Supplementary material

Long-term UVB-exposure promotes predator inspection behaviour in a fish

Biology Letters

Simon Vitt*1, Janina E. Zierul1, Theo C. M. Bakker1 and Ingolf P. Rick1

1Institute for Evolutionary Biology and Ecology, University of Bonn, An der Immenburg 1,

D-53121 Bonn, Germany

*Corresponding author e-mail:

Experimental Subjects and holding conditions

Twenty-eight full-sib families of lab-bred sticklebacks were split into two groups of 20 individuals each (1120 sticklebacks in total) and transferred into 56 enclosures (39 × 28 × 28 cm, water level 20 cm), placed into four 2500 l outdoor tanks (14 enclosures / outdoor tank). From 07 September 2015 on, one group of each family was exposed to enhanced ambient UVB radiation for two hours from 11 a.m. to 1 p.m., provided by UVB lamps (G8T5E, 8W, Sankyo Denki, Japan) installed 10 cm above the water surface whereas for the other group opaque plastic strips were installed instead of the lamps at the same position to provide similar shade conditions. UVB-winter conditions, from 19 November 2015 to 14 April 2016, were created by reducing the irradiation time to one hour (11 a.m. to 12 midday). During the experimental phase, UVR was logged by a UV data logger (UVA / UVB, UV Microlog, Scitec Instruments, Bradford on Avon, GB). For the behavioural experiments, one randomly chosen stickleback out of each tank was used (N = 56) whereas remaining fish served for another study (S. Vitt et al., unpublished data).

Trout were kept in two large outdoor tanks (750 l) with air ventilation and a constant supply of tap water at a flow rate of 3 l min−1 and fed with commercial trout pellets six days a week.

Experimental setup

The experimental tank (100 cm x 45 cm x 35 cm) consisted of a predator compartment (25 cm x 45 cm) and a test fish compartment (75 cm x 45 cm). The predator compartment was separated by a fixed sheet of perforated clear plexiglas and a removable sheet of opaque plastic, respectively. The test fish compartment was divided by a removable, perforated clear plexiglas sheet into a start zone (15 cm x 45 cm) and the experimental compartment (60 cm x 45 cm). The inspection zone (20 cm x 45 cm) was located next to the predator compartment, within the experimental compartment. A plastic plant in the middle of the start zone served as a hiding place and a webcam (QuickCam Pro 9000, Logitech) attached to a laptop (ThinkPad T60, Lenovo) was mounted above the tank. The aquarium was illuminated by two fluorescent lamps (True Light, Natural Daylight 5500, 36 W, Aura Light) at 43 cm distance from the water surface. The whole set-up was surrounded by a black curtain and both removable sheets could be lifted from outside the curtain by a string, using a pulley system. Before each experiment, the aquarium was filled with one day-old tap water up to a level of 16.5 cm.

Irradiance Measurement in the field

Measurements in the field were made by using a spectrophotometer (AvaSpec 2048, Avantes, the Netherlands) equipped with a cosine corrector (CC-UV/VIS, Avantes, the Netherlands). Irradiance was calibrated against an Avantes NIST traceable application standard. For measurements, the irradiance probe was kept in a fixed position at 10 cm water depth. Fifteen single spectra were recorded within one measuring series using the software AvaSoft (Version 7.5, Avantes, the Netherlands) and transferred to Microsoft Excel (Microsoft Office 2007) to calculate a mean spectrum.

Linear mixed effects models

Table S1: All linear mixed effects models (lme) calculated. In the lmes regarding standard length (SL) either sex or treatment were taken out of the model because both significantly influenced the SL. Time until leaving the start zone (latency), absolute activity (activityabs), relative activity (activityrel), absolute inspection time (ITabs) and relative inspection time (ITrel) were used as dependent variables. Treatment, sex and SL were used as explanatory variables. Hierarchical random effects were used in each model by nesting family within outdoor tank as random factors.

dependent variables / explanatory variable / random factors
treatment / sex / SL
χ2 / ∆d.f. / P / χ2 / ∆d.f. / P / χ2 / ∆d.f. / P
SL / 22.162 / 1 / < 0.001 / 18.092 / 1 / < 0.001 / - / - / - / outdoor tank
family
latency / 2.495 / 1 / 0.114 / 1.761 / 1 / 0.185 / 0.130 / 1 / 0.718 / outdoor tank
family
activityabs / 0.025 / 1 / 0.875 / 2.054 / 1 / 0.152 / 0.002 / 1 / 0.969 / outdoor tank
family
activityrel / 0.223 / 1 / 0.637 / 2.095 / 1 / 0.148 / < 0.001 / 1 / 0.985 / outdoor tank
family
ITabs / 5.538 / 1 / 0.019 / 0.359 / 1 / 0.549 / 0.109 / 1 / 0.741 / outdoor tank
family
ITrel / 5.213 / 1 / 0.022 / 0.361 / 1 / 0.548 / 0.137 / 1 / 0.711 / outdoor tank
family