Most methods for assessing reef fish assemblages at night require artificial light, but the use of different colors of light may influence the results. We used data from 135 underwater visual censuses (UVCs) performed with different colors of light (red, blue and white) to evaluate the structure of fish assemblages on subtropical rocky reefs along three depth intervals. We did not detect any effect of the color of light on total density or fish species richness per transect, nor on the structure of the entire assemblage. However, the density of some of the most abundant species varied according to the color used. Red light showed the highest values of frequency of occurrence for most species, while the white light resulted in decreased abundance of some fish species. Our results emphasize the importance of choosing the color of light depending on the type of studies to be conducted. This will depend on the objectives of the research (e.g. inventory, behavior or community dynamics) and the target fish fauna (e.g. mobile or sedentary).

大多数在夜间评估珊瑚鱼种类的方法都需要人工光，但是使用不同颜色的光可能会影响结果。我们使用来自135个水下视觉普查（UVC）的数据，以不同的颜色（红色，蓝色和白色）进行了评估，以评估沿三个深度间隔在亚热带岩石礁上鱼群的结构。我们没有发现光色对每个样条线的总密度或鱼类物种丰富度的任何影响，也没有检测到整个组合的结构。但是，某些最丰富物种的密度根据所用颜色而变化。红光显示大多数物种的出现频率最高值，而白光则导致某些鱼类的丰度降低。我们的结果强调了根据进行的研究类型选择光色的重要性。这将取决于研究的目标（例如清单，行为或社区动态）和目标鱼类区系（例如流动的或久坐的）。