A Light Pollution Assessment in the Fringing Reefs of San Andrés Island: Towards Reducing Stressful Conditions at Impacted Coral Reefs

When looking at the night sky, we quickly notice the presence of stars; we begin to count them one by one and notice that more of them appear between each count. In this process, we are also able to identify perhaps some planets, the star clusters that gave rise to the Pleiades, and constellations that inspired the great poets of ancient Greece and mythical Arab navigators, and filled with knowledge the astronomers and physicists of modern times, and even, astrologers! The interpretation of zodiacal lights has captivated the curiosity of many people, permeating many areas of daily life. For example, the moon phases are linked to the agricultural sowing and harvesting seasons. However, it has become increasingly difficult to observe and interpret the dynamics of the night sky due to light pollution, the most influential factor affecting astronomical observation (Hamidi et al. 2011).

This chapter focuses on a scientific perspective about natural and artificial light and how light pollution affects organisms from coral reefs around San Andrés Island. Based on astronomical information provided by NOAA and Unihedron Sky Quality Meter, and data collection of the sky’s brightness with in situ sensors, we were able to determine that the quality of natural light has been declining with the recent increase of the artificial light sources, due to constant population growth around the coast and Hurricane Iota’s impact decimating coastal vegetation, populations of marine species located in the fringe and leeward reefs will be more vulnerable, mainly when synchronizing their reproductive cycle with the vertical light spectrum.

The natural light observed during night hours comes from multiple sources such as the Milky Way, stars, and the Moon, which is fundamental in the diurnal, nocturnal, and seasonal cycles, and that plays a particular role in the behavior patterns of marine and terrestrial animals (Gaston et al. 2017; Luarte et al. 2016). However, in the last century, significant anthropogenic light sources gained relevance, disrupting the quality of natural light. This phenomenon is associated with the increase in the world population, specifically in the coastal areas where it is occurring faster and where demand for energy has increased with respect to other zones (Gaston et al. 2013, 2015). Consequently, the periods in which ecosystems are exposed to artificial light are longer and with more incidence, this is known as ALAN (Artificial Light Pollution at Night), and it has widespread negative effects on diurnal and nocturnal organisms (Sanders et al. 2021), including birds (McLaren et al. 2018), Fish (Pulgar et al. 2019), marine turtles (Dimitriadis et al. 2018) and insects communities (Grubisic and Van Grusven 2021). Many coastal marine ecosystems are currently exposed to artificial light at night, affecting the biological clocks of both marine and terrestrial organisms (Tamir et al. 2017). Many reef dweller species naturally synchronize their reproductive events with signals from the night sky (vertical light), mainly linked to changes in the moon’s phases.

ALAN can also intervene in the survival of coral symbionts (specifically green algae) in species such as Pocillopora damicornis and Acropora euristoma, that after 120 days of exposure to light, the density of their microbial communities exhibits significant differences (Ayalon et al. 2019; Levy et al. 2020). Studies indicate that lunar irradiance affects both the maturation of gametes and their release, and it is a precursor of speciation in populations distributed along a bathymetric/light gradient. However, ALAN also affects reproductive behavior in corals due to its dependency on lunar cycles. In the long term, with the intervention of natural light patterns, sexual selection, reproductive isolation, gene flow, and genetic drift can be altered (Hopkins et al. 2018), due to the close relationship between the quality and quantity of natural light with the responsible genes of the reproduction, as is the expression of the cryptochromes (CRY genes) (Poehn et al. 2021). Its expression mechanism is based on the perception of a specific wavelength known as blue pulse (Levy et al. 2007; Sweeney et al. 2011), this environmental signal is the most predominant light spectrum within the photic zone of the sea. For the Cryptochromes (Cry genes) to recognize the spectral signals and, thus, for the organism to be able to synchronize for reproduction, this wavelength or blue pulse must be specific and of “good” quality (Kronfeld-Schor et al. 2013).

Artificial light is a new, silent yet highly visible enemy, altering these natural cycles, in turn affecting the resilience capacity of these marine ecosystems. The increase of this anthropogenic pressure is poorly studied in marine ecosystems, even if they mask and deteriorate the spectral quality of the light penetrated in the column water (Davies et al. 2013). Precise moonlight intensities comprise the proximate reproductive cue for many marine sessile invertebrates (Coelho and Lasker 2014). Particularly in coral reefs, part of the reproductive success of benthic organisms is linked to the planulation and synchronized reproduction generated by circadian stimuli, including particular moon phases (Sorek and Levy 2014). Laboratory studies suggest that night light pollution has negative implications, preventing synchronization in their reproductive timing, including tidal (12.4 h), lunarian (24.8 h), semilunar (14.77 days), or lunar (29.53 days) patterns in marine organisms (Naylor 1999; Kaiser and Heckel 2012).

Despite growing evidence on the impacts of light pollution on different species, the areas exposed to ALAN are expected to increase in intensity and spatial extent in the coming years (Davies and Smyth 2018). Based on the current data, the International Hydrographic Organization has defined areas especially vulnerable to contamination, considering places where at least 10% of the reef area is exposed to light levels more than double the brightness of the natural night sky. These locations include the Gulf of California (21.6%), the Persian Gulf (20.4%), the Gulf of Thailand (24.2%), the Gulf of Aqaba/Eilat (18.1%), the Gulf of Oman (17.6%), the South Atlantic Ocean (14.4%), the Malacca Strait (10.0%), and the Singapore Strait (34.5%). However, it is still necessary to expand the coverage of these data since there is almost no information associated with this phenomenon in Colombia. Hence, starting the exploration at the local level shows a first brushstroke of the state of light pollution in marine areas of reef importance in Colombian territory.

Therefore, to assess the impact of ALAN on marine ecosystems, we focused on quantifying, in situ, the artificial light that penetrates the column water in the leeward and fringing reefs from San Andrés Island, Colombia, because the Seaflower Biosphere Reserve is of paramount importance in worldwide, for its role in the biodiversity of marine species and its potential as a marine corridor for Caribbean species and that recently it was impacted for Iota Huracan. The finding of this work suggests that there is a significant incidence of artificial light in (1) the high scattering from the island to the reefs and (2) the depth at which it was assessed, and (3) annually there is evidence of a constant increase in horizontal light. This furthers our knowledge of the natural light alterations in both terrestrial and marine ecosystems. Additionally, as a result of the growing interest in tourism in the Archipelago of San Andrés, Providencia, and Santa Catalina the demand for multiple urban services on the island has increased and, consequently, artificial light sources that support the nightly business routine. As an initiative, the information obtained becomes a fundamental resource when proposing new management and conservation plans that can be effectively maintained in the face of the unstoppable arrival of artificial light in the future of the human population.

To measure artificial and natural skylight levels, we took nocturnal data for twelve days, starting at 19:00 h until 5:00 h, every 5 min, from 9 to 20th December 2020. The sensors were installed with scuba diving equipment in 3 locations around San Andrés Island (Fig. 1), where tourism is the main diurnal and nocturnal activity. One sensor was located on the fringe reef near Johnny Cay (12°35′47″N–81°41′52″E) (Green), where the incidence of artificial light comes from the island’s downtown commercial center. The second sensor was installed near the area known as Acuario, or the Aquarium (12°32′47″N–81°41′23″E) (Red), where the incidence of horizontal light comes from the San Luis and Rocky Cay sectors. The last sensor was installed in the West View zone (12°30′49″N–81°43′54″W) (Blue), classified as a place with less artificial light with respect to the other sampled sites.

A map of San Andres Island presents the coral reef barrier from the northeast to the east. 3 locations marked feature Johnny Cay near the northeast, Acuario in the east, and West view in the southwest. On the right side, a map features the Pacific Ocean, Caribbean Sea, and Atlantic Ocean.

Measurements of moonlight intensity were conducted using a Sky Quality Meter (SQM-LU-DL, Unihedron) (Fig. 2), which is an electromagnetic device sensitive to the light coming from a night sky (denominated vertical light). The SQM-LU-DL can discriminate against light pollution (denominated horizontal light) because its autonomous data-logging allows it to register and analyze the intensity of the night sky for several days in real time.

2 photos of sky quality meters. 1, the instrument unihedron is written in the front. 2, features S Q M-L U-D L with a power button on the right.

According to the supplier’s specifications, the SQM-LU-DL works with an HWHM of angular sensitivity at around 10° and uncertainty of ±0.10 (msas) magnitudes per square arcseconds (mag/arcsec²), indicating that it is a logarithmic measurement. Therefore, significant changes in the brightness of the sky will be interpreted as small numerical values, i.e. a difference of 1 is defined to be a factor of (100) (1/5) in received photons (Unihedron). In terms of the brightness quality, a higher magnitude is indicated by a darker color (24 msas; Fig. 3), and the magnitude decreases as the color becomes brighter. Therefore, it is also possible to measure light levels in the transition between day and night (twilight), but data registered by SQM-LU-DL will be saturated with daylight. Data obtained were analyzed in OriginPro v.9.8 (OriginLab Corporation, USA). In R studio, a Shapiro test was carried out to determine the normality of the data and a Kruskal–Wallis test to see if there were significant differences between each sampled area.

A view of the night vector presents a darker moon in the top right and a brighter moon in the top left. At the bottom, a scale of numbers from 15 to 24 are marked from the left to right. In the upper right area clusters of stars are noted.

The sensors are designed to be used on land. Therefore, we generated a polycarbonate box for underwater measurements using O-rings to avoid water. Each sensor was installed at 11 m depth and fixed with two leads to avoid water motion destabilizing the system, which was oriented towards the zenith to carry out the vertical measurement (Fig. 4). To contrast our in-situ measurements at 11 m depth, we also obtained maps and graphs (Fig. 5) of light pollution at the terrestrial level from NOAA and Unihedron databases.

3 photos. Each features a polycarbonate enclosure with a lunar light sensor within and positioned in the deep waters at 3 locations.

6 light pollution contour maps. A to F, on a shaded scale of 15 to 24 m s a s, each map features a patch in the north, with approximately 17 m s a s. The patch varies from 2012 at 4 kilometers to 2021 at 11 kilometers.

On the other hand, with the natural disaster of Hurricane Iota in 2020, much of the coastal vegetation was affected with the most significant loss along the island’s west coast. The West View data (in this study) was compared with the light data previously collected by Chilma-Arias et al. (2021) in 2019 to corroborate whether the vegetation is a natural barrier to artificial light. Therefore, it is necessary to consider the following:

Records collected from the Unihedron and NOAA websites (Fig. 5) indicate that, visually, the amount of artificial light has increased on San Andrés Island in the last nine years (2012–2021). The heat map (Fig. 5) shows that the horizontal light coverage in the northeast area of the island for the year 2012 reached a distance of 4 km (yellow line). In this same place and by the year 2021 (Fig. 5), the artificial light has spread to a distance of 11 km. It is important to mention that the island of Johnny Cay is approximately 2 km from our sensor at the Spratt Bight beach and the commercial center of the island, and the barrier reef in the fringing reef is 3 km away from this area of the island. This same pattern of artificial light was evidenced in the area of Acuario, indicating that, by 2021, the scattering of anthropogenic light sources altered the quality of natural light received by both the fringing reef and that entering 11 km offshore. The horizontal light dispersion was lower than at the previous sites regarding West View. However, visually, there was a slight increase in it between the years 2012 and 2021. It was not possible to find numerical data to carry out the statistical analyses and thus observe if there were significant differences in the measurements made.

Plots of the data collected by the SQM sensors (Fig. 6) during the 12 days of monitoring show that, in West View (blue dots), the sensor recorded a maximum magnitude of 21 msas with the majority of values around 18 and 20 msas. The average was 14.97 msas for Johnny Cay, 18.22 msas for Acuario, and 18.70 msas for West View. In contrast, the data obtained from the Acuario sector (red dots), revealed that the highest concentration of readings was between 16 and 20 msas, which indicates that, visually, there are slight differences between West View and Acuario. The sensor installed in Johnny Cay (green points), the highest incidence of horizontal light coming from the island occurs in this site compared to the other sampling sites because the concentration of the data was between 18 and 14 msas. These results indicate that the leeward reef located in the West View area is where the least damage from artificial light occurs. The opposite occurs in Johnny Cay, since the presence of LED lights from commerce and tourism negatively impacts the quality of natural vertical light. The above is supported by a significant Kruskal–Wallis test (X2 = 642.56, df = 2 and a p-value < 2.2e−16).

A scatterplot of the M S A S versus time in days has the data for sky brightness between each zone. 3 shaded plots form multiple stacked vertical trends throughout the graph.

When comparing the data recorded between December 2019 and December 2020 (Fig. 7), at a depth of 14 m, there was a maximum magnitude of 22 msas compared to the magnitude recorded at 11 m, which was 21 msas. A Kruskal–Wallis test shows significant differences (X2 = 66.47, df = 1 and p-value = 3.54e−16), which could indicate that during the year and indeed with the loss of vegetation, the quality of vertical light on the island decreased by a factor of 1 (from 22 to 21 msas), noting that magnitudes per square arcsecond are a logarithmic measurement. Therefore, the reduction of 1 unit between 2019 and 2020 corresponds to a 20-fold increase in light intensity (see http://​www.​stjarnhimlen.​se/​comp/​radfaq.​html). It is necessary to highlight that the difference between magnitudes is possibly higher, because in December 2019, the data was recorded in the presence of a full moon, while in 2020, the data were recorded in a new moon, indicating that in 2020 the light quality can be less than 22 msas.

2 scatterplots of M S A S versus time. A, the plots follow multiple stacked vertical trends between 14 and 24 on the y-axis. B, the vertically stacked plots span between 4 and 22 on the y-axis.

Since the origin of life on the planet, living beings have responded to natural changes and extreme conditions in the history of the earth, which have directed their particular evolutionary trajectories and have allowed them to accumulate adaptations that, over millions of years, modeled organisms with incredible capacities to survive and exploit the resources of each particular niche. However, it is undeniable that the anthropogenic era has generated changes that are so rapid that they have not allowed many species to respond at a rate of adaptation as fast as the innovations that occur throughout the world (Allgeier et al. 2020). The influence of human activities on different natural fields and the understanding of the multiple ecological synergies have become more important day by day. The consequences of our actions on the environment is something largely unexplored, and today we observe new collateral effects resulting from changes in the landscape.

Reef organisms such as corals, fish, echinoderms, mollusks, and crustaceans have evolved to synchronize their reproductive with the lunar periodicity, which is by a set of genes called Cryptochromes (Cry) that is expressed by perceiving a particular specific wavelength and thus carrying out the release of gametes (Levy et al. 2007; Sweeney et al. 2011). However, the spectral dynamics change due to the increase and excessive use of artificial light in coastal and island cities. In this study, it was impossible to quantify the current state of the blue pulse present on the island of San Andrés since the SQM sensors do not quantify each wavelength separately. Nevertheless, we can analyze that there is a significant disruption in the natural light cycles, due to constant incidence of the anthropogenic light sources inside the sea. Therefore, it is possible that the resilience of the coral reef organisms can be highly affected by the unprecedented nature of this pressure (Swaddle et al. 2015).

The levels of understanding of how human actions affect the planet provide tools to implement effective responses. Our findings helped unveil the state of the intervention of artificial light in the marine ecosystems of the Seaflower Biosphere Reserve in points close to urban centers, offering new data on an unexplored threat to marine ecosystems in Colombia. Figure 5 illustrates the patterns of light increase on the island of San Andrés from 2012. We can see a growth in the coverage area of artificial light in the northeastern zone, where the range of extension of the light passed from 4 km in 2012 to 11 km in 2021. In proportion, these phenomena are similar to what occurs in areas such as the Gulf of Eliat in the Red Sea, where the coverage of the artificial light that affects it is 47% brighter than a natural night sky and rises to a maximum of 60 times brighter than starlight on the north shore (reef mean 470%) (Ayalon et al. 2021).

According to the Bortle scale (Sky and Telescope 2020), a class 1 in sky brightness is an excellent dark-sky site, and a class 9 in sky brightness is an inner-city sky. Therefore, the Johnny Cay sector can be categorized in class 5, defined as a suburban sky (magnitude value between 14.5 and 15 msas). At the same time, Acuario and West View probably belong to class 2, equivalent to the typical genuinely dark site. However, it is essential to highlight that this classification system was developed for terrestrial data. Therefore, the underwater data taken here should be adjusted to terrestrial data to observe the change in magnitude from the surface to 11 m depth. There was not an appropriate number of sensors to collect data at the two depths for the three sampled sites in this investigation. Additionally, here we make a possible approximation to the Bortle classification system. Although the Unihedron maps (Fig. 5) are updated, certain liminal information is part of the atlas presented by Falchi et al. (2016).

When exploring the local level, the significant changes in the island’s light conditions are in the area where trade is centered (North Zone). The less light-saturated part is in the South Zone, as shown in Fig. 6. The highest light records correspond to Johnny Cay, followed by Acuario, where there is also a tendency for the presence of two peaks of light during a 24-h period (morning and night). The Acuario area presents considerable disturbances but shows maximum darkness values similar to the places less affected by light pollution, which corresponds to West View. Despite this, differences higher than 1 arcsecond are present in the data from the three sites (Shapiro–Wilk > 0.05), with the West View being the place with ideal natural light conditions. These observations generate new hypotheses that must be explored to evaluate the problem at the local level. For example, it would be expected that the processes of coral reproduction would be somehow affected by lunar synchrony, so following the moments in which the reproductive event occurs in the areas where light records were taken, the connection between the problem of artificial light and its influence on corals could be quantified.

The temporary nature of the records made it possible to analyze the consequences of the Iota climate phenomenon, which still has many questions about how it affected marine ecosystems. The data during 2019 showed significant differences compared to those taken in 2020 in the West View area. Figure 7 focuses on the site where less intervention of artificial light had been perceived. We can see how, after the hurricane, a group of data with values less than 10 arcseconds begins to appear, leading to the conclusion that after the hurricane, the penetration of light was greater. One of the explanations for these differences lies in the substantial loss of vegetation found between the sea and the closest sources of light. This strip of trees, no greater than 25 m, served as a natural barrier to artificial light, which was reflected in higher darkness values at the beginning of the study. Given the rapid growth of many developing world economies, future increases are expected to be greater in these regions as compared with the developed world over the coming decades, with unknown consequences for some of the planet’s most biodiverse marine ecosystems (Aubrecht et al. 2008), as is the case of the Seaflower Biosphere Reserve.

An additional point to consider is how artificial light affects different members of coral communities. Previous research shows how many reef diseases are generated by the instability of their microbial communities, allowing opportunistic microorganisms to colonize the tissue and cause negative consequences (Boilard et al. 2020; MacKnight et al. 2021). However, the inner relationship between corals and their symbionts is fragile and increasingly threatened by anthropogenic stressors (Hoegh-Guldberg 2014). Artificial light induces photoinhibition of the symbionts, overproduction of reactive oxygen species (ROS), and increased oxidative damage to lipids in coral species (Levy et al. 2020). Similarly, the modification of the symbiont community by artificial light intervenes at the time of reproductive processes (Tamir et al. 2020). This is causing a potential loss of ecological barriers in coral reefs that spawn in similar time frames.

Under continuous light conditions, oxidative stress occurs accompanied by a reduction in the ability to perform photosynthesis, both in the coral and its community of microorganisms. This result indicates the potential danger of artificial light for corals’ adaptation to human pressure because of initial characteristics in coral bleaching processes (Suggett and Smith 2020). These observations show that there is indeed an interruption in the coral-symbiont symbiosis, mainly because the excess of light does not allow a complete cycle of the dark part of photosynthesis, where many endosymbionts complete cell recovery processes throughout the night (Hill et al. 2011).

Despite the small number of instruments deployed to measure light and the short sampling window, the results are conclusive in revealing the threat of artificial light in different locations to reef organisms in different locations on the island. As such, the contribution is of value and even provides valuable artificial light management recommendations that should be relatively easy to implement. This study is just an approximation regarding the new techniques that can be useful at the local level, taking data with a higher level of precision is recommended to access more robust conclusions about how we should respond to the new collateral challenges of population growth. Among the recommendations to address the problem of imminent population growth is an appropriate use of lighting resources that allow it to be in harmony with the marine environments, minimizing the effects of artificial light and given that reproductive cycles occur on specific days and times. An alternative could be based on reducing light levels by a certain percentage on spawning days. In this way, the island’s tourist activities will not be affected, and marine organisms’ dynamics will be less altered. The use of light sources based on high-pressure sodium or fluorescent lights with a wavelength that does not have the same effects as LEDs is one possible solution that can also be explored. Additionally, lower-intensity LED lights could specifically reduce the emission of blue light peaks. Finally, although the island’s vegetation has been recovering naturally (a slow succession process), it is pertinent to carry out reforestation programs in the coastal areas and thus recover this natural barrier, which reduces the interaction between horizontal light and vertical light.