The Karstlands of Antigua, Their Land Use and Conservation

By Day, Michael

The northeastern 110 km^sup 2^, or nearly 40%, of Antigua is underlain by impure limestones of the Oligocene-aged Antigua Formation, on which has developed a subdued karst landscape consisting essentially of shallow enclosed depressions (dolines), intermittently active stream valleys and widely scattered residual hills. The dispersed dolines are broad, shallow and clustered, especially in the central and southeastern sections of the limestone belt, where they attain densities of 7/km^sup 2^. The widely spaced residual hills attain heights of up to 40 m and localized densities of over 4/km^sup 2^. Five main valley systems up to 6 km in length traverse the limestone in a broadly northeast direction, carrying both autogenic drainage from within the karst area and allogenic drainage from the non-carbonate Central Plain. Karst and cave development has been constrained by the low purity of the limestones. Of the four types of carbonate islands identified within the Carbonate Island Karst Model, Antigua most closely resembles the Composite Island type. The karst has been much influenced by human activities, particularly agriculture and quarrying, and is now a focus of the burgeoning tourism industry. Virtually none of the karst is designated as protected areas, but several sites warrant protection, and several conservation strategies have been suggested. KEYWORDS: Antigua, karst, geomorphology, human impact, resources, management

Introduction

Although the Caribbean is one of the World’s premier karst landscapes, with a total karst area of nearly 130 000 km^sup 2^ (Day 1978a; Kueny and Day 1998), the karst of the smaller islands, particularly those in the Lesser Antilles, is relatively little known. These karstlands are devel- oped in the outer band of the Lesser Antilles islandarc, otherwise known as the Limestone Caribbees, and incorporate Anguilla, Antigua, Barbuda, Grande Terre Guadeloupe and nearby Marie Galante and Desirade. Perhaps the best known Lesser Antilles karst is that on Grande Terre, Guadeloupe, where the dry valleys were the object of a classic study by Lasserre (1954). Further south, the karst of Barbados has been studied in some detail (Fermor 1972; Day 1983) and that of Trinidad and Tobago has been the subject of preliminary studies (Day and Chenoweth 2004). The karst of Antigua is an important element of that island’s landscape, but has been the subject of only cursory previous research (Day 1978a 1986).

Small carbonate island karst has been an important focus of recent studies in karst geomorphology, and it has been shown to differ intrinsically from that on larger islands (Mylroie and Vacher 1999; Mylroie and Jenson 2002; Vacher and Mylroie 2002; Mylroie 2004). Such distinctive island karst develops where young carbonate rocks on small islands or coasts have not been buried beyond the range of meteoric diagenesis and interact with fresh and saline groundwater that is affected by glacio-eustatic and tectonic changes in sea level. This diagenetically immature karst is termed eogenetic and its hydrogeology is defined by the Carbonate Island Karst Model (CIKM) (Mylroie and Vacher 1999; Vacher and Mylroie 2002; Mylroie 2004). Four types of carbonate islands (simple, carbonate cover, composite and complex) are recognized within the CIKM on the basis of a spectrum of carbonate/non-carbonate complexity. Much of the Lesser Antilles karst is adjacent to volcanic islands of the inner island are and overlies older volcanic or crystalline basement rocks, thus providing a number of potential test cases of the universality of the CIKM.

Geology, soils and climate

The karst in Antigua is developed primarily on the Antigua Formation, which is composed mostly of limestones and marls with beds of calcareous mudstones and clays, plus small intrusions and localized lenses of siliceous reworked volcanic debris (Martin-Kaye 1959; Weiss 1994). Its age has been disputed, but it has been assigned to the Oligocene by Warneford (1949), Martin-Kaye (1959) and Weiss (1994). There are also limestones within the basal volcanic suite and in the Central Plain Group, but these are not extensive and are only locally karstified (Weiss 1994).

The stratigraphie succession on Antigua is the most complete in the Limestone Caribbees, illustrating declining episodic volcanism and intermittent but increasing marine transgression throughout the Oligocene (Weiss 1994). Rich coral communities formed as biostromes and patch reefs fringing or on a bank barrier in the lower part of the Antigua Formation, but the upper part is dominated by deeper water (30-100m), island slope deposits of mostly fine-grained limestones (Martin-Kaye 1959; Weiss 1994). The Formation attains a thickness of at least 550 m and dips regionally towards the northeast at an angle of less than 15[degrees] (Figure 1), although local exposures suggest that the dips are more depositional than structural.

Limestones within the Antigua Formation are variable, both in terms of purity and mechanical strength. Petrologically, some micrites and biosparites are present, but the limestones are dominantly biomicrit.es, often highly fossiliferous and containing small amounts of quartz, smectite clays and feldspars (Day 1978a; Weiss 1994). A series of insoluble residue determinations (Day 1978a) showed a range of carbonate contents from 62.0% to 96.2% (mean 82.0%), and a mean acid insoluble residue content of 1 8.1 % (s = 1 0.65, n = 1 0). Weiss (1994) characterized acid insoluble residues within the biomicrites as less than 10%, although noting that individual beds contained up to 40% of impurities. Acid insoluble residues in the reef limestones and the biosparites are generally low, but those in the marly limestones may approach 50% (Weiss 1994).

Surface compressive strength values, determined using a Schmidt Test Hammer (Day and Goudie 1977; Day 1980 1982), vary from 12 to 36 (mean 33.3) for fresh sections and from 27 to 41 (mean 37.8) for indurated surfaces. These variations in strength reflect both the conspicuous rhythmic sequences of island-slope limestone beds in the upper part of the Antigua Formation (Weiss 1994), and considerable variability in diagenetic cementation and surface case-hardening.

Soils are dominantly calcareous (up to 74% calcium carbonate), shallow (less than 0.3 m deep), extremely well-drained and stony grey-black rendzinas or mollisols. These are young, dark, calcimorphic soils, derived from weathering of the limestone and/or the build up of humus (Hill 1966; Watts 1987). Hill (1966) recognized two sub-types: (1) generally shallow calcareous clay soils in the north, which are productive in the deeper phases over the softer marls and, despite the high clay content, have good structure and high base saturation, with a pH of around 8.2; (2) a mix of shallow and deeper calcareous soils, mostly in the east, where the drier climate restricts productivity.

Natural woodland vegetation and the native fauna were decimated during the colonial period, and the limestone area is now characterized by a savanna-like grass and scrub vegetation with localized woodland (Beard 1948; Harris 1960; Loveless 1960; Caribbean Conservation Association (CCA) 1991; Lindsay and Horwith 1997). The dry scrub forests are dominated by Acacia spp. and Prosopis chilensis which were introduced and have become naturalized (FAO 2003). The karst area was used extensively for growing sugar cane and cotton until the second half of the twentieth century (Barker 1981; CCA 1991). Some areas of coastal mangrove persist within the karst, and ecosystem variety in the limestone area is enhanced by the presence of caves and by the seasonal drainage channels, streams, potholes and ponds (FAO 2003). Bats are the only native terrestrial mammal, with seven species documented (CCA 1991).

Antigua currently has a semi-arid tropical climate that is characterized by extreme rainfall uncertainty. Mean annual temperature in Antigua is 28[degrees]C and mean annual rainfall is 1110 mm (CCA 1991), with a wide range from 648 mm in 1930 to 1868 mm in 1899 (Martin-Kaye 1959). Rainfall totals generally increase with elevation, and most of the karst belt, which is the driest part of the country, receives less than 900 mm of rainfall annually. A marked dry season extends from January to April, with a wet season from about August to November. Evapotranspiration rates are relatively high, averaging from 87 mm/month in November to 143 mm/ month in March, and annual potential evapotranspiration exceeds effective precipitation (McMillan 1985), with an average annual effective precipitation to evapotranspiration ratio of about 0.57 (UNCCD 2005).

Meteorological drought, defined by OAS (2001a) as when yearly rainfall falls below 80% of the annual average, is a recurrent problem for Antigua. Such meteorological drought is in turn associated with agricultural drought, where extended dryness reduces moisture in the soil to the point that plants and crops are adversely affected, and hydrological drought, where the reduction of water levels in wells, surface reservoirs and dams severely affect livestock, or where water must be rationed to households, hotels and other sectors.

Drought occurs every 5-10 years (Martin-Kaye 1956; Lewis 1984; CCA 1991; FAO 2004), with perhaps the most notable examples in 1930, 1948, between 1964 and 1969 and in 1983-84. Drought risk is generally highest within the karst belt (OAS 2001a). Conversely, heavy rains and flooding may also be ephemeral problems, particularly during hurricanes such as Hugo in 1989 and Luis in 1995. Hurricane damage may also disrupt water supplies. Karst development in Antigua has occurred throughout a time of considerable paleoclimatic variability. Progressive global cooling culminated in Quaternary glaciations by about 1 .6 m per year and glacial- interglacial fluctuations affected the Caribbean, with glacial climates cooling by up to 5[degrees]C, with increased aridity, and sea levels declining by 100 m or more (Graham 2003). These glacial periods would generally have impeded karst development, although carbonates which are now below sea level would have been exposed to subaerial karstification. The impacts of contemporary convergent, compressional and collisional tectonic activity on the Antigua Formation limestones have not been investigated in detail, but they are likely to have overprinted the effects of glacioeustacy (Multer et al. 1986).

Karst landforms

Karst landforms are developed where the Antigua Formation crops out over about 110 km2 (nearly 40% of the total land area) in the northeast of the island (Figure 1). ‘Karst development affected the Antigua Formation deposits in a major way, but limestone volumes in the Basal Volcanic Suite and the Central Plain Group are not large enough to yield a karst topography’ (Weiss 1994, 26). Broadly, the karst area is an undulating limestone upland, which rises fairly steeply from the adjacent Central Plain to elevations over 100 m. There are no permanent streams or rivers. The dominant features are enclosed depressions or dolines, intermittently active valleys and residual hills (Day 1986), with a few small caves (Gurnee 1961) and natural bridges on Soldier Point and at Devils Bridge, the latter now a national park. The karst is essentially Neogene in age, although some features may be related to the initial Oligocene carbonate deposition. The conspicuous rhythmic sequences of island- slope limestone beds in the upper part of the Antigua Formation give rise to characteristic ledge-and-notch’ profiles on some slopes, and locally there are small channels formed by turbidity currents and slumped masses perhaps produced by earthquake disturbance of semi- coherent carbonate sediments (Weiss 1994). Although there has been no definitive work on the subject, it is also possible that at least some of the larger enclosed depressions may be tectonic or constructional features produced by variability in the carbonate depositional environment, and representing fossil lagoons or swales between dunes (Mylroie 2004), rather than being dissolutional in origin. Karst development was also doubtless facilitated by fluctuating sea levels during the Quaternary, as described by the CIKM (Mylroie and Vacher 1999; Vacher and Mylroie 2002; Mylroie 2004), and contemporary tectonic activity, although not specifically studied in detail, doubtless overprinted the effects of glacioeustacy (Multer ef al. 1986).

The karst area is delimited by a discontinuous southwest-facing escarpment, in places 60-80 m high, which trends approximately northwest to southeast from Wetherills Point to Willoughby Bay (Figure 1). The escarpment is breached in four locations by dry and seasonally active alluviated valleys which carry allogenic drainage from the Central Plain across the limestone towards the northeast (Figure 1 and Table 1). The elevation of the Central Plain is generally less than 30 m, and the valleys traverse the karst from about this elevation to sea level, with higher karst blocks between them. Surface flow in the valleys is intermittent or ephemeral and, although alluvium appears to render many of the valley floors impermeable, flow appears in places to be lost into the streambeds, resulting in wet and dry reaches. No detailed studies of these through-valleys have been undertaken, nor has their hydrology been investigated thoroughly, although Martin-Kaye (1959) suggested that part of the Central Plain drainage was superimposed. Drainage densities (Table 1) are at least one or two orders of magnitude lower than those characteristic of tropical areas in general (Gregory 2000), reflecting available moisture and the importance of infiltration into the limestone.

Within the karst blocks there are numerous short dry valleys, some of which terminate in small sinks. Of these, only Carr’s Gut (or ghaut) is prominent (Figure 1 and Table 1). Although no detailed studies of these autogenic valleys have been undertaken, they are poorly organized, with only a suggestion of a dendritic pattern. Otherwise, they resemble both morphologically and probably functionally the comparable dry valleys in Guadeloupe (Lasserre 1954), Barbados (Fermor 1972; Day 1983) and Jamaica (Day 1985), and their ephemeral floodwaters represent an occasional hazard to local residents (Day 1978b).

The Antigua dolines warrant attention because of their potential surface collapse hazard (Day 1978b), and because their origins and development have not been established. Given the eogenetic and primarily autogenic groundwater recharge environment, it is possible that at least some of the closed depressions are actually depositional in origin, and some of the carbonate stratigraphy described by Weiss (1994) supports this possibility. In contrast, there is little evidence of large-scale active or former doline collapse, which is common in other carbonate and composite islands (Mylroie and Vacher 1999), nor are there deep water-filled shafts or blue holes, which also characterize such settings.

Doline numbers and densities are very low by comparison with other karst areas in the Caribbean (Day 1978a). Dolines are few in the northeast, but are more numerous in the central and southeast parts of the karst area, with the most striking examples occurring in the vicinity of Freetown (Figure 1). In the karst area of 110 km^sup 2^, 45 dolines have been located and surveyed in the field (Day 1978a); the morphometric properties of these dolines are summarized in Table 2. Overall doline density is low at 0.39/km^sup 2^, but the distribution, as measured by the single nearest neighbour statistic (Williams 1972), is clustered, so local densities are greater, reaching 7/km^sup 2^ in the southeast (Figure 1), where karst development is promoted by a locally steeper hydraulic gradient and where there is little competition from dry valley systems (see Fermor 1972; Day 1983). The dolines are shallow, generally less than 2 m deep, although some near Freetown exceed 3 m in depth. Doline diameters are much greater than their depths, and so the depth/diameter ratios are very low. Doline area per km2 is very low and the index of pitting (the inverse of the percentage of the total area occupied by dolines) is correspondingly high. Only two (4.5%) of the dolines occur in a complex, that is where two or more individuals make up a larger, compound doline. This is lower than in similar terrain in Jamaica, but similar to the situation in Barbados and the Yucatan (Day 1978a 1983).

Planimetric shapes are quite variable, with no clear overall pattern. Smaller dolines tend towards circularity, but many larger individuals are notably elongated in one direction. Long axis orientations are concentrated in three directional classes: 31- 40[degrees] (approximately southwest-northeast), 81-90[degrees] (approximately east-west) and 351-360[degrees] (near north-south) (Day 1978a 1986). This suggests a broad structural control, as regional dip is towards the northeast and the three dominant jointing directions in the Antigua Formation are in the segments 10- 40[degrees], 50-80[degrees] and 280-300[degrees] (Martin-Kaye 1959). However, the correspondence between doline long axes and jointing directions is poor, and the preferred long-axis orientations are not statistically significant as determined by Tanner’s (1955) test.

In terms of overall morphology and landscape position, most of the smaller dolines, those less than 20 m in width and 2 m in depth (Plate 1), appear to be the result of near surface dissolution, rather than having a collapse or depositional origin. Low doline numbers, limited depths and low density reflect the inhibition of point-focused near-surface dissolution by diffuse input and limited recharge areas. Near-surface flow concentration may be associated with more insoluble and less permeable units within the limestone, and the clustering of dolines supports the possibility. However, further investigation is necessary to clarify the situation. In contrast, some of the larger depressions, particularly those exceeding 100 m in diameter, may have a constructional, depositional origin, although detailed investigation of this possibility also remains to be undertaken. The variability of planimetrie shapes and the poor correspondence between doline long axes and jointing directions also point to variations in the depositional environment, rather than long-term dissolutional development.

Several dolines in coastal locations have been breached by marine erosion, exposing their poorly cemented sediment infill. Particularly good examples of this are evident at Soldier Point and on Great Bird, Green, Smith and York Islands.

The other karst landforms are isolated, possibly residual hills (Plate 2), which Martin-Kaye (1959, 1) described succinctly as ‘scattered, rather abrupt hills’. Day (1978a 1979) grouped this landscape on general morphometric grounds with tower karst, but there are clear morphological and functional differences, with the closest analogues being the conical residual hills of Grande Terre, Guadeloupe (Lasserre 1954) and the broader residual limestone hills in Trinidad (Day and Chenoweth 2004).

Most of the hills rise from an undulating, or near-flat, surrounding limestone surface, but some are surrounded by Plio- Pleistocene clays (Martin-Kaye 1959) or by recent alluvial deposits. Typically, these surface deposits occur only peripherally, and do not separate individual hills. Locally, the clays produce surface runoff during heavy rain, and this may abut against the hill bases, but no basal depressions, channels or sinks are evident, which contrasts with the situation in tower karst areas such as in Puerto Rico (Day 1978c), Belize (McDonald 1976 1979) and elsewhere (Day and Tang 2004). The clays also act as reservoirs for water, and may perhaps promote dissolution at the clay/limestone interface, but the exact relationship between the surface deposits and the limestone hills is unclear, and the role of the deposits in the formation of the hills appears limited (Day 1978a 1986). Weiss (1994) attributed the irregular hills and ridges to differential, patchy cementation within the Antigua Formation, a hypothesis which merits further investigation given the eogenetic setting. A field survey of a 4 km^sup 2^ area 2 km southeast of Parham (Figure 1) located 17 hills with a density of 4.25/km^sup 2^ (Day 1978a). Densities in other areas are lower; four randomly located 1 km^sup 2^ quadrats had hill densities of 1, 1, 2 and 4/km^sup 2^ (mean 2/km^sup 2^, s = 1.225). Hill heights range from 9 to 40 m, with a mean of 19 m (s = 1 1 .8 m, n = 1 7) and median of 16 m. Diameters range from 75 to 520 m, with a mean of 1 75 m (s = 1 21 m, n = 1 7) and median of 140 m. Height/diameter ratios are low, ranging from 0.06 to 0.17, with a mean of 0.12 (s = 0.029, n = 17) and median of 0.11. A strong positive relationship exists between hill height and diameter (Spearman rank correlation coefficient r^sup s^ = 0.85, significant at the 0.01 level).

Hill width/length ratios (R^sub wt^) range from 0.46 to 0.75, with a mean of 0.61 (s = 0.083, n = 17) and median 0.60. Twelve of the hills (71% of the measured population) have ratios between 0.51 and 0.70. The hills exhibit two preferred long-axis orientations which are statistically significant as determined by Tanner’s (1955) test. These are in the classes 1-10[degrees] and 341-350[degrees], both approximately north-south. This is similar to the long axis orientations of some of the dolines, and may reflect the influence of joints. Single nearest neighbour measurements (Williams 1972) indicate that the 17 measured hills, represented by points located at their summits, exhibit a distribution similar to that which may result from a uniformly acting process. The actual mean distance between summits is 0.340 km (s = 0.118, n = 17) and the expected distance in a population of equal density is 0.236 km. The ratio of actual to expected distance is thus 1.441, which is significantly different from random (ratio 1.0) at the 0.05 level. Throughout the broader karst area, however, the hills appear to be markedly clustered.

Given the differential, patchy cementation within the Antigua Formation (Weiss 1994) and the possibility that at least some of the larger enclosed depressions are not dissolutional in origin, it is possible that the hills are partly depositional, rather than erosional in origin, and thus characterizing them as residual hills may be inappropriate. They are certainly morphologically and functionally different from karst towers, but at least some of them are capped by harder, more resistant limestone beds (Multer ef al. 1986), suggesting that erosional processes have played at least some role in their development.

Caves are poorly developed in Antigua because of the impurity of the limestone and the limited allogenic recharge. Nevertheless, there has been considerable potential for the development of flank margin or other hypogenic (mixing zone) caves (Mylroie and Carew 1995 1997; Mylroie et al. 1995; Mylroie 2004). Pending further investigation, these appear to be essentially absent, and certainly neither numerous nor extensive, probably because of the inhibition of speleogenesis by carbonate impurities, although mixing zone dissolution is generally rather insensitive to carbonate impurities. There is also no evidence of the high-density subsurface voids or ‘banana holes’ that characterize other small carbonate islands and represent a serious surface collapse risk (Mylroie et al. 1995).

The paucity of autogenic and allogenic recharge, combined with carbonate impurity, similarly, has inhibited epigenic speleogenesis, although the presence of sinks, such as that at the terminus of Carrs Gut, suggests that there is at least some local potential for autogenic development, and the apparent loss of water locally into the beds of the allogenic streams is also suggestive of underlying conduit formation.

Perhaps the best known cave is Bat (or Bats) Cave, which is located in St Paul’s Parish, near Shirley Heights, and is now a national park. Historically, the cave, which was shown on a map of 1749 (Shaw 2004), was a valuable source of saltpeter, and limited amounts of bat guano were extracted from it into the twentieth century for agricultural fertilizer. Martin’s Hole, near Cedar Grove, is named after an escaped prisoner who reputedly hid there for several months (Gurnee 1961).

There are also extensive contemporary coral reefs (Multer ef al. 1986), and the coastline of the karst area is markedly indented with some fine beaches (Bush et al. 2001). The coastal limestones, where they are subject to seawater spray, have developed upon them suites of littoral eogenic karren (Toborosi et al. 2005), although neither these nor the karst on the islands offshore from the mainland karst have been studied in any detail. The karren commonly take the form of small pinnacles less than 0.25 m in height, and, like those in other island karst areas, they are polygenetic, involving the interaction of salt water and rainwater dissolution, mechanical salt weathering and biological weathering (Viles 1988; Toborosi et al. 2005). Kamenitzas, shallow solution pans, are also developed locally. The conspicuous rhythmic sequences of the island-slope limestone beds in the upper part of the Antigua Formation give rise to characteristic ‘ledge-and-notch’ profiles on coastal cliffs that are subject to wave action, typically producing ledges 0.1-0.2 m thick and notches of similar or slightly smaller dimensions. The dimensions and morphology of both the karren and the ‘ledge-and- notch’ profiles may be related to variations in insoluble residues within the limestone, but this relationship has not yet been examined in detail.

Overall, the Antigua karst appears to be at an intermediate stage between eogenetic small island karst (Mylroie and Vacher 1999; Vacher and Mylroie 2002) and karst developed on the larger Caribbean islands, which is similar to that in continental settings. Eogenic island karst forms in young carbonates on small islands and coastlines that are affected by glacio-eustatic sea level fluctuations, but which have not been buried beyond the range of meteoric diagenesis. Such eogenic carbonates have not been extensively cemented or compacted and retain much of their primary depositional porosity, and island karst is defined by the CIKM (Mylroie and Vacher 1999; Vacher and Mylroie 2002), in which diagenetically immature carbonates interact with fresh and saline groundwater whose levels are controlled by glacio-eustacy and tectonic uplift.

Of the four types of carbonate islands identified within the CIKM, Antigua most closely resembles the Composite Island type, exemplified by Barbados. The non-carbonate/carbonate ratio is much higher in Antigua than Barbados, and although the non-carbonate rocks partition the subsurface freshwater lens, they contribute very limited allogenic groundwater recharge. Thus, the eogentic CIKM has less significance in Antigua than might otherwise be the case, although it is more applicable to Barbuda and to other small Caribbean carbonate islands.

Natural resources, land use, development and conservation

Human impacts on Caribbean karst landscapes have been long term and severe (Day 1993), in particular through forest clearing, species introduction, agriculture, utilization of water resources, urbanization and industrial activities, including mining and quarrying. These human impacts, all illustrated by Antigua, are particularly significant when integrated with the background of natural environmental instability, such as Holocene climatic change and Quaternary eustatic sea level variations (Watts 1987; Day 1993; Mylroie and Carew 1997).

The present population of Antigua is about 75 000, with a high population density of about 225/km^sup 2^ and an annual population growth rate of 0.55% (CIA 2006). Pre-European populations exerted early impacts on the karst environment, but these were overshadowed by an almost total destruction of the native vegetation for sugar cane cultivation during the colonial period. Subsequently, agriculture has declined in importance and tourism has become a major focus in the karst, necessitating construction of hotels and other infrastructure, and placing increasing pressures on the limited water resources. Development has been largely uncontrolled, and appropriate planning and conservation measures have been implemented only recently (Zachariah 1988).

Pre-European impacts

Antigua has a rich legacy of pre-Colombian settlement, and was an important regional source of flint and chert for fabrication of stone tools (Davis 1993). Most archaeological sites are within the karst on the east and northeast coasts (CCA 1991), presumably reflecting positive habitat selection by the pre-European populations. Native inhabitants had profound influences on the natural vegetation, using native species and introducing others (Harris 1965; Watts 1987). One-third of the total number of species represented as fossils in limestone fissure fills are either completely extinct or have never been recorded historically, suggesting extinction as a result of human-caused environmental degradation (Steadman ef a/. 1984). Agriculture

Colonial history was dominated by the production of sugar cane, which started around 1675 and lasted until the 1970s. Cotton was introduced shortly after colonization but gave way to sugar cane soon afterwards. It was reintroduced at the start of the twentieth century, but production dwindled and was essentially abandoned after Hurricane Hugo in 1989 (CCA 1991).

The primary human impact on Antigua’s karst has been the colonial era clearance of natural vegetation and the institution of commercial agriculture. Agriculture was the dominant traditional land use within the karst, and was dominated formerly by sugar cane and cotton. The Antigua Syndicate Estates, which dominated cane production in the early twentieth century, were located primarily in the karst (Augelli 1953). The intensity of this monoculture and the associated unsustainable agricultural practices degraded the vegetation and landscape of Antigua severely (Coram 1993; Watts 1987 1993), and it has been estimated that 92% of the total land area, including virtually all of that within the karst, was cleared for sugarcane cultivation and the fuelling of the evaporators used in sugar production (Cooper and Bowen 2001). Following the demise of these traditional crops, agriculture in the karst is now dominated by livestock production and subsistence cultivation of fruits and vegetables. There is a recent trend towards small-scale intensive production of organic produce for upmarket hotels and restaurants, and the potential for agricultural improvement through the use of irrigation is high, but water supply problems constrain such investment (Borden 1976). Overall, agriculture has declined from 20% of national GDP in 1960 to less than 4.0% of CDP in 2005 (CCA 1991; CIA 2006), but it is still important within the karst area, some 80% of which remains in some form of agricultural use (Day 1993). The limestone hills are also important areas for Antigua’s limited remnant woodland (Loveless 1960), and the mangroves are ecologically important but under increasing human pressure (CCA 1991). Accelerated soil erosion is a local problem within the karst, particularly in areas with large numbers of goats or other livestock, or where the vegetation is cleared or burned to encourage younger, more edible fodder (Plate 3). Sixty percent of the land is government owned, and farmers make little long-term investment (Lernet ef al. 1988; FAO 2003), reflecting former colonial land use policies (Augelli 1953). Land degradation is locally a problem within the karst, particularly as a result of vegetation clearance and overgrazing by livestock (UNCCD 2005), and there are also increasing concerns about contamination by agricultural chemicals (Rawlins ef al. 1998). Erosion hazard maps at a scale of 1 :5000 have been prepared for Antigua by OAS (2001b), and the susceptibility in the karst to inland erosion hazards is generally low to moderate.

Tourism

Tourism, which began in the 1950s and boomed in the 1970s and 1980s, now dominates the national economy, accounting for about 60% of GDP, 75% of foreign exchange earnings, and 25% of the labour force (Government of Antigua and Barbuda 2004a). Over 245 000 tourists visited Antigua in 2004 (Government of Antigua and Barbuda 2005). Although it is difficult to quantify exactly to what extent tourism is focused on the limestone area, tourism and its attendant infrastructure have had a major impact on the karst area, particularly through increasing demands for water and construction materials, and through the degradation of water quality.

Although tourism has provided the financial resources for infrastructure development and the enhancement of social services, it has also placed particular burdens upon the karst landscape. For example, the average visitor uses up to 10 times the amount of water as a resident, and nearshore waters are increasingly polluted by hotel activities, such as the discharge of wastewater into the sea, and by the application of pesticides and fertilizers to lawns and gardens (Sebastian 1998; Baldwin 2000). Hotel, marina and road construction have altered ecosystems, reduced biodiversity, disturbed archaeological sites and damaged historic artifacts throughout the karst and elsewhere (McElroy and de Albuquerque 1997). In particular, intertidal wetlands, which mediate terrestrial biological nutrient and sediment flows, have been severely disrupted, leading to a loss of productivity by coralgal communities and reduced sediment budgets (Baldwin 2000).

Particular controversy developed in 1992 over the proposed construction by a foreign developer of a large-scale resort complex at Coconut Hall. In response to the considerable public opposition, the Antigua Labour Party government conducted a belated Environmental and Social Impact Assessment (ESIA) in 1994, but an expanded and even more controversial project, encompassing Guiana Island, and termed the Asian Village project, was unveiled in 1997, subsequently to be vetoed by the United Progressive Party government, which was elected in 2004 (Government of Antigua and Barbuda 2004b).

Sewage and solid waste disposal are also increasing problems; for example, on Crabbs Peninsula and at Freetown (CCA 1991). The international airport, the main desalinization plants, a major marina, golf courses and about a third of Antigua’s hotels are all located within the karst belt.

Both natural resources and agriculture within the karst are beginning to be recognized as significant components of ecotourism (Cooper and Bowen 2001) and the former plantation system in the karst is itself a major tourist attraction (Weaver 1988). Investment in eco- and agro-tourism has, however, been limited by economic and labour constraints, and these sectors also remain vulnerable to environmental degradation (Lorah 1996; McElroy and de Albuquerque 1997). Scuba diving is an important element of tourism and is focused on the coral reefs. Beach erosion resulting from inappropriate development and sand extraction was formerly regarded as a local problem (CCA 1991), but has now become significant and endemic (Baldwin 2000). For instance, at Dickenson Bay, one of the most popular tourist beaches of northwestern Antigua, the beach eroded at a rate of 0.9 m per year between 1991 and 1994 (UNESCO 1996). Marine pollution and flooding have also been associated with resort development in the coastal karst (CCA 1991). Coastal erosion hazard maps at a scale of 1 :5000 have been prepared for Antigua by OAS, with the coastal erosion hazard in the karst being generally moderate to high (OAS 2001b). In this context, marine erosion and flooding are recognized as threats to structures on the coastal portions of the karst. Setbacks of 15 m from cliff edges are regarded as adequate protection against cliff retreat for the economic life (30 years) of most development projects, but a setback of 30 m from the vegetation line has been recommended for protection of rocky limestone shores against flooding during tropical storms and hurricanes (UNESCO 2006).

Ground subsidence

Locally, infrastructure within the karst is at risk of ground surface subsidence and collapse, although the risk of this can be minimized by appropriate land use planning and site investigation (Mylroie and Carew 1997; Waltham ef al. 2005). Some of the limestones are quite hard and competent (Martin-Kaye 1959), but they have localized cavities and zones of weakness within them. Around the village of Liberta, a distinctive green limestone is widely used as a decorative facing and provides sound bearing. In contrast, a white powdery limestone known as rock flour is found in small pockets in hard rock. This can be problematic, since it has little or no bearing capacity, but fortunately it is uncommon.

Quarrying

Although their output is small by Caribbean standards, quarrying and mining, mostly of limestone, contributed nearly US$30 million or about 1.7% of Antigua’s GDP in 2002 (IMF 2004). A cement factory was established on the Crabbs Peninsula in 1964 (Blume 1974), but Antigua still imports much of its cement, particularly from Trinidad. There are several operating limestone quarries, most notably the Army Base and the Piggots Hill quarries in the north of the island, and numerous abandoned operations. Softer limestone, generally known as marl, is quarried widely from small pits, and is used for house foundations and roadbed construction. The potential economic significance of limestone quarrying was emphasized by Mowat (1972) and Tomblin and Tomblin (1995), and the production of crushed stone, concrete blocks and cement increased by 50-75% between 1984 and 1988 (CCA 1991). In excess of 130 000 tonnes of crushed stone, the vast majority of it limestone, was produced in 1988, along with nearly 27 000 m^sup 3^ of concrete (CCA 1991). Historically, limestone was quarried widely for local construction, for example of public buildings and of windmills which were used in sugar production. The production of slaked or hydrated lime has also been important traditionally within the karst area, the lime being used for a wide variety of purposes including whitewash, water purification, liming of acidic soils and cooking. Sand has also been taken from beaches within the karst area for use in construction (Mowat 1972), although the government has recently increased efforts to stop this because of accelerated beach erosion and its incompatibility with tourism.

Water supply

Water supply is of paramount importance in Antigua, and particularly within the karst area, which has no permanent streams and receives the least rainfall in the island. There is a long history of groundwater development, the earlier phase of which has been reviewed by Martin-Kaye (1956 1959). Maintaining reliable water supplies is particularly difficult within the karst because there is only a limited fresh water supply and there is always the risk of salt water intrusion throughout the karst aquifer. There is an increasing need for comprehensive water resource and landscape management to meet rapidly increasing water demands (Williams 1994; Sebastian 1998; Cooper and Bowen 2001). Currently, the Ministry of Agriculture has the authority to construct water catchments, while the Antigua Public Utilities Authority (APUA) is responsible for coordinating water resource management and development. Drought and flooding hazard maps at a scale of 1 :5000 have been prepared for Antigua by OAS (OAS 2001b). Vulnerability to drought throughout the karst is rated as high or very high, whereas its vulnerability to flooding is rated as generally low, except in the North Sound catchment.

In terms of surface water supply, the Parham and Fitches Creek watersheds are among the six most significant in the country, and the Collins Bristol Springs well field is the most productive groundwater reservoir, with an annual production of 266 00Om^sup 3^ (Cooper and Bowen 2001). The potential annual groundwater yield is approximately 1.3 million m^sup 3^, and the APUA currently produces about 0.6 million m^sup 3^ of groundwater each year (UNCCD 2005).

Saltwater intrusion is a problem throughout the karst area, particularly in the vicinity of the valleys, and high salinity levels pose a problem for much of the groundwater, particularly following drought or excessive pumping (Sandberg and Barnes 2004). High levels of total dissolved solids also generally limit the direct use of ground water for human consumption, although less stringent limits are placed on the use of the water for agricultural purposes. A multi-variant, matrix approach for preliminary assessment of regional fractured bedrock aquifer potential (Bisson and Maharaj 2003) suggests that the limestones possess substantial fracture permeabilities benefiting from limited rainfed recharge, warranting further exploration of the potential for sustainable brackish water sources. Such brackish water in the fractured bedrock aquifers is potentially of considerable importance to the tourism and agricultural sectors (Bisson and Maharaj 2003).

There are two desalinization plants in the karst on Crabbs Peninsula. The larger thermal plant, which was opened in 1970 (Blume 1974), has a capacity of 10 600m^sup 3^/day. The second plant, commissioned in 1993, is a reverse osmosis plant with a capacity of 3040 m^sup 3^/day (UNCCD 2005). During drought, the desalinisation plants provide up to 83% of all freshwater (UN 2002). Discharge of hyperthermal and hypersaline water from the thermal plant poses a threat to the local marine ecosystem (CCA 1991).

Many individual residences have cisterns which provide part or all of the household water needs. By law, all new houses are supposed to be equipped with rainwater collection and storage systems, with an average storage of 19 m^sup 3^ (5000 gallons) (UNCCD 2005). Other needs are supplied by surface dams, numerous small ponds (many in dolines) and well fields (McMillan 1985).

Waste water disposal also poses a threat to karst groundwater in Antigua. Some rural houses have only rudimentary, or no disposal systems, but domestic sewage is increasingly collected in underground septic tanks and the effluent piped to rock-filled soak- aways. Septic tanks may fail, especially if not emptied regularly, and soils with high clay contents allow very little absorption, such that the soak-away does not function well. In lieu of a soak-away, a drain field consisting of perforated PVC pipes laid in a shallow rock fill may be used to spread the effluent over a large area where it may be absorbed by the topsoil, transpired and evaporated.

Conservation and protected areas

A total of 66 km^sup 2^ or about 1 5% of Antigua’s land area is conserved within 1 1 designated protected areas, but of the approximately 110 km^sup 2^ of karst in Antigua, less than 10 km^sup 2^ (8%) is afforded any kind of protection or conservation (Kueny and Day 1998). The National Parks Act (1984) is the primary protected areas legislation, although the country is a signatory to the Convention on Biological Diversity, the Convention of International Trade in Endangered Species of Wild Fauna and Flora and the Specially Protected Areas and Wetlands Protocol. The Government of Antigua and Barbuda has recently developed a biodiversity strategy and action plan which, among other things, calls for the establishment of an integrated national protected areas system (Government of Antigua and Barbuda 2001).

To date, the only formal karst and cave protection implemented is at the small national park sites of Bats Cave, Great Bird Island, Devils Bridge and Jabberwock Beach. Great Bird Island is the base for the Antigua Racer Conservation Project (ARCP) (now known as the Off Shore Island Conservation Project (OICP), which was formed in 1995 to save the Antiguan racer snake (Alsophis antiguae) from extinction. The project also aims to conserve other kinds of rare wildlife that depend on Antigua’s limestone islands. OAS funded the project to establish a Marine Reserve and Wildlife Sanctuary on Great Bird Island, and a management plan for the proposed protected area was drawn up by the Island Resources Foundation as part of the Biodiversity Conservation Programme for the Eastern Caribbean (Potter ef al. 1997).

Additionally, several areas within the karst, including Santa Maria Hill, Ayres Creek, the Potworks and Collins Reservoirs and several of the offshore islands have been identified as potential biodiversity protection sites, and Corbison’s Point has been proposed as a national park (CCA 1991). Other small areas currently being protected are privately owned and managed. The Environmental Awareness Group, an NGO, was established in the late 1980s, and a number of community-based sustainability projects are in the earliest stages. An Environmental Division in the Ministry of Tourism and Environment is now charged with the executive responsibility of managing and administering various aspects of the nation’s physical, biological and cultural environment.

Antigua’s coastline, including that within the karst area, is characterized by bays and headlands which represent natural ‘coastal compartments’ (Bush ef al. 2001) and which provide opportunities for prioritizing localized collective land management plans through a Coastal Compartment Management Plan (CCMP). This considers the uniqueness of compartments based on rapid geological analyses of beach settings, hydrodynamics and hazard geoindicators (Bush ef al. 2001). Dickinson Bay (Figure 1) has been identified as one such priority compartment based on significant shoreline length, wave exposure, historical storm response, state of the shoreline, number of existing hotels and pressure for further development, and such an approach is applicable throughout the carbonate area.

Future prospects

Pressures on Antigua’s karstlands are already severe, and they will only be exacerbated by predicted anthropogenic climate change and other increasing human pressures. General predictions for the Caribbean are that anthropogenic climatic change will lead to increasing air and water temperatures, rising sea levels and changing weather patterns, including decreasing precipitation totals, and the increasing frequency of extreme events, such as droughts and hurricanes (Nurse and Sem 2001). The effects of all these changes will be magnified in the Antigua karst, where water resources are already highly limiting, and disruption of the karst hydrological cycle may lead to increasing aridity and desertification, with concomitant impacts on geomorphic processes, ecology and potential land use.

Increasing population and economic development, particularly through tourism, will further exacerbate human impacts on the karst landscapes, in particular through species introduction, expanding agriculture, increasing utilization and contamination of water resources, urbanization and quarrying. The severity of these impacts can best be reduced by appropriate land management and land use planning in the karst, including the expansion and maintenance of protected areas.

Conclusion

Studies of the Antiguan karst to date are preliminary, and there is considerable potential for further research. Although an understanding of the variation within the Antigua Formation limestones has improved considerably since the 1970s, the role of these variations in influencing karst landform development still remains uncertain. The role of the surficial deposits also remains unclear.

Development of the valley systems is poorly understood, and their contemporary hydrology merits attention in the context of water resources and flooding hazards. The contribution of the allogenic recharge to the karst hydrology appears to be minor, but it may be significant locally. Similarly, the precise nature of the autogenic recharge mechanism is unclear, although the vast majority of the water appears to recharge diffusely, with only localized focusing of near surface flow in dolines or at the termini of dry valley systems such as Carr’s Gut.

The dolines initially appear unremarkable by Caribbean standards, but it is possible that they have a variety of origins. The smaller individuals appear to have been produced by near surface dissolution, and their small numbers, limited depths and low density reflect the inhibition of point-focused near-surface dissolution by diffuse input and limited recharge areas. Near-surface flow concentration may be associated with more insoluble and less permeable units within the limestone, and the clustering of dolines supports the possibility, but further investigation is necessary to clarify the situation. In contrast, some of the larger depressions, particularly those exceeding 100 m in diameter, may have a depositional origin. These contrasting origins also relate to contemporary doline activity, which appears to represent a localized hazard only by way of limited subsidence or collapse in the smaller, dissolutional depressions. There is no evidence of the numerous subsurface voids or ‘banana holes’ that characterize other small carbonate islands (Mylroie ef al. 1995). The limestone hills represent a subdued form of karst residual, and their development also warrants further study. It seems likely that depositional and diagenetic variations within the Antigua Formation account for many of the hills and ridges, and it is possible that some hills are constructional, rather than erosional in origin, and thus characterizing them as residual hills may be inappropriate. They are certainly morphologically and functionally different from karst towers, but the caps of harder, more resistant limestone beds (Multer ef al. 1986) suggest that erosional processes have played at least some role in their development.

Speleogenesis in Antigua has been restricted by the impurity of the limestone, the limited allogenic recharge and the diffuse nature of autogenic recharge. Moreover, the potential for the development of flank margin or other hypogenic (mixing zone) caves appears to have gone unrealized, and there is no evidence of the variety of caves that characterize other small carbonate islands.

Overall, the Antiguan karst morphology is intermediate between types that have been described and quantified elsewhere in the Caribbean (Day 1978a) and fits within the composite island group in the CIKM (Mylroie and Vacher 1999; Vacher and Mylroie 2002), although karst and cave development appears to have been restricted by the impurity of the limestone and the limited or poorly focused recharge.

There is a long history of considerable and increasingly disruptive human activity within the karst area. Recent declines in intensive agricultural activity within the karst provide significant opportunities for increased conservation and landscape restoration activities, which are compatible with the emphasis on sustainable tourism, but these have been slow to develop. Tourism itself poses problems within the karst, particularly along the coastline, around the airport and in the vicinity of St Johns. Groundwater supplies remain of paramount importance, but rising demands increasingly necessitate use of supplementary resources, including desalinization. Pressures on Antigua’s karstlands are already severe, and they will only be exacerbated by predicted anthropogenic climate change and other increasing human pressures. Successful and sustainable land use in the Antiguan karst will require understanding of both the karst landscape and contemporary karst processes.

Acknowledgements

Research in Antigua has been funded by a scholarship from the Natural Environment Research Council (UK) and by grants from the University of Wisconsin-Milwaukee’s Center for Latin American and Caribbean Studies. I am grateful to John Briggs and the anonymous reviewers for their suggestions which improved the original manuscript. I dedicate this contribution to the late Desmond Nicholson, whose knowledge of Antigua was encyclopaedic, and whose enthusiasm for the island was contagious.

References

Augelli J P 1953 Patterns and problems of land tenure in the Lesser Antilles: Antigua, B.W.I. Economic Geography 29 362-7

Baldwin J 2000 Tourism development, wetland degradation and beach erosion in Antigua, West Indies Tourism Geographies 2 193-218

Barker G H 1981 Antigua and Barbuda: an agricultural profile Caribbean Agricultural Research and Development Institute, St Augustine

Beard J S 1948 The natural vegetation of the Windward and Leeward Islands Oxford Forestry Memoirs 21

Bisson R A and Maharaj U 2003 ‘Megawatersheds of the Caribbean: new water from old rocks’ Paper presented at the Caribbean Water and Wastewater Association (CWWA) 1 2th Annual Conference, Bahamas 29 September-3 October 2003 (http://www.earthwaterglobal.com/pubs/ Megawatersheds%20of20the%20Caribbean- %20New%20Water%20from%20Old%20Rocks.pdf) Accessed 10 May 2006

Blume H 1974 The Caribbean Islands Longman, London

Borden 1976 A method for determining the irrigation requirement in Antigua, W.I. Tropical Agriculture 53 41-5

Bush D M, Young R S and Neal W J 2001 Managing Caribbean coasts: the Coastal Compartment Management Plan Geological Society of America Abstracts with Programs 33 1 05

Caribbean Conservation Association (CCA) 1991 Antigua and Barbuda: country environmental profile CCA, St Michael, Barbados

CIA 2006 The World fact book (httpyAvww.cia.gov/cia/publications/ factbook/geos/ac.html) Accessed 10 April 2006

Cooper B and Bowen V 2001 Integrating management of watersheds and coastal areas in small island developing states of the Caribbean Environment Division Ministry of Tourism and Environment, Government of Antigua and Barbuda

Coram R 1993 Caribbean time bomb William Morrow, New York

Davis D O 1993 Archaic blade production on Antigua, West Indies American Antiquity 58 688-97

Day M J 1978a The morphology of tropical humid karst with particular reference to the Caribbean and Central America Unpublished DPhM thesis Department of Geography, University of Oxford

Day M J 1978b Engineering hazards in tropical karst terrain Applied Geography Conferences 1 288-98

Day M J 1978c Morphology and distribution of residual limestone hills (mogotes) in the karst of northern Puerto Rico Bulletin of the Geological Society of America 89 426-32

Day M J 1979 Surface roughness as a discriminator of tropical karst styles Zeitschrift fur Geomorphologie Supplementbande 32 1-8

Day M J 1980 Rock hardness: field assessment and geomorphic importance The Professional Geographer 32 72-81

Day M J 1982 The influence of some material properties on the development of tropical karst terrain Transactions British Cave Research Association 9 27-37

Day M J 1983 Doline morphology and development in Barbados Annals of the Association of American Geographers 73 206-19

Day M J 1985 Limestone valley systems in north-central Jamaica Caribbean Geography 2 16-33

Day M J 1986 Karst in Antigua, West Indies Proceedings 9th International Congress of Speleology 1 218-220

Day M J 1993 Human impacts on Caribbean and Central American karst in Williams P W ed Karst terrains: environmental changes and human impact Catena Supplement 25, Cremlingen-Destedt 109-25

Day M J and Chenoweth M S 2004 The karstlands of Trinidad and Tobago, their land use and conservation The Geographical Journal 170 256-66

Day M J and Goudie A S 1977 Field assessment of rock hardness using the Schmidt Test Hammer British Geomorphological Research Group Technical Bulletin 18 19-29

Day M J and Tang T 2004 Tower karst in Gunn J ed Encyclopedia of caves and karst science Taylor and Francis, New York 734-6

FAO 2003 National Forestry Action Plan, Antigua and Barbuda (http://www.fao.org/docrep/X5681 E/x5681e00.HTM) Accessed 5 September 2005

FAO 2004 Irrigation in Antigua and Barbuda (http^/www.fao.org/ Regional/LAmerica/paises/h2o/antigua.htm) Accessed 4 September 2005

Fermor J 1972 The dry valleys of Barbados: a critical review of their pattern and origin Transactions of the Institute of British Geographers 57 153-65

Government of Antigua and Barbuda 2001 Biodiversity strategy and action plan for Antigua and Barbuda (http://biodiv.org/doc/world/ag/ ag-nbsap-01-en.doc) Accessed 10 January 2006

Government of Antigua and Barbuda 2004a Antigua and Barbuda: national strategy to strengthen trade-related capacity Government Printer, St Johns

Government of Antigua and Barbuda 2004b Statement by the Attorney General, Honourable Justin Simon delivered 18 August 2004, Guiana Island Development Project (http://www.antigua-barbuda.com/ business_politics/attorneygeneral_asianvillage.asp) Accessed 10 January 2006

Government of Antigua and Barbuda 2005 Tourist arrivals 2004 (http://www.ab.gov.ag/gov_v2/government/statsandreports/index.html) Accessed 12 April 2006

Graham A 2003 Geohistory models and Cenozoic paleoenvironments of the Caribbean region Systematic Botany 28 378-86

Gregory K J 2000 Drainage basins in Hancock P L and Skinner B | eds The Oxford companion to the Earth Oxford University Press, Oxford 256-62

Gurnee R H 1961 The caves of Antigua and Barbuda International Speleologist 1 14-18

Harris D R 1960 The vegetation of Antigua and Barbuda, Leeward Islands, the West Indies University of California Press, Berkeley

Harris D R 1965 Plants, animals and man in the outer Leeward Islands, West Indies University of California Press, Berkeley

Hill I D 1966 Antigua Soil land-use survey of the British Caribbean, World Land Use Survey 19 Ebbingford, Bude

IMF 2004 IMF country report no. 04/369: Antigua and Barbuda: statistical appendix (http://www.imf.org/external/pubs/ft/scr/2004/ cr04369.pdf) Accessed 2 September 2005

Kueny J A and Day M J 1998 An assessment of protected karst landscapes in the Caribbean Caribbean Geography 9 87-100

Lasserre G 1954 Notes sur Ia karst de Ia Guadeloupe Erdkunde 8 115-18

Lemel H W, Stanfield D and Deterville H 1988 A study of the land tenure structure of Antigua and Barbuda: public and private interests in land Land Tenure Center, University of Wisconsin- Madison

Lewis J 1984 Multi-hazard history of Antigua Disasters 8 190-7

Lindsay K and Horwlth B 1997 A vegetation classification of Antigua, Barbuda and Redonda Island Resources Foundation, Eastern Caribbean Biodiversity Programme Publication 2

Lorah P 1996 An unsustainable path: tourism’s vulnerability to environmental decline in Antigua Caribbean Geography 6 28-39

Loveless A 1960 The vegetation of Antigua, West Indies Journal of Ecology 48 495-527

Martin-Kaye PHA 1956 The water resources of Antigua and Barbuda, B.W.I British Guiana Lithographic Co La Penitence, British Guiana

Martin-Kaye PHA 1959 Reports on the geology of the Leeward and British Virgin Islands Voice Publishing Co, Castries

McDonald R C 1976 Hillslope base depressions in tower karst topography of Belize Zeitschrift fur Geomorpholgie Supplementbande 26 98-103 McDonald R C 1979 Tower karst geomorphology in Belize Zeitschrift fur Geomorphologie Supplementbande 32 35-45

McElroy J E and de Albuquerque K 1997 Community and NGO initiatives in coastal conservation: lessons from Antigua and Barbuda Caribbean Geography 8 18-31

McMillan J 1985 Evaluation of agricultural water supplies in Antigua and Barbuda OAS, Washington DC

Mowat R 1972 An economic geological investigation of nonmetallic mineral resources for the construction industry in the Leeward- Windward Islands and British Honduras Caribbean Development Bank Technical Report 8 Bridgetown, Barbados

Multer G, Weiss M and Nicholson D 1986 Antigua: reefs, rocks and highroads of history Leeward Islands Science Associates, St John’s

Mylroie J E 2004 Speleogenesis: coastal and oceanic settings in Gunn J ed Encyclopedia of caves and karst science Taylor and Francis, New York 674-7

Mylroie J E and Carew J L 1995 Karst development on carbonate islands in Budd D A, Harris P M and Staller A eds Unconformities and porosity in carbonate strata American Association of Petroleum Geologists, Tulsa 55-76

Mylroie J E and Carew J L 1997 Land use and carbonate island karst in Beck B F and Stephenson J B eds The engineering geology and hydrogeology of karst terranes A A Balkema, Rotterdam 3-12

Mylroie J E, Carew J L and Vacher H L 1995 Karst development in the Bahamas and Bermuda in Curran H A and White B eds Terrestrial and shallow marine geology of the Bahamas and Bermuda Geological Society of America, Boulder 251-67

Mylroie J E and Jenson J W 2002 Karst flow systems in young carbonate islands in Martin J B, Wicks C M and Sasowsky I D eds Hydrogeology and biology of post Paleozoic carbonate aquifers Karst Waters Institute, Charles Town, West Virginia 107-10

Mylroie J E and Vacher H L 1999 A conceptual view of carbonate island karst in Palmer A N, Palmer M V and Sasowsky I D eds Karst modelling Karst Waters Institute, Charles Town, West Virginia 48-57

Nurse L A and Sem G 2001 Small island states in McCarthy J J, Canziani O F, Leary N A, Dokken D J and White K S eds Climate change 2001: impacts, adaptation, and vulnerability IPCC Cambridge University Press (httpy/www.grida.no/climate/ipcc_tar/wg2/ index.htm) Accessed 10 May 2006

OAS 2001a Drought risk Antigua (http://www.oas.org/pgdm/hazmap/ drought/ant_map.pdf) Accessed 10 September 2005

OAS 2001b Post-Georges disaster mitigation hazard vulnerability assessment: Antigua and Barbuda (http://www.oas.org/PGDM/hazmap/ ab_va.htm) Accessed 10 September 2005

Potter B, Horwith B and Lindsay K 1997 Management plan for the Bird Island (Antigua) Marine Reserve Island Resources Foundation, Eastern Caribbean Biodiversity Programme

Rawlins B G, Ferguson A J, Chilton P J, Arthurton R S, Rees J G and Baldock J W 1998 Review of agricultural pollution in the Caribbean with particular emphasis on small island developing states Marine Pollution Bulletin 36 658-68

Sandberg S K and Barnes H 2004 Geophysical methods applied to saltwater intrusion in Antigua in Allred B ed Symposium on the application of geophysics to engineering and environmental problems Environmental and Engineering Geophysical Society, Wheat Ridge, Colorado 936-43

Sebastian C T 1998 Water and tourism: the Antiguan-Barbudan context Caribbean Currents 6 2-3

Shaw T 2004 Caribbean islands: history in Gunn J ed Encyclopedia of caves and karst science Taylor and Francis, New York 191-2

Steadman D W, Pregili G K and Olson SL 1984 Fossil vertebrates from Antigua, Lesser Antilles: evidence for late Holocene human- caused extinctions in the West Indies Proceedings of the National Academy of Sciences 8 1 444851

Tanner W F 1955 Paleographie reconstructions for crossbedding studies Bulletin American Association of Petroleum Geologists 39 2547-63

Toborosi D, Jenson J W and Mylroie J E 2005 Karren features in island karst: Guam, Mariana Islands Zeitschrift fur Geomorphologie NF 48 369-89

Tomblin J F and Tomblin J M 1995 Mineral resources of the Lesser Antilles, West Indies British Geological Survey Technical Report WC/ 95/58R, Nottingham

UN 2002 Antigua and Barbuda: country profile (http:///www.un.org/ esa/agenda21/natlinfo/wssd/antigua.pdf) Accessed 6 April 2006

UNCCD 2005 Draft national action plan for Antigua and Barbuda(httpy/www.unccd.int/actionprogrammes/lac/national/2005/ antigua_and_barbuda-eng.pdf) Accessed 6 May 2006

UNESCO 1996 Beach erosion in Antigua-Barbuda (http:// www.unesco.org/csi/act/cosalc/project9.htm) Accessed 18 April 2006

UNESCO 2006 Environment and development in coastal regions and in small islands (http://www.unesco.org/csi/pub/info/info410.htm) Accessed 5 May 2006

Vacher H L and Mylroie J E 2002 Eogenetic karst from the perspective of an equivalent porous medium Carbonates and Evaporites 17 182-96

Viles H A 1988 Biogeomorphology Basil Blackwell, Oxford

Waltham T, Bell F and Culshaw M 2005 Sinkholes and subsidence Springer, Berlin

Wameford FHS 1949 An introduction to the geology of Antigua Private Publication St Johns, Antigua

Watts D 1987 The West Indies: patterns of development, culture and environmental change since 1492 Cambridge University Press, Cambridge

Watts D 1993 Long-term environmental influences on development in islands of the Lesser Antilles Scottish Geographical Magazine 109 133-41

Weaver D 1988 The evolution of a plantation tourism landscape on the Caribbean island of Antigua Tiddschrift voor Economische en Sociale Geographie 79 319-31

Weiss M P 1994 Oligocene limestones of Antigua, West Indies: Neptune succeeds Vulcan Caribbean Journal of Science 30 1-29

Williams C 1994 Towards the sustainable