The effect of patch disturbance on stream invertebrate community structure: the influence of disturbance history.

The effect of disturbance history on the recovery of benthic invertebrate communities following disturbance was investigated in four streams in the Southern Alps of New Zealand. Two of the streams had a history of fluctuating discharge and temperature while the others did not. Recovery from disturbance was tested experimentally using baskets of cobbles, a third of which were disturbed every week for 9 weeks, a further third every 3 weeks and the final third left undisturbed. Algal biiomass, number of invertebrate taxa and total number of invertebrates all declined in baskets disturbed more frequently. Although the relative abundance of some taxa declined with time since the last disturbance, no taxa showed a significant decline in absolute abundance. However, several taxa showed marked increases in relative abundance in the less disturbed treatments particularly at the more stable sites. In contrast to the predictions of ecological theory, numbers of taxa and total invertebrates appeared to recover more quickly in the more complex communities at the stable sites. However, if these communities are considered to represent only stable communities, they do support the view that more complex communities will be more resilient. Community structure at the stable sites was also more similar between baskets in the undisturbed treatment than at the unstable sites, suggesting communities had reached a constant state more quickly. The more rapid recovery of communities measured at the stable sites may have been a consequence of experimental scale; disturbed patches were only 0.045 m2 in area and the higher densities of invertebrates at the stable sites meant a larger pool of colonists was available following each experimental disturbance. Nevertheless, ideas of stability in ecological theory and the scale of most spate events suggest this is the appropriate scale for examining community recovery. Furthermore, the larger pool of available colonists could not explain all the differences in community response, as patterns of change in community structure at the stable sites differed considerably more from those expected by purely random colonisation processes than at the unstable sites.