Retrofitting streetlights with LEDs has limited impacts on urban wildlife

Abstract

Artificial light at night (ALAN) causes a wide range of ecological impacts across diverse ecosystems. Most concentrated in urban areas, ALAN poses a particular risk to associated wildlife by disrupting physiology, behaviour and ultimately survival. This risk is predicted to shift as nighttime lightscapes in many cities undergo change. Globally, streetlights are currently being retrofitted with newer technologies that differ in the spectrum and intensity of their emissions, but there is a dearth of in situ urban experiments on the ecological impacts of this change. We monitored timing of dawn and dusk bird song; frequency of owl vocalisations; avian diversity, relative abundance and community composition; small invasive mammal and ground insect activity; and invertebrate relative abundance at 26 residential properties over an 18-month period that coincided with a retrofit from high-pressure sodium (HPS) to white light-emitting diode (LED) streetlights. Initiation time of dawn song was advanced or delayed for two bird species following the retrofit and backyard avian community composition was altered. Avian species richness, relative abundances of three bird species and ground insect activity increased in the presence of LED streetlights. No other retrofit effects were found. Our study suggests that retrofitting streetlights with white LEDs may lead to both positive and negative conservation outcomes for urban wildlife, but direct impacts are relatively small and may be mitigated by changes in lighting characteristics, such as dimming. Streetlight retrofits could provide an opportunity to reduce the impacts of ALAN on urban wildlife if intentionally designed with conservation benefits in mind.

Introduction

Artificial light at night (ALAN) is a global threat to biodiversity that is worsening with increasing urbanisation (Kyba et al., 2017). ALAN disrupts natural light cycles, impacting the physiology, behaviour and survival of numerous taxa across diverse ecosystems (Gaston et al., 2014; Gaston et al., 2017). Known impacts include the advancement of dawn song in birds (Da Silva and Kempenaers, 2017), the reduction of activity in small mammals (Farnworth et al., 2016; Spoelstra et al., 2015) and the attraction to light sources and subsequent death of phototaxic insects (Eisenbeis, 2006). These kinds of impacts on individual behaviour and physiology are widespread across taxa and regions, and can disrupt interspecific interactions and lead to the restructuring of entire communities (Davies et al., 2012; Owens and Lewis, 2018; Sanders et al., 2020). Conservation of biodiversity in and around cities therefore needs to account for the effects of ALAN, but limited understanding of knock-on effects makes prediction of community response to large-scale changes in the night-time lightscape difficult (Sanders and Gaston, 2018).

Artificial lightscapes in cities are currently undergoing large-scale change, with a global shift towards increased use of broad-spectrum white lights. This shift is particularly noticeable in city-wide streetlight retrofits, where old streetlight technology (such as high-pressure sodium [HPS] lamps) is replaced by more advanced, energy-efficient lamps such as white light-emitting diodes (LEDs; IPWEA, 2014; Mizon, 2012; Schubert and Kim, 2005). These new streetlights have different spectral signatures, with emission of wavelengths peaking in different parts of the spectrum (Elvidge et al., 2010). White LEDs emit a broad spectrum light that differs noticeably from the orange/amber coloured light of previous technologies such as HPS, and contains a greater proportion of blue light (Aubé et al., 2013; Luginbuhl et al., 2014). Broad-spectrum, blue-rich lights tend to have the most ecological impact, as the emission of a broad range of wavelengths is likely to trigger more wavelength-specific biological responses than lights with narrower spectra, while short wavelengths in particular affect wide ranges of physiological and biological processes (Davies et al., 2013; Gaston et al., 2013; Gaston et al., 2017). Consequently, widespread adoption of white >2000 K LEDs is expected to worsen the impact of ALAN on biodiversity (Falchi et al., 2016; Gaston et al., 2012; Longcore et al., 2018; Luginbuhl et al., 2014).

On the other hand, the newer LED technology also allows for dimming and immediate switch on of lights without a warm-up phase, as well as timed control via central management systems (Gaston et al., 2012; Kyba et al., 2012). Dimmed or low-intensity lights can reduce the trespass of light and aid in creating dark refugia amongst otherwise illuminated landscapes (Gaston et al., 2012), although very low light intensities can still have effects on wildlife (e.g. Dominoni et al., 2013; Dominoni et al., 2013a; Lewanzik and Voigt, 2014). LED luminaires often have more directed light pools, emitting less above the horizontal plane than previous streetlight designs, and limiting the extent of the area impacted by direct ALAN (Wakefield et al., 2018; but cf. Gaston et al., 2012). The net effect of white LED streetlights is therefore difficult to predict, as their light distribution and intensity are expected to have less impact on urban wildlife than older streetlights, contrasting the expected exacerbated impact of their spectrum.

Despite their potential to alter the existing ecological impact of ALAN, the consequences of streetlight retrofits are not well known (Stanley et al., 2015). Such consequences are potentially borne out differently dependent on taxa and scale, but to date most retrofit studies have only focused on bats. Changes between low-pressure sodium (LPS), mercury vapour, metal halide and LED streetlights have resulted in various responses of altered activity, abundance or diversity (Haddock et al., 2019; Lewanzik and Voight, 2017; Stone et al., 2015; Plummer et al., 2016; van Grunsven et al., 2019) or in no response at all (Rowse et al., 2016). Experiments comparing the impacts of HPS and LED streetlights have produced mixed results across a range of taxa, with LEDs being found to have both more (Grubisic et al., 2018; Pawson and Bader, 2014; Rodríguez et al., 2017) and less (Eisenbeis and Eick, 2011) of an effect on wildlife than HPS lights. However, most studies are either conducted in the lab, or in rural areas, and results are extrapolated to impacts on urban wildlife. To our knowledge, no-one has conducted an in situ urban experiment on an HPS–LED retrofit, so there is a limited understanding of how the wildlife impacts of the retrofit might play out in a real urban setting. Such understanding is critical for managing and conserving urban biodiversity exposed to the effects of ALAN.

In this study, we utilised an ongoing city-wide streetlight retrofit from HPS streetlights to LED lights in Auckland, New Zealand to investigate the response of a range of taxa to the change in ALAN. Specifically, we aimed to determine whether the change to the less intense, but more blue-rich LED streetlights altered: 1) timing of dawn and dusk bird song, 2) nocturnal vocalisations of an owl species, 3) avian community composition, 4) small invasive mammal and ground insect activity, and 5) invertebrate abundance. Due to the higher proportion of blue in the spectrum, we predicted that LED streetlights would worsen the impact of ALAN; namely that birds would sing earlier in the morning, later at night, and have altered frequency of nocturnal vocalisations, small invasive mammals would be detected less often due to increased light-avoidance, and conversely, positively phototaxic insects would show increased light-attraction and be more abundant under lights. It was hypothesised that individual-level effects could have knock-on effects on the abundance and diversity of bird species, thus altering avian community composition.