Partial night lighting may reduce the physiological impact of artificial light at night on captive zebra finches

Abstract

Background: Artificial light at night (ALAN) continues to increase at an unprecedented rate globally every year. ALAN can disrupt circadian rhythms and cause behavioural and physiological changes which may have knock on effects for health, yet we still understand very little about these effects. It is becoming increasingly important to investigate potential mitigation strategies, through understanding what aspects of ALAN negatively impact wildlife health.

Methods: Here we present the results of an experiment where we investigated the impact of ALAN on various health biomarkers in 44 captive adult zebra finches (Taeniopygia guttata) over the course of 4 months. The health biomarkers measured included glucose concentration, change in relative telomere length, malondialdehyde, and antioxidant capacity of plasma. The birds were separated into three treatment groups consisting of 13-16 individuals and were either exposed to full light at night, partial light at night or darkness (control).

Results: We show that exposure to full light at night impacted the circadian pattern of glucose levels, with glucose concentrations remaining elevated later into the night compared to the control group. Full light at night also accelerated telomere shortening. However, the relationship between telomere length and treatment only became significant when the partial light at night treatment group was removed and should therefore be interpreted with caution. These effects were not observed under partial light at night.

Conclusion: Our findings suggest that partial night lighting may mitigate some of ALAN's negative impacts on wildlife health. This approach could be a valuable tool in future strategies to reduce the ecological effects of light pollution in urban environments and should be investigated further.

Keywords: artificial light; avian health; avian physiology; light pollution; urban ecology.

FIGURE 1

The relationship between treatment group and glucose levels in zebra finches at various times of day. The DARK group are shown in blue, the PLAN group are shown in green, and the FLAN group are shown in yellow. The means are shown on the plot along with the standard errors. The raw data points can also be seen. Time of day is shown on the x-axis and glucose concentration is shown on the y-axis. The first panel of the plot represents the females, and the second panel represents the males. Sample sizes are as follows DARK Female 1 p.m. (3), DARK Male 1 p.m. (3), PLAN Female 1 p.m. (3), PLAN Male 1 p.m. (2), FLAN Female 1 p.m. (2), FLAN Male 1 p.m. (4), DARK Female 8 p.m. (3), DARK Male 8 p.m. (3), PLAN Female 8 p.m. (3), PLAN Male 8 p.m. (3), FLAN Female 8 p.m. (4), FLAN Male 8 p.m. (3), DARK Female 1 a.m. (3), DARK Male 1 a.m. (3), PLAN Female 1 a.m. (2), PLAN Male 1am (4), FLAN Female 1 a.m. (3), FLAN Male 1 a.m. (4), DARK Female 6 a.m. (3), DARK Male 6 a.m. (3), PLAN Female 6 a.m. (3), PLAN Male 6 a.m. (4), FLAN Female 6 a.m. (4), FLAN Male 6 a.m. (4). Tukey’s post hoc analysis revealed significant and marginally significant differences between the FLAN and DARK groups in males at 1 p.m. (P = 0.042) and at 1 a.m. (P = 0.062), as well as in females at 1 a.m. (P = 0.070).

FIGURE 2

Mean glucose concentrations along with standard errors are shown at four time points for each treatment group. Glucose concentration is shown on the y-axis and time of day is shown on the x-axis. The DARK group are shown in blue, the PLAN group are shown in green, and the FLAN group are shown in orange. Samples size are as follows: DARK (12), PLAN (15), FLAN (16).

FIGURE 3

The relationship between treatment groups and glucose levels in zebra finches from the start to the end of the experiment. The DARK group are shown in blue, the PLAN group are shown in green, and the FLAN group are shown in yellow. The panels represent the different sexes. The means are shown along with the standard errors, the raw data points are also shown. Experiment timepoint is shown on the x-axis and glucose concentration is shown on the y-axis. Samples sizes are as follows DARK Female (6), DARK Male (6), PLAN Female (6), PLAN Male (5), FLAN Female (6), FLAN Male (7). Tukey’s post hoc test revealed significant differences (when PLAN was removed from the model) between males and females in the DARK group at the end of the experiment (P = 0.049).

FIGURE 4

The relationship between treatment groups and telomere length in zebra finches at the start and end of the experiment. The DARK group are shown in blue, the PLAN group are shown in green, and the FLAN group are shown in yellow. The model residuals (without any plate effects) corrected by mean telomere length along with the standard errors, the raw data points are also shown. Experiment timepoint is shown on the x-axis and relative telomere length is shown on the y-axis. Sample sizes are DARK (12), PLAN (15), FLAN (16). LRT analysis revealed a significant interaction between treatment and experimental timepoint when the PLAN group was removed from the full model (P = 0.016).

FIGURE 5

The relationship between treatment groups and MDA in zebra finches at the start and end of the experiment. The DARK group are shown in blue, the PLAN group are shown in green, and the FLAN group are shown in yellow. The means are shown along with the standard errors, the raw data points are also shown. Experiment timepoint is shown on the x-axis and MDA is shown on the y-axis. The separate panels represent the different sexes. Sample sizes are as follows are DARK Female (6), DARK Male (5), PLAN Female (4), PLAN Male (5), FLAN Female (7), FLAN Male (7). LRT analysis revealed no significant effect of treatment on MDA levels although there was a significant effect of sex (P = 0.019).

FIGURE 6

The relationship between treatment groups and OXY in zebra finches at the start and end of the experiment. The DARK group are shown in blue, the PLAN group are shown in green, and the FLAN group are shown in yellow. The model residuals (without any plate effects) corrected for the mean OXY levels are shown the standard errors, the raw data points are also shown. Experiment timepoint is shown on the x-axis and OXY is shown on the y-axis. The separate panels represent the different sexes. Sample sizes are as follows DARK Female (6), DARK Male (6), PLAN Female (6), PLAN Male (5), FLAN Female (7), FLAN Male (7). No significant treatment effects were detected.