.

The majority of Perak’s karsts lie on state land and most are unprotected. A few in the Lenggong Valley are protected as a UNESCO Archaeological Heritage Site because of the valley’s importance as the site of the Perak Man, the earliest prehistoric human remains in Peninsular Malaysia (UNESCO 2012). Another three are protected as National Heritage Sites under the National Heritage Act, including neolithic rock paintings at G. Panjang Prk 25/Gua Tambun (Gua = cave). However, karsts to the south fall within the Batu Gajah economic zone, where leases for land use for quarries have been and continue to be issued.

In 2018, the Kinta Valley national geopark was declared, covering 1,952 km2 and 18 sites (Bunyan 2018). Ten of the geopark sites are karsts, selected as suitable for geo-tourism packages to attract both domestic and foreign visitors. The aim of the geoparks is to “advance protection and use of geological heritage in a sustainable way” and to “open opportunities for the state to add value to its existing tourist products”. The balance between exploitation of tourist potential and conservation is challenging because there is no legal framework for enforcement of geopark provisions. Geopark status does not have the same protection levels as state parks (Mohanakrishnan 2021). Further, several karsts were excluded from the Kinta Valley Geopark as they fall within the Mineral/Rock Basic Industry Economic Zone (Sharul-Hafiz 2021).

Important Plant Areas and conservation planning

Important Plant Areas (IPAs) are defined as the most important places globally for wild plant and fungal diversity that can be protected and managed as specific sites (Plantlife International 2004; Plantlife 2018). The designation of IPAs highlights sites that contain (1) important populations of threatened or highly range-restricted species; (2) exceptional botanical richness; and/or (3) important extents of threatened or range-restricted habitats. The IPA approach has been a successful tool widely utilised in Europe, North Africa and the Middle East where identified IPAs have resulted in recognition of new protected habitats or expansion of protected area networks (Darbyshire et al. 2017). With the clear need to conserve critical karst sites in the face of increased commercial exploitation, a combination of conservation responses encompassing sound scientific research, sustainable management and legal protection is required (Baillie et al. 2004), hence the identification of IPAs is an appropriate step towards enabling effective conservation decisions in this environment.

However, progress in commitment to conservation is lagging behind development. To date, applications for sustainable development planning, and accompanying Environment Impact Assessments (EIA), have emphasised commercial aspects at the expense of conservation measures. Malaysia’s National Policy on Biodiversity 2016 – 2025 (Ministry of Natural Resources and Environment 2016) nevertheless provides a framework for action through five goals. Goal 3 specifies safeguarding all key ecosystems, species and genetic diversity. Within this goal, Target 6 recognises the need to conserve at least 20% of terrestrial areas by 2025 through the establishment of protected areas, whilst Target 7 (directly relevant to identifying IPAs) calls for the protection of vulnerable ecosystems and habitats. Limestone habitat is specifically mentioned in Targets 6 & 7 and is also specified as an Environmentally Sensitive Area in the National Physical Plan 3 (Ministry of Housing and Local Government 2015) that recommends the protection of sensitive habitats and areas. In the Malaysian National Strategy of Plant Conservation (Saw et al. 2009), Target 5 recommends 10% of each ecological habitat be effectively conserved, with limestone areas again specified, while Target 6 specifies that 50% of the most important areas, including limestone areas, be totally protected through legislation. While limestone is singled out as a habitat of importance in all these conventions, legal protection has been slow. National Parks, State Parks and Wildlife Sanctuaries provide the strongest legal protection in Peninsular Malaysia. However, no karst in Perak has been gazetted for protection under these enactments. Although IPAs do not constitute a legal designation in Peninsular Malaysia, their identification is an effective tool that can be used in prioritising decision-making and to support and underpin legislation to conserve vulnerable and biodiverse habitats (Hamidah et al. 2020).

Assessing hills for conservation importance and balancing utilisation and conservation

Two principles can be considered when selecting limestone karsts for conservation (Kiew 1991a). The first, and the one that attracts most attention, is the presence of rare and threatened species. The second is that the karsts selected should be species rich, still pristine and harbour a good representation of species that make up the basic limestone flora.

The first step in the process of effective conservation planning is to close the knowledge gap for plants by documenting the species and habitats of a given area and their geographic distribution, conservation status and threats. These data together enable the identification of IPAs and the formulation of an effective action plan. In the case of Perak karst sites, a network of IPAs is needed to achieve maximum protection of the Perak limestone flora and habitats in light of the serious threats they face.

Biodiversity data required for conservation decisions and management strategies need to be available in a timely manner. Botanical field surveys are time consuming, expensive and require specialist knowledge that is in short supply in Malaysia. Herbarium holdings where specimens are correctly identified and databased are, therefore, the most reliable primary source of knowledge for rapidly obtaining this information, supplemented by appropriate literature and harnessing local and international expertise. Such data can help to inform future targeted fieldwork to infill knowledge gaps and to monitor management efforts. However, despite increased digitisation and data-sharing platforms that have greatly facilitated retrieving botanical data, there remain major challenges to collate and make readily available the botanical data needed for holistic conservation measures (Given 1994; Saw et al. 2009).

The aims of the current study are, therefore, to collate and harness existing data on the plant and habitat diversity of Perak’s karst environments in order to (a) identify which sites trigger IPA status; (b) amongst those sites, identify which are the highest priorities for conservation efforts and what is the minimum number of sites required to effectively protect the Perak karst flora, and so to enable effective conservation prioritisation; and (c) to consider the relationships between a range of parameters including collecting effort, site size and site isolation in order to better understand the distribution of Perak’s karst flora and where future research efforts may be best directed.

Methods

Sources of information

The baseline data for this study was derived primarily from herbarium holdings, taxonomic monographs and other relevant literature.

Geographic locations

Price (2014) and Liew et al. (2016) provide the basis for information on the location, size, geological and other features such as anthropogenic disturbance for each karst site, from which the checklist for karsts in Perak could be extracted (Appendix 1). All karsts were listed using the codes assigned by Price (2014) and Liew et al. (2016). Historic and herbarium label names, where they differ from the current standard name (Kiew 2013; Price 2014), are also included.

Herbarium holdings

Only specimens associated with limestone substrate in Perak were included in the collation of botanical data. Weedy species and dubious plant accessions were excluded. Specimens with duplicates bearing differently determined botanical names in different herbaria were re-identified and, for poorly understood species, checked with the expert for that particular family (see Acknowledgements).

Historically, the earliest collections from limestone in Perak date from 1880s (Kiew 2013). Most of the herbarium specimens are deposited in the Kepong Herbarium of Forest Research Institute Malaysia (KEP) and the Singapore Botanic Gardens Herbarium (SING) with early collections (pre-1900) being deposited in the herbarium of the Royal Botanic Gardens, Kew (K) (abbreviations follow Thiers, continuously updated). A plant checklist for Perak limestone (Appendix 2) was compiled from the BRAHMS (Botanical Research and Herbarium Management System) herbarium specimen databases of KEP and SING. Plant accessions associated with any limestone karst in Perak were extracted using karst names. Type and historic specimens at K were databased directly from specimen label data. For some historic specimens, this was not possible because of vaguely recorded localities, inconsistent use of locality names, and handwriting illegibility.

In order to obtain as complete a checklist as possible, specimens under ‘informal’ locality names that provided sufficient details to relate to a particular limestone karst were included. For example, the batch of specimens collected by I. H. Burkill dated 10 September 1920 only bears the name of a town in Perak, ‘Tambun limestone cliff’. These are included as plant records for Gua Datok Prk 36 because this location is reasonably identified based on several plant specimens, e.g. Gomphostemma crinitum and Homalomena humilis, that are restricted to shaded moist habitats in the foothills, a habitat available only on parts of Gua Datok Prk 36 and not on the smaller karsts in the Tambun area.

Botanical literature and online resources

A key source of data was the checklist of the limestone flora for Peninsular Malaysia by Chin (1977, 1979, 1983a, 1983b), but this only records distribution data for rare and endemic species. Since his account, many new species have been described as collecting continues. Kiew (1991b) included preliminary checklists for some individual karsts in Perak.

Plant records were also obtained from published literature, unpublished field survey reports by KEP botanists (i.e., Kiew et al. 2017) and the biodiversity data sharing platform, the Global Biodiversity Information Facility (GBIF Secretariat 2016). Taxonomic names are based on The Flora of Peninsular Malaysia (FPM 2010 – 2022), The Plant List (2017) and accounts by Turner (1997) and Ong et al. (2017). These sources, together with the Plants of the World Online portal (POWO 2022), were also consulted for the distribution and range of endemic taxa in Peninsular Malaysia.

Assessing extinction risk status

Assessment of the extinction risk (conservation) status of species follows the IUCN Red List Categories & Criteria version 3.1 (IUCN 2012). Assessments were obtained from FPM (2010 – 2021), the IUCN Red List of Threatened Species, and published botanical monographs (full list of references available on request from the authors). However, the majority of karst species (85.7%) has yet to be assessed (Yong et al. 2021). In this study, each species was provisionally assessed following the IUCN categories and criteria (IUCN 2012; IUCN Standards and Petitions Committee 2022) based on the collated data. Generally, the data available were most applicable to thresholds defined under Criterion B: geographic range, in the form of B1 (extent of occurrence — EOO) and/or B2 (area of occupancy — AOO) and its conditions (e.g. number of locations and population decline or decline in quality and extent of habitat). These are regional conservation assessments as they apply only to their occurrence in Malaysia, except for endemic species for which the conservation status is a global assessment (Chua 2010).

The distribution of each species in Peninsular Malaysia was geo-referenced and uploaded to the Geospatial Conservation Assessment Tool (GeoCAT) (Bachman et al. 2011) to estimate (EOO). For AOO, three approaches were adopted,: (1) for karsts less than 4 km2, AOO was taken to be the size of the karst (Liew et al. 2016), because about 90% of karsts in Peninsular Malaysia are smaller than 1 km2 (Liew et al. 2016) and the foothills have been cleared of surrounding forest (Chin 1977; Kiew 1991b; Kiew & Rafidah 2021); (2) for karsts larger than 4 km2, AOO was taken as 4 km2, because many limestone species are typically restricted to particular microhabitats that are not available throughout a limestone massif (Kiew et al. 2014; Kiew & Rafidah 2021); (3) for forest types other than limestone forest, a 2 × 2 km grid cell was used, which is the recommended size for most situations (IUCN Standards and Petitions Committee 2022). Forest cover was observed using Google Earth Pro (Google Earth Pro 2016) and with reference to information published in the annual report of Forestry Department Peninsular Malaysia (2016) and from field surveys.

Species with 15 or more occurrences in five or more states of Peninsular Malaysia, often with EOO > 20,000 km2, likely do not meet the thresholds of threatened categories or Near Threatened (NT). Therefore, all species with this distribution range were designated as Least Concern (LC) without geo-referencing. Species already recorded as ‘widespread’ or ‘throughout the country’ in Turner (1997) and monographs (e.g., Seidenfaden & Wood 1992; Holttum 1968) were also assigned as LC. The Data Deficient (DD) category was assigned to taxonomically uncertain species or those that lacked accurate locality data. Plant species that were not assessed were assigned as Not Evaluated (NE). Specialists for a particular family (see Acknowledgements) were consulted for provisional conservation status based on personal knowledge and reviewing the preliminary extinction risk status. Of the 340 species provisionally assessed in this study (Appendix 2), 289 species were assessed based on literature and the herbarium specimen dataset (cited as ‘provisional’); and 51 species were assessed by experts (cited as ‘pers. comm.’).

Data analysis

Important Plant Areas Assessment

The revised global Important Plant Area (IPA) criteria (Darbyshire et al. 2017) were applied to assess whether each of the Perak karsts triggers IPA status, in order to address aim (a) of this study. The gathered information of each karst, such as recorded species with threatened extinction risk status, adaptation to a specific habitat and endemism are factors directly related to the thresholds under the IPA criteria.

Conservation Priority Scores (CPS) to identify IPAs

In order to address aim (b) of the study, a new priority scoring scheme (Table 1) was developed for this study, to provide a quantitative basis for effectively highlighting sites of exceptional botanical importance from the many karsts that qualify as IPAs (Radford et al. 2011). The formulated scheme quantifies categories of threat and levels of endemism for every species recorded on each karst. The accumulated total of the weighted score of each species recorded at each locality provides a Conservation Priority Score (CPS) which can be used to highlight karsts of highest biodiversity value and so priorities for protection.

Table 1. Conservation priority scores for different categories of threat and levels of endemism.
Full size table

To ascertain if CPS is an optimal indicator for IPAs, we first collated botanical data for each karst (i.e., CPS, species richness, endemism and number of threatened species), karst parameters (i.e., size in km2 and degree of isolation in 10 km radius) and collection effort (i.e., number of specimens). Prior to analysis, all data were log transformed and scaled to allow their effects to be compared. We then performed a multiple linear regression model analysis, with CPS for 36 hills as the dependent variable and number of threatened species and endemics for each hill as independent variables. The same analyses were also performed using other parameters as the dependent variable, i.e., species richness, number of specimens, and karst parameters by the same set of independent variables. The following model equation was applied:

$${\text{log}}\left(\mathrm{dependent\; variable}\right)={\beta }_{0}+{\beta }_{1}\times {\text{log}}\left({\text{threatened}}\right)+{\beta }_{2}\times {\text{log}}({\text{endemic}})$$

We selected models with best adjusted R-squared (R2), standard error (SE) and Akaike Information Criterion (AIC). R2 is a continuous measure that indicates degree to which a model fits with the data, where values closer to 1 indicate better fit and closer to 0 indicate poorer fit. The AIC method was performed to further confirm best-fit regression models of parameters against both independent variables (threatened and endemic species).

Drivers of CPS

In order to address aim (c) of the study, relationships between all of the karst parameters were investigated. All variables were fitted in single linear regression models for analysis, with CPS as the dependent variable, while the independent variables included species richness, number of specimens, karst size, and degree of isolation. The adjusted R2, SE and AIC of all models were compared to determine how these drivers contribute to CPS. The following model equation was applied:

$${\text{log}}\left({\text{CPS}}\right)={\beta }_{0}+{\beta }_{1}\times {\text{log}}(\mathrm{independent\; variable})$$

All of the above analyses were performed using R version 4.0.5 (R Core Team 2021).

Results and Discussion

Physical details of the 95 karsts included in this study are listed in Appendix 1. For the majority of sites (59 karsts), botanical information was found to be completely lacking. Data on species richness, endemism, and conservation status for the 36 karsts that have been botanically surveyed are provided in Appendix 3.

Species richness

From herbarium holdings and the literature, from the earliest collections dating from the 1880s (Kiew 2013) until today, 54 plant collectors have accumulated approximately 1,200 collections from the karsts of Perak. From these, a total of 538 species (44.2% of the Peninsula’s limestone flora) from 110 families and 334 genera were recorded (Appendix 1).

The most speciose families on Perak limestone are Orchidaceae (75 species), Annonaceae (29 species), Gesneriaceae and Araceae (21 species each), Acanthaceae, Phyllanthaceae and Rubiaceae (18 species each), Euphorbiaceae (17 species), Apocynaceae and Moraceae (16 species each), Arecaceae (12 species), and Lamiaceae, Meliaceae and Urticaceae (11 species each). The most speciose genera are Ficus (13 species), Selaginella and Diospyros (8 species each), and Asplenium, Cleistanthus, Dendrobium and Paraboea (7 species each).

The eight best-collected karsts in Perak, i.e., those with more than 40 species recorded (Appendix 3), are G. Kanthan Prk 47 (191 species), G. Tempurung Prk 01 (114 species), G. Rapat Prk 23 (101 species), Gua Datok Prk 36 and G. Kuang Prk 46 (each with 70 species), Pulau Batu Putih mykarst-172 (66 species), G. Pondok Prk 55 (65 species) and G. Mesah Prk 06 (45 species). Each individually harbours fewer than 36% of the total Perak limestone flora (Fig. S1).

These totals compare rather poorly with the most extensively surveyed karsts in Peninsular Malaysia, namely Batu Caves, Selangor, from where 360 species have been collected (Kiew et al. 2023) and Gua Musang, Kelantan, with 223 species (Aliaa-Athirah et al. 2019). However, rather than reflecting true differences in species richness, this may indicate that the Perak limestone flora is poorly documented relative to these Selangor and Kelantan sites.

Aliaa-Athirah et al. (2019) have demonstrated the importance of collecting effort, with species totals increasing with more intensive collecting as the karsts are more thoroughly explored. Batu Caves is a test case in that it has been continuously collected by a range of specialists over more than a hundred years and was the focus of a year-long expedition in 2019. The current total of 360 species (Kiew et al. 2023) represents only 28% of the total limestone flora of Peninsular Malaysia, indicating that no single karst will harbour more than a fraction of the limestone flora. On the other hand, under-collecting presents an impediment to identifying IPAs when such a high number of karsts in Perak are under-collected or even, for the majority, not at all.

Endemism

Of the 538 species on Perak limestone, 71 (13.2%) are endemic of which 10 (1.9%; Table 2) are restricted to a single karst site, 23 (4.3%) are endemic in Perak, and 38 (7.1%) are endemic in Malaysia.

Table 2. Critically Endangered and narrowly endemic species on Perak limestone karsts
Full size table

The high level of endemism is also reflected in other groups of organisms that include rare and hyper-endemic species, such as land snails (Foon et al. 2017); bent-toed geckos (Grismer et al. 2016) and trapdoor spiders (Whitten et al. 2013).

Threatened species

Among the Perak limestone flora, 526 species (97.8%) have been assessed for their regional or (for endemic species) global extinction risk. The remaining 12 species were not assessed because of insufficient data. However, only 185 species have a published status. The flora on karsts has a high proportion of threatened species; of the 538 species, 108 (20.1%) are threatened, of which 21 are Critically Endangered (CR), 47 are Endangered (EN), and 40 are Vulnerable (VU) (Table 3, Fig. S2).

Table 3. Threatened species on limestone karsts in Perak. CR = Critically Endangered; DD = Data Deficient; EN = Endangered; EX = Extinct; LC = Least Concern; NE = Not Evaluated; NT = Near Threatened; VU = Vulnerable.
Full size table

Among the Perak threatened species, concern is highest for the 21 CR species, of which 10 are hyper-endemics and 7 endemic to particular clusters of karsts in Perak (Table 2). These species are most threatened because a single anthropogenic disturbance event may lead to extinction. Information on narrow endemic species on isolated karsts is often not readily available, being embedded in scientific publications, which has led to them being overlooked in site-level surveys and EIAs for development purposes and to stakeholders involved. For this reason, a list of endemic species known from fewer than five karsts was published by Kiew et al. (2017).

The extreme vulnerability of CR species is demonstrated by Pseuduvaria glossopetala and Vatica kanthanensis that are both hyper-endemics on the actively quarried limestone massifs of G. Pondok Prk 55 and G. Kanthan Prk 47 respectively. At present, the continued existence of P. glossopetala on G. Pondok Prk 55 is doubtful as its most recently recorded occurrence was in 1971 (Su et al. 2010) and since then the summit and northern part have been quarried away. Despite G. Kanthan Prk 47 being extensively explored, V. kanthanensis is known only from a population on the ridge of the hill (Tan et al. 2014).

Extinct species

Six species were assigned as extinct or probably extinct (Table 4). The main inferences for their extinction are extensive degradation of their only recorded locality and/or not having been recollected for more than a hundred years. Among species considered as ‘probably extinct’ are Ardisia meziana, Cleistanthus parvifolius and Piper collinum that have not been re-found since they were first documented by Kunstler, who collected in the south of Perak between 1880 and 1886 (Kiew 2013). The area from where they were collected has since suffered widespread forest clearing and habitat degradation by tin-mining and agriculture.

Table 4. Extinct and possibly extinct species in the Perak limestone flora and the locality from where they were known. 1 Locality as written on the only available specimen.
Full size table

In addition, there are two foothill-associated species that have been collected only once over the last hundred years, namely Justicia subalternans and Strobilanthes pachyphyllus from the base of G. Tempurung Prk 01 and G. Mesah Prk 06 respectively, although the upper parts of these karsts are still covered by undisturbed vegetation. A thorough survey of the foothill areas of these two karsts is required before these species can be assessed as extinct.

Assessing the importance of species based on CPS

Two thirds (360 species, 67%) of the Perak limestone flora is widespread and most of these species are not endemic. These species together contribute only 21 of the total CPS. Conversely, there are 108 (20.1%) threatened species recorded, which account for nearly 80% of the CPS for the Perak karst flora (Fig. 3).

Fig. 3.
figure 3

Density scatter plot to visualise cumulative CPS of individual species based on extinction risk status and endemism. Area of deeper colour tone represents a higher proportion of accumulated CPS.

Full size image

Species with the highest CPS are those that are Critically Endangered (CR) and endemic in one locality (individual CPS score 12) followed by CR species endemic in Perak (CPS 11) (Table 1). These hyper-endemic species and species endemic in Perak are of greatest conservation concern for Perak as they are most in danger of extinction due to the prevalence of quarries and lack of legally enforceable protected status. In addition, many have extremely small and local populations, such as Monophyllaea elongata which is known from one cave entrance that is a popular tourist site. A single disturbance event can lead to the extinction of such species. Although there are only 21 CR species, the cumulative impact of the CR species is notable in the CPS, yielding the second highest score of 208 amongst the Perak karst flora.

Endangered (EN) species are assigned with a lower CPS than CR species, i.e., a score of 10 for endemic and 5 for not endemic respectively. However, the cumulative CPS for all EN species yields the highest CPS score of 350 and accounts for 39% of the total CPS for the Perak karst flora, because the number of EN species on Perak karst is double that of CR species. Many of these EN species are endemics and limestone-obligate species, such as Corybas calcicola and Stichorkis calcicola.

By including endemism level of individual species to quantify conservation importance, some species with a lower threat status but that are highly range-restricted are given a high CPS, equable to that of a threatened but non-endemic species. For instance, Scaphochlamys kunstleri var. kunstleri, a Near Threatened (NT) endemic Perak species, has a CPS 5 that is the same as a non-endemic EN species. The CPS thus highlights endemism in its own right, irrespective of threat status, as worthy of conservation attention as part of a more comprehensive conservation strategy.

Benefits of CPS for conservation planning

The use of CPS can improve the prioritisation of karsts harbouring high proportions of restricted range species. By modelling the relationships between different parameters, we confirmed that threatened and endemic species composition of limestone hills in Perak shapes CPS (Fig. 4) more than species richness, number of specimens, karst size and karst isolation. CPS modelled as a function of both threatened and endemic species accounts for high variability (R2 = 0.9346), has high precision (Standard error = 0.2634) and model fit (AIC = 11).

Fig. 4.
figure 4

Relationship of CPS with a number of threatened species, and b number of endemic species, for Perak karst sites. Number of species per karst is represented by relative point size.

Full size image

Models for other parameters were found to have lower R2 values (0.8398 for species richness, 0.8534 for number of specimens, 0.2523 for karst size and 0.1851 for karst isolation); higher standard errors (0.4122, 0.3943, 0.8905, 0.9297 for the four parameters respectively); and higher AIC values (43.2, 40, 98.7 and 102 respectively). This demonstrates that CPS helps identify high conservation value karst sites to ensure maximum coverage of the range restricted limestone flora.

Significance of parameters on the distribution of species of conservation importance

  1. (a)

    Species richness and collecting effort

CPS was found to be significantly positively correlated with number of species (rs = 0.8949, p < 0.001), and number of specimens (rs = 0.8941, p < 0.001) at a site. Single regression analysis showed number of species and number of specimens have similar values in R2 (0.8039 vs 0.8108), SE (0.4482 vs 0.435), and AIC (47.5 vs 46.2), suggesting both are important variables to make realistic CPS predictions. However, misinformation might occur if karsts are designated for protection based mainly on species richness or specimen number, as botanical surveys of many hills have been limited and non-exhaustive and some karsts are dominated by widespread species. Taking the example of two karsts of c. 0.3 km2, G. Mesah Prk 06 has a total CPS of 84 which is higher than G. Kuang Prk 46 with a CPS of 63 despite the latter having more species reported, i.e., 70 species compared to 45 species. G. Mesah Prk 06, a site that has been repeatedly surveyed botanically for over a decade, has a higher CPS because of the higher concentration of restricted range species despite the lower recorded plant diversity.

  1. (b)

    Karst size and isolation

For land snails, Clements et al. (2008) and Liew et al. (2021) suggested that karst size and isolation are the main determinants for the occurrence of narrowly endemic species, both with a positive correlation, and that therefore larger and/or more isolated karsts would have higher conservation value. The inventories of Foon et al. (2017) for land snails, however, showed a correlation between karst size and number of hyper-endemic species, but not with species richness. Further, they found that degree of isolation of individual karsts was not statistically correlated either with presence of hyper-endemic species or with species richness.

For plants, results from our study show that CPS has a weak correlation with karst size (rs= 0.3113, p = 0.04025) , with very low R2 (0.092), and very high standard error (0.9529) and AIC value (103). However, this is possibly due in part to variable collection effort of Perak limestone karsts. For instance, G. Kanthan Prk 47, one of the smaller karsts covering only c. 0.8 km2, has the highest CPS of 248, with 191 species recorded. Conversely, the largest hill, G. Tempurung Prk 01 (8.7 km2), has a lower CPS of 244 and only 114 species recorded. Both karsts have been popular collection sites for over a century, but G. Kanthan Prk 47 was repeatedly surveyed by the same botanist team from FRIM with different fields of expertise during different seasons from 2013 to 2014, and in 2017. Utteridge & Edwards (2009) suggest that the same collection team repeatedly surveying a given site will result in a wider range of plants being recorded as they become more knowledgeable about the site’s flora. In contrast, G. Tempurung Prk 01 was only occasionally surveyed and a large portion of the karst has not been botanised due to inaccessibility.

Kiew (1991b), Kiew & Rafidah (2021) and Fitzsimons & Michael (2017) have demonstrated the importance of diversity of microhabitats and niches as the main determinant of plant diversity. While it might be inferred that larger karsts would have more microhabitats and niches, this is not necessarily always the case, as it is highly dissected karsts that support the most microhabitats and hence harbour higher species richness.

Regarding karst isolation, this does not correlate with CPS in our study (rs = -0.2779, p = 0.1949), with very low R2 (0.209), and very high standard error (0.9895) and AIC value (105). This supports the findings of Foon et al. (2017) for land snails and indicates that isolation is not a reliable indicator of conservation importance. However, these results should be regarded as tentative, because many of the most isolated karsts have either not been surveyed at all or very few collections exist for them. Additionally, there are a notable number of limestone obligate species known from a single karst and to a specific habitat niche and confined area. For example, Grewia huluperakensis is a very rare scandent shrub only found at an isolated limestone island near Temengor Dam. Through prolonged geographic isolation, speciation occurs due to poor interbreeding and change in characteristics for adaptation to survive and reproduce (Clements et al. 2006) hence we might expect that isolated karsts would have more hyper-endemic species.

Network of priority IPA sites for Perak karst

In Peninsular Malaysia, adjacent karsts or those in close proximity cannot be assumed to harbour the same assemblage of species and the same endangered species, due to differences in the topography and the availability of microhabitats (Kiew et al. 2019). This has implications for conservation management because quarrying companies often suggest that conserving a nearby karst will conserve the equivalent biodiversity as the karst to be quarried.

The limestone flora varies between different karsts, each supporting different combinations of species so a single karst typically supports at most just 35.5% of the total species richness (Fig. S1). To safeguard maximum plant diversity, a network of protected karsts is therefore required. Of the 36 sites for which we have botanical data, 27 potentially trigger IPA status under criteria A(i,ii,iii,iv), B(ii) and/or C(iii), i.e. they harbour important populations of threatened species, outstanding botanical richness and/or are important examples of threatened habitats that in Perak include the foothills and limestone forest. Given that most of the karsts for which we have data trigger IPA status (27 out of 36), the CPS can help to further prioritise these sites for conservation planning. 17 karsts are here prioritised as IPAs based on the CPS and species richness (Table 5; Figs 5 & S1). Note that the table is based on karsts for which data are available. For a further 19, data are limited and 59 karsts have yet to be explored. These hills are therefore excluded from the selection, although they are likely to be of conservation value.

Table 5. Perak limestone karsts listed for conservation priority as IPAs based on conservation priority scores and species richness
Full size table
Fig. 5.
figure 5

Map of 17 karsts prioritised as IPAs based on species richness and conservation priority scores.

Full size image

The 15 karsts ranked highest by the CPS support populations of all the CR species, hyper-endemics and 17 of the 18 species endemic to karst in Perak. These include G. Ginting Prk 26 which, although with a lower total species richness recorded, is highlighted as important using the CPS as it harbours a significant number of species of conservation importance not known from other karsts. These include three range restricted species, one of which is CR (Iguanura bicornis), one that is EN and endemic to Perak (Tarenna sp. 10), and another that is endemic to Perak (Calamus setulosus).

In contrast, the top 15 karsts ranked by number of species (species richness), although having a more diverse flora than the top 15 using the CPS, support fewer of the CR species and Perak endemics, and only support an additional two EN species (Cleistanthus praetermissus on Tunggal Prk 02 and Impatiens macrosepala on Gua Badak Prk 61) and two VU species (Trichoglottis bipunctata and Momordica clarkeana), when compared to karsts selected using the CPS (Table 5). In theory, Tunggal Prk 02 and Gua Badak Prk 61 should be considered within the karst IPAs network in Perak to ensure maximum representation of threatened species is safeguarded (Figs S1 & S3). Tunggal Prk 02 is, however, currently being quarried and its population of C. praetermissus has quite probably been extirpated, hence this latter site no longer qualifies as an IPA. Gua Badak Prk 61 (Lenggong) is protected as a UNESCO Archaeological Site and, whilst not managed for biodiversity conservation, it is nevertheless viable as an IPA. In addition, Hill KF Prk 53, which does not feature in Table 5, should be considered as an IPA as it harbours the CR and Perak endemic herb, Gymnostachyum kanthanense, for which the other two of its recorded localities are threatened by active quarrying or agricultural activities. Three other threatened species are also recorded at this latter site, i.e., Paraboea capitata var. capitata (EN), P. paniculata (EN) and Fagraea curtisii var. curtisii (VU).

Therefore, the 15 karsts with the highest CPS, together with Gua Badak Prk 61 and Hill KF Prk 53 (Fig. 5), are recommended as a network of 17 IPAs that will together conserve 90% (483 species) of the Perak karst flora including 94.4% (102 species) of its threatened species and 93% (66 species) of the endemic species, as well as including all those that are CR (Table 2) and almost all the EN species.

Among the 17 selected IPAs, G. Kanthan Prk 47 has the highest CPS (248), holding approximately a quarter (26.7%) of Perak’s limestone endemics (19 species) and a quarter of the threatened species (28 species). G. Kanthan Prk 47 also has the highest total number of species, with 191 species representing 35.5% of Perak’s limestone flora (Kiew et al. 2014).

Seven limestone karsts have a CPS greater than 70 (Table 6). These sites harbour the highest species richness, and each has one to three hyper- or Perak endemic species (Table 2), but none has received any protection. Several of the IPA sites, including Gunung Kanthan Prk 47, G. Rapat Prk 23, G. Pondok Prk 55 and G. Datok Prk 36, are currently under threat from on-going quarrying.

Table 6. Summary of endemism, conservation status and conservation priority scores for the assessed limestone karsts in Perak
Full size table

The CPS for plants can also be an important indicator of the conservation value of karst sites for fauna. For example, Davison (1991) inferred that Perak’s land snail species represent about one-third of those known from Peninsular Malaysia. Foon et al. (2017) inventoried 12 karst sites in Perak and found a total of 122 land snail species with 63 species in Perak found on a single karst, of which 34 species are hyper-endemic being known from a single karst in Malaysia. Four karsts, G. Tempurung Prk 01, G. Pondok Prk 55, G. Kanthan Prk 47 and Bt Kepala Gajah Prk 64, were pinpointed as in need of conservation attention as these support 91% of mollusc diversity in Perak and 27 of the 34 species unique to a single karst. Hence three of the top five sites and four of the top ten sites of importance for plants using the CPS method correspond to the most important karst sites for land snails. The high conservation value of G. Kanthan Prk 47 is also supported by other faunal groups, including the presence of the Vulnerable Sumatran serow (Capricornis sumatraensis, Phan et al. 2020), hyper-endemic bent-toed gecko (Cyrtodactylus guakanthanensis, Grismer et al. 2014), and the Critically Endangered and hyper-endemic trapdoor spider (Liphistius kanthan, Whitten et al. 2013).

Conclusions and next steps

Balancing commercial enterprise with effective environmental management and biodiversity protection is one of the great challenges facing the world today. Striking this balance is particularly urgent in fragile and highly restricted environments such as the karsts of Peninsular Malaysia, where damage is often irreversible. It is not practically possible to protect every karst due to development demands, hence target-orientated conservation measures are necessary to ensure long-term survival of the limestone flora.

Karsts in Perak harbour impressive plant diversity with 538 species representing 44.2% of Peninsular Malaysia’s limestone flora, including 108 globally or regionally threatened species. However, as with all other studies of the limestone flora in Peninsular Malaysia, individual karst harbour only a subset of this plant diversity, such that in order to safeguard the maximum diversity of Perak limestone plants including the threatened and endemic species, a network of protected karsts is required. The identification of a network of IPAs, supported by quantitative conservation priority scoring, effectively highlights the highest priority karst sites for legal protection across the state that can form the basis for a conservation strategy that will capture maximum plant species richness, endemism and threatened species. Critically, this study demonstrates that the CPS method is more effective in identifying high conservation value karsts compared with methods based on karst size, isolation or species richness. If effectively managed and protected, the 17 karst IPAs identified here will conserve 90% (483 species) of Perak’s limestone plant diversity including 93% (66 species) of the state’s limestone endemic species and 94.4% (102 species) of its threatened species.

Although the Perak limestone is well-documented botanically by Malaysian standards, many Perak karsts still need surveying, and those identified as highest conservation priority require continued monitoring. Field surveys are required to update many of the now 30+ year old surveys to evaluate the effect of on-going encroachment. Intensive field surveys are necessary to identify populations of threatened species, especially those uncollected for many years. If appropriate, ex situ conservation can be carried out, as was done for the Pahang hyper-endemic Paraboea bakeri (Gesneriaceae) which is now extinct in the wild due to quarrying (Tan 2014).

With increasing demand for limestone products, and the resultant increase in number and size of quarries, the need to botanically explore karsts for which no data exists becomes more urgent. These karsts listed as a priority for scientific surveys should facilitate better orientated and comprehensive limestone plant diversity inventories.

Results of this study serve as a comprehensive and accessible information source to facilitate all stakeholders in these karst environments to achieve sustainable development. Dissemination of these data to key stakeholders, particularly to state government planning departments, can facilitate effective planning and can provide a model that can be applied in the other Malaysian states with limestone karsts.