Testing the efficiency of suction samplers (G-vacs) on spiders: the effect of increasing nozzle size and suction time

The efficiency of two G-vacs (Ryobi RSV 3100 and Flymo BVL 320) was tested to examine the effects of increasing the suction nozzle area and the sampling time whilst keeping total sampling area constant (0.49 m). We tested an a priori hypothesis that increasing nozzle area was significant but that reducing sampling time was not, using a planned comparisons approach. We found that when the nozzle diameter was doubled in size, significantly fewer species and individuals of spiders and numbers of Pachygnatha degeeri, Centromerita concinna and Lepthyphantes tenuis were collected. However, the effect of increasing the suction time tenfold did not significantly increase numbers collected. This indicates that a significant proportion of the catch is collected in the first second of sampling. We conclude that in studies of short grasslands, small differences in suction time are unlikely to introduce confounding effects of under-sampling. Increasing the nozzle area may have serious and unwanted effects. G-vac users should be encouraged to give as much detail about their machines and their sampling method as possible and should see that reporting the nozzle size is paramount.