Title: Comparative assessment of adult dragonfly indices for the evaluation of the health of artificial wetlands in South Africa

Author: Philippe Kwitonda

Supervising Institution: IHE Delft  - Institute for Water Education

Year: 2018



Wetlands provide a variety of ecosystem services to human well-being. However, they are threatened by anthropogenic pressures and there is a need to assess them, determine their health status as first step towards their restoration and conservation. There are three commonly used adult dragonfly indices in assessing wetland and river health: These indices were developed differently and in different geographic areas: (1) Odonata Index of Wetland Integrity (OIWI) was developed in USA to assess wetlands and uses coefficient of conservatism (CoC) assigned to species by looking at their affinity with degradation status of wetlands; (2) Dragonfly Association Index (DAI) was developed in Austria to assess lowland rivers and assigns ecological statuses to rivers based on deviation between reference condition species and status quo; (3) Dragonfly Biotic Index (DBI) was developed in South Africa and measures the health condition of rivers and wetlands by using DBI score allocated to species by taking into consideration their sensitivity, their geographic distribution and their threat status. This study was conducted in South Africa to carry out a comparative assessment of those three commonly used adult dragonfly indices for assessing the ecological integrity of artificial wetlands and come up with possible recommendations for their betterment. A sample dataset of 68 artificial wetlands and 118 dragonfly species from across South Africa was used to develop the adult dragonfly indices. Additionally, land cover in 300 meter buffer around each artificial wetland was surveyed for determination of wetland degradation status which helped to classify the wetlands and then generate species CoC which is required to get OIWI. Among dragonfly species, Trithemis arteriosa had the higest CoC of 6.10 while Orthetrum caffrum had the lowest of 4.63. Using its building blocks, DBI score was determined as well. Pseudagrion furcigerum had the highest DBI score of 7 while species like Crocothemis erythraea, Ishnura senegalensis and Trithemis arteriosa among others have BDI score of zero. The adult dragonfly indices were tested at thirty artificial wetlands in Stellenbosch, the Western Cape Province. By using Wilcoxon rank sum test, this study found that although DBI and OIWI provided different ecological integrity information due possibly to their different building frameworks, they were both sensitive to alteration of artificial wetlands (OIWI, n = 30, p = 0.005; DBI, n = 30, p = 0.004). Spearman correlation test revealed that DBI and OIWI showed strong negative correlation (r = -0.72, n = 13, p = 0.005) especially at natural sites where there were more stenotopic dragonfly species. OIWI had very strong negative correlation (r = -0.85, n = 13, p < 0.001) with species richness at natural sites while DBI was strongly and positively correlated (r = 0.63, n = 13, p = 0.02) with species richness at natural sites. At disturbed sites, the correlation between DBI and species richness was very weak and not significant (r = 0.19, n = 17, p = 0.49) possibly because of a high occurrence of eurytopic species at disturbed sites. Compared to species richness and OIWI, DBI seems to be more robust as it provides additional information about species sensitivity. For the first application, the adult dragonfly indices are laborious and their transfer from one region to another is not straightforward since they are biogeographic-specific and adjusting CoC and DBI scores may be required. This study recommends to DBI to use sum of scores at sites where there are rare or/and critically endangered species and to OIWI to carry out informative assessment about species sensitivities before application. Due to insufficiency of data, this study didn’t develop fully DAI.