Abstract: Cereus peruvianus (Peruvian apple cactus) is a large erect and thorny succulent cactus characterized by column-like (cereus [L]: column), that is, candle-shaped, appendages. For three successive years (1100 days), between early April and late November, we studied the flowering patterns of eight cacti growing in public gardens and rural areas of north and central Tunisia, far from nighttime artificial illumination, in relation to natural environmental light, temperature, relative humidity and precipitation parameters. Flower blooming was assessed nightly between 23:00 h and until at least 02:00 h, and additionally around-the-clock at ~1 h intervals for 30 consecutive days during the late summer of each year of study to quantify both nyctohemeral (day-night) and lunar patterns. During the summer months of prolonged daytime photoperiod, flower blooming of C. peruvianus exhibited predictable-in-time variation as “waves” with average period of 29.5 days synchronized by the light of the full moon. The large-sized flower (~16 cm diameter) opens almost exclusively at night, between sunset and sunrise, as a 24 h rhythm during a specific 3-4-day span of the lunar cycle (full moon), with a strong correlation between moon phase and number and proportion of flowers in bloom (ranging from r = +0.59 to +0.91). Black, blue and red cotton sheets were used to filter specific spectral bands of nighttime moonlight from illuminating randomly selected plant appendages as a means to test the hypothesis of a “gating” 24 h rhythm phenomenon of photoreceptors at the bud level. Relative to control conditions (no light filtering), black sheet covering inhibited flower bud induction by 87.5%, red sheet covering by 46.6% and blue sheet covering by 34%, and the respective inhibiting effects on number of flowers in bloom were essentially 100%, ~81% and ~44%. C. peruvianus is a unique example of a terrestrial plant that exhibits a circadian flowering rhythm (peak ~00:00 h) “gated” by 24 h, lunar 29.5-day (bright light of full moon) and annual 365.25-day (prolonged summertime day length) environmental photoperiod cycles.

Abstract: Many broadcast spawning corals in multiple reef regions release their gametes with incredible temporal precision just once per year, using the lunar cycle to set the night of spawning. Moonlight, rather than tides or other lunar-regulated processes, is thought to be the proximate factor responsible for linking the night of spawning to the phase of the Moon. We compared patterns of gene expression among colonies of the broadcast spawning coral Acropora millepora at different phases of the lunar cycle, and when they were maintained under one of three experimentally simulated lunar lighting treatments: i) lunar lighting conditions matching those on the reef, or lunar patterns mimicking either ii) constant full Moon conditions, or iii) constant new Moon conditions. Normal lunar illumination was found to shift both the level and timing of clock gene transcription cycles between new and full moons, with the peak hour of expression for a number of genes occurring earlier in the evening under a new Moon when compared to a full Moon. When the normal lunar cycle is replaced with nighttime patterns equivalent to either a full Moon or a new Moon every evening, the normal monthlong changes in the level of expression are destroyed for most genes. In combination, these results indicate that daily changes in moonlight that occur over the lunar cycle are essential for maintaining normal lunar periodicity of clock gene transcription, and this may play a role in regulating spawn timing. These data also show that low levels of light pollution may have an impact on coral biological clocks.