The study area is located at the northern end of the Gulf of Aqaba (Figure 1), which is the northward extension of the desert-enclosed Red Sea. The Gulf is a semiisolated basin separated from the Red Sea proper by the Straits of Tiran, which are about 240 m deep. The maximum depth of the Gulf is 1830 m; its 180 km long and 5 - 26 km wide.

The Gulf is located within the very warm portion of the Sahara bio-climatic zone. The climate is arid with high evaporation (~400 cm/year) and negligible precipitation (~2.2 cm/year) and runoff [27]. The mean sea surface temperatures (SSTs) are 23.5˚C and mean salinity values in the upper waters are 40.4‰ - 40.6‰ [28].

Extremely oligotrophic conditions are prevailing in the Gulf due to the arid climate and because it receives its waters from the nutrient-depleted Red Sea surface waters through the Straits of Tiran [27]. The deep light penetration and high transparency due to low amount of resuspended materials and fresh water flux [29] results in extending the depth limit of massive hermatypic corals such as Porites down to 40 - 50 m [30,31].

All investigated corals in this study were collected from a depth transect from the reef complex in front of the Marine Science Station in Aqaba (29˚27′N, 34˚58′E), (Figure 1), except the corals AQB-Big, AQB-18H, AQB-ph. The transect was laid out on the reef between the reef flat and the deep fore-reef down to 42 m. Scleractinian corals were distributed along this transect at all depths.

A total of 32 columns of Porites spp. colonies (Table 1) were collected along this transect from different depths between 1 and 42 m and at different time periods (April 1999, June 2000, May 2005 and June 2010). Colonies were harvested by scuba diving by removal of hemispherical-shaped colonies (5 - 10 cm in diameter) underwater using hammer and chisel. After sampling, the corals were cleaned under high-pressure tap water to remove the residual organic matter and then dried under the sun.

The corals were sectioned longitudinally into slabs of 4 mm thickness parallel to the axis of maximum growth. The slabs were cleaned and X-rayed using industrial X-ray machine. Black-and-white positive prints of the x-radiographs were obtained for extension rate measurement.

The annual skeletal extension rates of all corals were directly measured along the major (vertical) growth axes (Figure 2) from the positive prints of the x-radiographs as the distance between the top edges of low-density bands [8]. Each couplet of high-/low-density bands represents an annual growth increment. As a double check for this method, the annual extension rates of corals were calculated from the seasonal cycle of δ¹⁸O (or Sr/Ca) as the distance from the maximum δ¹⁸O (Sr/Ca) value (which corresponds to the seasonal temperature minimum) in a given year to the maximum value in the fol-

lowing year [31-33]. The mean extension rate of a given colony was calculated as the average of its annual values.

Corals investigated in the present study reveal clear and regular skeletal density banding patterns of alternating bands of high and low-density (Figure 2). All coral samples were retrieved between April and June and they showed their high-density annual growth bands in the outermost parts suggesting that they were formed during winter season.

The timing of annual density bands formation from this study is similar to that obtained by Rosenfeld et al. [34] whom found that low-density skeleton is produced during summer and high-density skeleton is produced during winter in the same Porites corals grown in both shallow and deep growth phases (corals transplanted from 6 to 40 m). However, Klein et al. [30] documented the deposition of high-density bands during winter only in shallow water (3 m) Porites colony and opposite pattern in deep water (51 m) coral.

It was suggested that seawater temperature and light intensity are the major factors influencing the density patterns in Porites corals from the Red Sea [35]. The seawater temperature record in the study area shows strong seasonal pattern with highest temperature occurring in August (26.32˚C) and lowest in March (21.18˚C), [31].

The annual extension rates measured using the x-radiography and the δ¹⁸O (or Sr/Ca) curves reveal identical results. Mean annual extension rates of the 32 Porites spp. corals as obtained from 181 annual growth bands reveal a large inter-colony variability with values ranged between 2.36 to 20.0 mm/year (average 9.48 ± 4.98 mm/year). The maximum extension rate value of 20.0 mm/year was recorded in corals growing at a depth of 19 m and the minimum of 2.36, 2.40 mm/year in corals from 27 and 42 m depth, respectively (Table 1, Figure 3). These values fall within the global average of 12 mm/year in massive corals under normal conditions [14,15] and within the range of values presented by other authors from the Gulf of Aqaba [30,33,36-40]. For example, Heiss et al. [36] found that the growth rates of Porites corals in shallow water from Aqaba (1 - 3 m) vary between 8.64 - 12.26 mm/year, which is high for reefs at this latitude.

Most interesting, is the highest extension rates (15.2 - 20 mm/year) measured in corals growing at depths around 20 m (Aq19-A, Aq19a, Aq19b, AQB-T20-B1 and AQB-T-20-C1). This could be attributed to the position of these colonies, since they stand in the upper fore reef and the top of these colonies is about 1.5 m above the sea bottom. This will elevate the colony to a position of either more light, less abrasion by moving sand or both [23]. The same reason could be also responsible for the

high extension rate value (16.3 mm/year) obtained from coral AQB-Big which is a huge colony and lies only 1 m below seawater surface. Similarly, Heiss [41] found that growth rates higher than 8 mm/year could occur in colonies derived from a free-standing Porites pinnacle.

In Caribbean and Indo-Pacific regions, Porites corals exhibit annual growth increments of about 10 mm/year for shallow water region and a decrease to 4 mm/year in more deeper waters. However, it is noticed that even in the deeper water environment (e.g., 10 - 25 m), the annual growth rates is still high with up to 7 mm/year [10]. The decline in coral extension rates in deeper waters could reflect the position of a light threshold below which calcification is reduced as suggested earlier by Chalker et al. [42]. For Caribbean reefs, Adey [43] cited 15 m as a controlling depth for reef initiation after Holocene sea level rise slowed; it is also the cut off between shallow and deep reef species on many Caribbean reefs.

In general, coral extension rates from this study decreased with increasing water depth (excluding the high values from 19 - 20 m depth). The inverse relationship between mean annual extension rates and depth for Porites spp. corals is non-linear and can be explained by a simple exponential model as follows (Figure 3):

where coral extension rate in mm/year and d is the water depth in meter.

The values tended to cluster around two groups, one at 10.86 mm/year in water depths less than 10 m (range between 7.8 - 16.3 mm/year), and the other around 5.23 mm/year below 12 m depth (range between 2.36 - 9.63 mm/year). The analysis of variance (one way ANOVA) showed that there is statistically significant difference in the mean extension rate of corals between the two groups (p < 0.0001).

The environmental factors that may control extension rates in massive coral include water depth, sedimentation and turbidity, nutrient concentration, seawater temperature and salinity [17,44,45]. In the Gulf of Aqaba, environmental factors such as precipitation, cloud cover and freshwater discharge (terrigenous sediment input) and its effect on salinity, sedimentation and turbidity can be eliminated since the area is a desert-enclosed sea, and annual precipitation is less than 2.2 cm/year [27]. Therefore, the Gulf of Aqaba is characterized by high water clarity and deep light penetration [29].

Depth is considered as a mixed environmental variable including the effect of light, water movement and resuspension of sediment and organic matter, and in some places temperature and/or salinity [10,14]. Seawater temperature effects are most probably not the reason for the decrease in coral extension rate with depth in this study since the Gulf of Aqaba is isothermal over the studied depth range. In the study area and from the same reef transect, the annual water temperature decreases with depth between the depths 7 and 42 m by less than 0.25˚C, and the salinity variations was less than 0.3‰ [31]. Light intensity (mE·m⁻²·s⁻¹) was measured at monthly basis at a steep reef slope (3 - 40 m depth) from the study area by Kampmann [46]. The record shows that light levels decreases exponentially with depth and coral extension rates from this study follow the same trend (Figure 4).

At 40 m depths, light intensity drops to about 50 mE·m⁻²·s⁻¹ in which extension rate of corals is minimum (2.40 mm/year). Therefore, light seems to be the primary environmental factor responsible for decreasing coral extension rate with depth and other parameters are fairly constant throughout the water column from the Gulf of Aqaba. Light enhanced calcification is responsible for most of the skeletal growth and carbonate production of reef building corals [47,48].

In inter-annual time scale, coral extension rate varies within the same colony over the studied time span. The time series record showed a fluctuated pattern with slightly increasing trend of the extension rate in both shallow and deep corals for the period of time represented despite the significant difference in extension rate between them (Figure 5).

Local sea surface temperature record (SST) from the study site shows similar increasing trend between 1988- 2009 (Figure 5). However, the correlation between the SST record and extension rates from both corals is not strong (r = 0.38, 0.24 for AQB-Big and AQB-T-42-A2, respectively). Several studies have documented the sensitivity of extension in Porites corals to temperature [16,17] over a wide SST ranges (22˚C - 29˚C) and they reported a significant positive correlation between both variables [17]. In our study, the range of inter-annual SST variability during the overall period (1988-2009) is relatively too small (fluctuated by less than 1.2˚C) to be a factor of primary importance affecting coral annual extension rates. Therefore, longer time series records of both SST and annual coral extension rate are needed to validate this relation from the northern Gulf of Aqaba.