The impacts of artificial light at night on the ecology of temperate and tropical reefs

Primary production is a core process underpinning the functioning of coral [14] and temperate reefs [15]. Impacts to primary production can result in the gain or loss of primary habitat, altered production and assimilation of biomass, and transfer of nutrients to higher trophic levels within the reef system [14]. The magnitude of primary production is largely driven by the light intensity in the environment [16], as increases in light irradiance on a reef can enhance the photosynthetic activity and growth of photoautotrophs (e.g. biofilms, micro and macroalgae, and corals and their endosymbiotic algae; [17–19]).

While little research to date has specifically assessed the impacts of ALAN on primary productivity on subtidal reefs, the addition of light to the reef environment will probably lead to increased primary productivity of the system. Studies on terrestrial plants confirm that relatively low levels of short durations of ALAN can effectively influence the growth and reproduction of terrestrial autotrophs [20]. Similarly, ALAN from white LED lighting has been shown to significantly increase maximum photosynthetic efficiency and photosynthetic biomass of epilithic microphytobenthos on rocky intertidal shorelines [21], and increase biomass and alter assemblage diversity of phytoplankton communities [22]. Therefore, by increasing the availability of light for photosynthesis, ALAN probably enhances primary productivity on coral and temperate reefs. Increased growth and productivity have bottom-up effects on the system, as more nutrients become available for higher trophic levels. However, the net impact of ALAN on primary production is not necessarily positive. ALAN can also reduce primary production on a reef indirectly through influencing top-down effects, i.e. increasing grazing pressure on primary producers. For example, in a temperate rocky intertidal food-web, Maggi & Benedetti-Cecchi [21] found increased consumption of epilithic microphytobenthos by herbivorous gastropods compensated for the positive effect of ALAN on primary productivity. Other studies, however, found no effects of ALAN on consumption rates of grazers, such as sea-urchins [23], highlighting the fact that responses to ALAN are species and context specific, and emphasizing the need to move beyond single organism studies in assessing the impacts of ALAN on trophic interactions, and community and ecosystem level dynamics.

Light is a critical environmental cue for the regulation of key hormone pathways responsible for the physiological functioning of an individual. In both vertebrates [24] and invertebrates [25], light is the primary environmental cue regulating the synthesis of melatonin, a key hormone involved in the regulation of circadian rhythms [24]. In a large range of species, the synthesis and release of melatonin occurs in darkness and is suppressed during daylight hours. Suppression or alteration of melatonin synthesis owing to changes in the night light environment can directly impact sleep/rest periods, and can have cascading effects on other hormonal pathways, including those affecting metabolism and energy expenditure, reproductive timing and fitness, stress, oxidative damage and survival in vertebrates [24], as well as moulting and metamorphosis in invertebrates [25].

On coral reefs, Scleractinian or hard, corals are critical to the structure and resilience of the reef ecosystem, and corals and their symbionts are highly photosensitive [26]. Corals are adapted to the natural light present in their environment, and coral growth, photosynthetic activity and health is dependent on both irradiance and wavelength of light (e.g. deeper corals exhibit higher photosynthetic rates in blue, short-wavelength light, compared to shallow-living corals which have higher photosynthetic rates in broad-spectrum lighting; [27]). The light and dark diel cycle is also critical for stress recovery and repair of coral photosynthetic symbionts [28,29], and therefore the presence of ALAN can erode these relationships. Indeed, ALAN has been found to elicit oxidative stress and reduce photosynthetic performance in Red Sea corals [28,30], with impacts found to be more extreme under shorter wavelength light (blue and white spectrum LED lights compared to yellow spectrum LED lights; [30]). ALAN can also disrupt the coral-dinoflagellate symbiosis [28,31], and alter pathways of gene expression, with approximately 25 times more differentially expressed genes that regulate cell cycle, cell proliferation, cell growth and protein synthesis under ALAN [26]. Furthermore, ALAN can affect gametogenesis and the timing of gamete release, impacting coral reproduction ([32]; see §2c Biological timings). Though research on the physiological impacts of ALAN on corals is only in its infancy, these studies demonstrate the potential for ALAN as a significant threat to coral fitness and survival, and consequently the resilience of coral reef ecosystems.

ALAN has also been shown to impact the physiology and fitness of reef fishes at various life stages, with research demonstrating significant effects of ALAN on metabolism [33,34], embryo quality [35], and post-settlement growth and survival [36,37]. As size is linked to probability of survival in the larval and post-settlement stages, with larger fish of the same age having significantly higher survival [38,39], impacts of ALAN on the size and growth in the early life stages of reef fishes may have carry-over effects for fitness and survival in later life-stages. For example, O'Connor et al. [36] found juvenile convict surgeonfish, Acanthurus triostegus, grew faster and were heavier under ALAN conditions, probably owing to increased time spent foraging under ALAN. However, the faster-growing fish under ALAN conditions had a significantly reduced probability of survival in the first 10 days post-settlement, in both the absence and presence of predators. By contrast, a long-term in situ study found long-term exposure to ALAN significantly reduced the growth of juvenile orange-fin anemonefish, Amphiprion chrysopterus [37], however, this study similarly found significantly reduced survival rates of the juvenile fish under ALAN conditions. Although the mechanisms behind the observed changes in growth under ALAN in juvenile reef fishes by both O'Connor et al. [36] and Schligler et al. [37] are unknown, the differences in findings are probably the result of many complex interacting physiological, behavioural and ecological factors. These studies both highlight the need for a greater mechanistic understanding of the impacts of ALAN on reef fishes and provide congruent evidence that the net result of ALAN exposure for juvenile coral reef fishes is decreased survival.

Many lower trophic level reef fishes and invertebrates (including the examples above), are sessile or highly site attached in juvenile and/or adult stages, and are therefore likely more susceptible to physiological impacts from ALAN as they cannot move away from artificial light in their environment. However, ALAN can also directly affect higher level consumers with greater mobility by influencing physiological responses, such as increased stress or higher metabolic rate, or indirectly through changes in resource availability and predation pressure (bottom-up and top-down effects, respectively) [20,40]. Although there has yet to be any research in the marine environment exploring how these direct impacts of ALAN on individual physiology can scale up to affect populations and/or reef communities, effects on individual growth, reproduction, and survival of organisms across trophic levels suggests that the impacts of ALAN on physiology could fundamentally alter the composition and functioning of reef communities.

One of the most significant ways in which ALAN can impact species on land and under water is by disrupting the timing of key biological events [3]. Many critical biological events in the ocean are highly attuned to daily, lunar, and seasonal changes in light produced by the reliable movements of the Earth and Moon around the Sun. Therefore, the addition of ALAN to an environment can mask or imitate natural light cues, resulting in an ALAN driven mismatch in life cycle timing.

Many mobile marine organisms, including marine larvae, are attracted to light [68]. Small fishes, invertebrates and/or plankton will move towards a light source, through positive phototaxis, disorientation or curiosity [69], and larger fishes and invertebrates from higher trophic levels are either attracted to the light itself, or the abundance of prey it attracts [70]. Many marine organisms also exhibit negative phototaxis, and will move away from a light source, if possible (e.g. [36,71]). For example, as both a releasing and directional stimulus ALAN can increase the abundance of individual amphipod species and alter amphipod community composition during the diurnal vertical migration of emergent zooplankton [72]. Amphipods are the dominant taxa within emergent zooplankton, which is an important food source for nocturnal planktivorous predators (e.g. corals, crinoids, fishes; [73]). The presence of ALAN is also known to attract both small shoaling fishes (less than 100 mm total length) and large-bodied predators (greater than 500 mm total length; [70]). This attraction and repulsion to light means ALAN can directly alter the composition of a nocturnal reef community, where the abundance of positively phototactic species will increase and negatively phototactic species will decrease, creating both unnatural bottom-up and top-down regulation of reef ecosystems.

Phototactic behaviour can be seen in marine organisms across life-stages. Many marine fishes and invertebrates have photo receptors and exhibit behavioural responses to light in the larval settlement stage. Therefore, ALAN has the potential to influence movement and settlement patterns of organisms that are sessile or site attached as adults, including habitat-forming species such as corals (e.g. [74,75]), oysters (e.g. [76]), barnacles (e.g. [77,78]), and polychaetes [79], as well as settlement stage larval reef fishes (e.g. [36,80]). For example, a manipulative experiment showed that ALAN significantly altered the composition of temperate epifaunal assemblages, with decreased abundances of the colonial ascidian Botrylloides leachii and the hydroid Plumularia setacea under ALAN treatments. By contrast, polychaete worms Spirobranchus lamarcki, the copepod Metis ignea and amphipods from the genus Corophium spp were more abundant on ALAN exposed habitats [81]. Furthermore, larvae of several coral species are known to respond to light intensity (Goniastrea aspera, Acropora tenuis, Ac. millepora, Oxypora lacera, Pseudodiploria strigose) and/or wavelength (Goniastrea favulus, Montipora peltiformis, Ps. strigose, Ac. millepora) during settlement [74,75], and the presence of ALAN has been shown to reduce settlement success of key habitat-forming species (e.g. corals, Stylophora pistillata [82]; Pacific oyster, Magallana gigas [76]; barnacles, Semibalanus balanoides [78], Notochthamalus scabrosus and Jehlius cirratus [77]). Interestingly, many coral reef fishes that are known to settle during the new moon (i.e. under darkness) have also shown positive phototactic behaviour in their larval settlement stage in response to intense light (e.g. light traps are used extensively for sampling reef fish larvae; [80]). This emphasizes the need for a greater understanding of how light intensity influences movement of settlement stage reef organisms, as the presence of ALAN could disrupt the natural light cues used to make informed decisions at settlement, fundamentally altering the habitat structure and community composition of both tropical and temperate reefs.

Temporal niche partitioning between competitors and between predators and their prey is a mechanism of coexistence in some ecological communities [84]. The addition of anthropogenic light to the nocturnal environment creates what has been termed a ‘night light niche'—when diurnal species remain active at night around artificial light (initially described by Garber [85]). This change in activity could result from direct impacts of ALAN on organism physiology (e.g. suppression of melatonin production), from ALAN masking the natural light cues that organisms normally respond to (i.e. darkness, or mimicking full moon), and/or a behavioural response to take advantage of new foraging opportunities. Regardless of the mechanisms, this exploitation of a new temporal niche can result in new predator–prey and competitive interactions between diurnal and nocturnal species that do not typically interact. Furthermore, this temporal niche shift can alter the strength of existing interactions, through the effects of ALAN on different species physiology and behaviour (e.g. increase efficiency of visual predators or decrease escape response of prey owing to stress), and/or the duration of interactions (and thus the outcomes), as ALAN extends the time that diurnal species remain active.

The night light niche could also provoke non-consumptive effects—changes in behaviour or physiology in prey species as a result of the perceived increased threat of predation under ALAN—i.e. changing the landscape (or seascape) for fear of nocturnal foragers. The presence of ALAN might be expected to influence the perceived predation risk for nocturnal foragers by affecting both their ability to detect approaching predators and the visual abilities of the predators themselves. This augmented landscape of fear can alter the behaviour of reef organisms in ways that we might not predict by considering the direct impacts of ALAN alone. For example, Manríquez et al. [86] found feeding rates of the nocturnal Chilean abalone, Concholepas concholepas, were three to four times lower when exposed to ALAN, and abalone were significantly more likely to be hiding in a dark refuge compared to control conditions. Therefore, while ALAN may not lead to increased predation of this species, enhanced vigilance (i.e. refuge-seeking) under ALAN reduces time allocated to foraging, and this may have long-term impacts on fitness and productivity.

The night light niche created by ALAN in shallow marine environments could therefore create novel ecosystem dynamics by (i) augmenting the amount of time visual predators/herbivores can spend foraging, increasing the duration of their activity and thus top-down pressure on the reef; (ii) creating novel interactions between diurnal and nocturnal organisms, thus altering competition and predator–prey relationships; and (iii) altering the landscape of fear for nocturnal foragers. These novel ecosystem dynamics have simultaneous costs and benefits for reef organisms, as ALAN can both increase growth and fitness of individuals if sufficient prey resources are available to support increased foraging activity, and negatively affect physiology, reproduction and survival of organisms through altered consumptive and non-consumptive effects. Therefore, ALAN and the night light niche can shift the balance of trade-offs for reef organisms, creating novel community and ecosystem dynamics.

ALAN could cause a collapse in species or functional diversity, by imposing selective pressures on species traits that limit fitness consequences for reproduction, recruitment and survival under ALAN conditions. For example, on coral reefs, through desynchronization of spawning, ALAN can alter the composition and function of coral communities. In a recent study that examined the timing of spawning in the five most abundant coral species in the Red Sea (Acropora eurystoma, Galaxea fascicularis, Platygyra lamellina, Dipsastraea favus and Dipsastraea favus), Schlesinger et al. [44] found three of the five coral species showed no consistent, synchronized spawning pattern relative to moon-phase, temperature or wind-speed—spawning had become out of sync. When coral recruits were assessed, despite high annual recruitment rates overall, no new recruits or juveniles from two of the three impacted coral species were detected in surveys, whereas the two coral genera with the highest number of recruits (Stylophora and Leptastrea) include species that are brooders and do not employ broadcast spawning (i.e. employ internal fertilization). Ultimately, a breakdown in spawning synchrony will reduce the probability of successful coral fertilization, resulting in a potentially insufficient supply of new recruits to sustain a stable population in the long term, and coral communities may shift towards dominance of species with life-history traits that make them less vulnerable to destabilization of environmental cues, such as brooding coral species.

While the authors hypothesized the desynchronization in this study is a result of changing temperature regimes, light pollution has the potential to comparably interfere with spawning patterns by masking the lunar cues used to synchronize mass spawning events. Furthermore, as different coral species possess different tolerances to changes in the light environment [27], ALAN can further shift the balance of ‘winners' and ‘losers' in coral communities through its direct influence on settlement patterns and impact on physiological stress and symbiotic relationships of some species, and indirectly through changes to competitive pressure for space (increased growth of turfing and macroalgae) and altered grazing pressure from primary consumers. This shift in coral community composition is likely to result in a loss of coral diversity and an overall reduction in coral cover. Loss of coral diversity and cover will have implications for the physical structure of the reef, which influences the composition, diversity and functioning of the reef community it can support. Changes in the physical structure of the reef can also directly impact ecosystem services—benefits provided to humans—through altering the capacity of reefs, for example, to protect coastal assets and support fisheries.

Since many marine organisms are broadcast spawners and/or undergo a pelagic larval dispersal phase where larvae disperse in the open ocean for several days, weeks or months before settling back to a reef [87], reduced reproductive fitness from ALAN exposure at one site may not directly affect the recruitment and population demographics at that same site. As a result, the presence of ALAN at one site has the potential to alter the supply of larvae to neighbouring reefs, sometimes 10–100s of km away, even if those downstream reefs are not exposed to ALAN themselves. Similarly, the presence of ALAN on a reef can influence whether incoming larvae from distant reefs will settle (e.g. [76–78,82]), as well as their probability of post-settlement survival and recruiting to the population (e.g. [36,37]). Therefore, through direct impacts to spawning and reproduction, settlement, and post-settlement survival, ALAN could create or alter the ‘source' and ‘sink' dynamics of connected populations. Altered source-sink dynamics could alter connectivity within a metapopulation, with potentially significant implications for the stability and resilience of local populations. Furthermore, if the impacted organisms are habitat-forming species or ecosystem engineers (e.g. corals, kelp, shellfish, barnacles), ALAN can fundamentally alter the structure of the reef and habitat availability, further impacting source–sink dynamics of the supported reef species. One way ALAN may lead to a population sink is through creation of an ecological trap—a habitat that animals prefer or equally prefer, but where their fitness is lower relative to other available habitats [88]. As many species of coral reef fishes are attracted to light, particularly in their larval settlement stage [80], reefs exposed to light pollution could experience an increased influx of settling larvae, which could contribute to reef resilience. However, as light at night can also reduce individual fitness of settlement stage larvae through physiological changes [36,37], enhanced predation risk [36] and increased top-down pressure from attraction of large predators [70], the attraction of larval reef fish to light-polluted habitats could be creating an ecological trap, significantly altering recruitment dynamics on tropical reefs and population persistence of vulnerable species.

The largest animal migration on the Earth is the diel vertical migration (DVM) of mesopelagic zooplankton from deeper waters during the day to near-surface waters during the night to feed [89]. Migrating zooplankton are followed by their predators, producing complex predator–prey migration patterns across multiple trophic levels [89]. This large-scale migration, which occurs in all oceans at all latitudes (e.g. [90]), plays a key role in the flow of energy and material between oceanic zones, and in biogeochemical cycles including carbon sequestration [91,92], and hence buffering global climate change [89]. DVM behaviour is generally controlled by ambient irradiance, with illumination from above inhibiting movement, and the setting sun, and thus a reduction in surface light levels, triggering an upward migration. The migrating zooplankton reach their shallowest distribution on the darkest nights of the lunar cycle, i.e. during the new moon, whereas the full moon is known to suppress the upward movement of all trophic levels [11]. Because of the nuanced relationships between ambient light levels in the ocean and the vertical movement of zooplankton and higher trophic level predators, ALAN can interfere with and disrupt the timing of the diel migration patterns of key organisms. Ludvigsen et al. [93] have shown that light from large vessels, or even from headlamps on a small open vessel, introduce enough light pollution to induce an avoidance response in the pelagic community down to more than 80 m depth. In addition to interfering with the initiation of zooplankton migration, ALAN can mediate changes in DVM through changing the behaviour of higher trophic organisms (i.e. increased/decreased consumption rates), through changes in species distributions (e.g. [72]), and through shifts in trophic dynamics. Impacts of ALAN on DVM will also probably affect carbon transport and biogeochemical cycles, with consequences to the functioning and connectivity of marine systems. As a result, disruption to DVM patterns from ALAN could have widespread consequences for the marine environment, potentially creating entirely novel ecosystems. However, attempts to quantify such impacts will need to consider sampling methods, as any nocturnal sampling method that introduces artificial light into the sampling environment will alter outcomes [93].

Increasingly, the marine environment is under threat from numerous stressors associated with human activities (e.g. chemical and noise pollution, ocean warming, acidification and de-oxygenation) [94]. These stressors more often than not co-occur with ALAN in space and time, making it challenging to assess the isolated impacts of ALAN in natural systems. Despite this, very few studies have actually considered the effects of ALAN alongside other co-occurring stressors in the marine environment [94–96]. Without empirical data, it is difficult to predict the combined effects of ALAN and multiple other stressors, as individual stressors may have species specific impacts, can impact different sensory systems (e.g. visual (light pollution), auditory (noise pollution) and olfactory (chemical pollution)) within individuals (e.g. [97]), and can have interactive, additive or antagonistic effects [98]. For example, observed impacts of ALAN on population dynamics and their consequences to ecosystem functioning are likely to be exacerbated with ocean warming, since thermal stress and ALAN similarly affect photosynthesis in primary producers [99], alter microbial communities [100] and alter growth and fitness of reef fishes (e.g. [101]), resulting in functional changes of systems [102]. This is probably true for corals, where increasing temperatures and ALAN similarly increase coral stress and mortality, and negatively influence the timing of gametogenic and spawning cycles [42–44], with potential for severe impacts to coral reproduction success and, consequently, overall diversity on a reef. However, the impacts of ALAN and ocean warming on sea urchins are probably antagonistic. Through grazing, sea urchins play a key role in the top-down control of macroalgal abundance on temperate reefs [103,104], and in high abundances, urchins can create barrens—large areas devoid of habitat-forming macroalgae [105,106]. While increased temperatures has been shown to increase urchin consumption rates [107], ALAN is predicted to decrease consumption rates of nocturnal urchins (e.g. through a seascape of fear). Likewise, these stressors can affect the kelp themselves, through e.g. changes in the kelp-associate microbiota [108] and their photosynthetic rates [109]. Such changes can not only have consequences to local primary productivity, but also influence the chemical defences of algae and their nutritional value, impacting interactions with herbivores, such as sea-urchins. At the community and/or ecosystem level, these interacting, additive or antagonistic stressors will probably create clear ‘winners' and ‘losers', which will ultimately determine the net effects of these stressors. More studies assessing combined effects of ALAN and other co-occurring stressors are therefore imperative to advance our knowledge and guide future management actions.

Our current understanding of how tropical and temperate reef organisms respond to ALAN is derived from short-term experimental or observational studies, spanning days to months in duration, and all within a single generation. The lack of long-term, multi-generational studies presents a significant gap in our understanding of how organisms may alter their behaviour, activity and/or physiology to maintain function under chronic ALAN exposure, and how ALAN might shift the selection pressure on certain phenotypic or genotypic traits. Though challenging, longer-term and multi-generational studies are necessary to determine whether phenotypic plasticity, parental effects or evolutionary compensation mechanisms allow marine populations to adapt to novel light environments. These questions can be addressed through laboratory-based multi-generational studies, as has been done for other environmental pollutants and climate stressors (e.g. [110–112]).

Interestingly, most of the research on the impacts of ALAN on subtidal reefs has been on tropical (e.g. corals and coral reef fishes) and subtropical (e.g. sea urchins) reef organisms, whereas many of the studies we have included in this synthesis on ALAN in temperate regions have focused on intertidal organisms (e.g. barnacles, shellfish and microphytobenthos). Intertidal communities experience very different environmental conditions than subtidal reef communities, including likely exposure to higher ALAN intensities; however, we have had to rely on these studies as research on the impacts of ALAN in subtidal temperate reefs is still minimal. This difference in research focus between tropical and temperate marine habitats is potentially owing to the fact that coral reefs are more likely to be vulnerable to impacts of light pollution. Water clarity on tropical reefs is typically much greater than in temperate marine habitats, and thus light will penetrate deeper through the water column and is more likely to reach the reef. This difference in light environments at different latitudes could potentially also result in a stronger reliance on nocturnal illumination as a cue for many critical biological events and processes on tropical reefs, whereas variation in moonlight in higher, temperate, latitudes may not be as important a cue as the more considerable changes in daylength or water temperature, both of which are more variable in temperate environments. Alternatively, the predictable but more annually variable lunar irradiance in temperate regions (e.g. [113]) means organisms can determine the specific time of year and month using lunar cues, and therefore temperate species may be equally or more sensitive to small changes in light environment as their tropical counterparts. This synthesis highlights the need for more research on temperate reefs to determine if and how the sensitivity of temperate reef organisms and trophic interactions to ALAN differ from those at lower latitudes.

This article has no additional data.