Scientist, Author, Educator, Tree Canopy Biologist

The Conservation Value of Ethiopian Church Forests


Dr. Neville Winchester, Department of Geography, University of Victoria, British Columbia, Canada V8W 3N5
Alameyuhu Wassie Eshete, Forestry Department, Bahir Dar, Ethiopia
Meg Lowman, PAMS, North Carolina State University, USA


Ethiopia is experiencing continual deforestation, and could lose its remaining afro-montane forests within the foreseeable future. In the northern highlands of Ethiopia, remnants of old-aged afro-montane forests can be found associated with Ethiopian Orthodox Tewahido Churches (Alemayehu 2010). Rising population pressures and deforestation mainly due to agricultural practices are the primary stressors impacting the rich fauna and flora of the region. Loss of these forests is the greatest threat to the biological diversity of the region and small remnant fragments (e.g., 1-100 ha) of these afro-montane forests exist but will likely lose much of their biodiversity as the surrounding habitat matrix continues to be eroded. We know that in general, species loss is accelerated in small (e.g. <100ha) and isolated fragments (Turner and Corlett 1996). Local extinction related to habitat fragmentation has been well documented (see Biological Dynamics of Forest Fragments Project, Manaus, Brazil) and compelling evidence suggests that many species of vertebrates cannot persist within small fragments but it is, as yet not clear what proportion of the original arthropod fauna is viable in small and often isolated fragments. Several factors influence arthropods that are likely to survive in a forest fragment. Size and connectivity are important, but frequency and intensity of human disturbance in the surrounding habitat matrix and encroachment are central drivers affecting biodiversity patterns and processes in these Church Forests.

Here, I report on a current biodiversity initiative and highlight preliminary results from a microarthropod program conducted in 2 Church forests. This research program, organized by Dr. Meg Lowman, was predicated on the partnership that she formed in 2009 with the Christian Orthodox clergy… the caretakers of these forests… and Dr. Alemayehu Wassie, a well known Ethiopian ecologist and point contact for this expedition. Information on other components of this pilot project and participants/research from the 2010 expedition can be found here.


Arthropod communities associated with suspended soils/epiphytes and ground debris in Ethiopian Church Forests were the target groups of interest. In August, 2010, I sampled 2 forest fragments, Debresena and Zhara. Debresna (11o, 51’N;37o,59’E) is a 11.5 hectare fragment containing 34 tree species at an elevation of 2690 metres. Zahar (11o 48’N;37o34’E) is lower (1950 m) in elevation and has less tree species (29). Although both fragments are small, they differ markedly in the degree of human impact within and outside of the forest fragment, Debresna represents a fragment that is significantly more intact with reduced human impacts when compared to Zahara (Figure 1). Additional details pertaining to Church Forests can be found in Alemayehu (2010).

I used single rope techniques to access the canopy and samples were collected with a soil corer, 3cm diameter x 5 cm deep (Figure 2). Contents of each sample core were extracted in the field. Microarthropods were extracted into 75% EtOH using Berlese funnels which were run for 48 hours. Since power was unavailable, samples were extracted using moth balls, a variation on the standard extraction technique but one that is proving to be valuable for areas where a power source is unavailable.

This study incorporates a stratified random sampling design where at each site the forest was divided into 3 strata based on distance to the edge. In each strata 3 trees were randomly chosen and single samples were taken from the ground and from the canopy (approximately 10 m from the ground). In total 18 samples were collected from each site and an additional 2 samples were collected from Zahar.

Extracted microarthropods were sorted into the following taxonomic groups: mites (Acari), springtails (Collembola), beetles (Coleoptera), flies (Diptera), bees, wasps and ants (Hymenoptera), pseudoscorpions (Pseudoscorpiones, millipedes and centipedes (Myriopoda), spiders (Araneae) and snails (Gastropoda) (Figure 4). The Acari were further identified to suborder and adult oribatid mites were identified to species by Dr. Valerie Behan-Pelletier. All specimens are deposited in the Canadian National Collection, Ottawa, Canada.

Independent sample t-tests were used to test for differences in oribatid richness and abundance between sites (canopy and ground samples were pooled). Using F-tests I compared the effect of distance on oribatid species richness at each site. Theoretical total species richness for each site was calculated using first- and second-order Jackknife, ACE, ICE and Chao 1 and 2 estimators. All estimates were performed using EstimateS (Colwell 2005).


From 6472 microarthropods collected from Berlese extractions a total of 51 species representing 30 families were identified from 1043 adult oribatid mites (741 from Debresna, 302 from Zahara). Observed and theoretical species richness using first- and second-order Jackknife, ACE, ICE and Chao 1 and 2 estimates were consistently higher in Debresena compared to Zahar (Table 1).

Twenty-two species of oribatid mites were observed from Zahar, of which 12 species were unique to this site and not found in Debresna. Thirty-six species of oribatid mites were observed from Debresna, of which 19 species were not found in Zahar, and the remaining 20 species of oribatid mites were shared in common between study sites. No significance difference was found between mean species richness (t17, 0.05 = 1.25 P = 0.229) and mean species abundance (t17, 0.05 = 1.62, P = 0.62) between sites. I detected a significant effect of distance from the edge for species richness at both sites (Zahar, F3,9,0.05 = 9.12; P = 0.004; Debresna, F3,8,0.05 = 5.74; P = 0.022) and a Tukey multiple comparison test indicated that the difference in both cases was between the samples from the forest fragment and the samples collected outside of the forest fragment. Distance from edge within the forest fragment showed little association with species composition.


Trends found in oribatid mite communities recorded in this study likely act as a surrogate for all microarthropods that require decomposing litter as a habitat template to complete their life histories. I detected no difference in species richness with distance from the edge within each forest site which likely is related to fragment size… these fragments were just too small to observe an effect of distance on diversity. However, distinct differences in species richness were found when samples from each fragment strata were compared to samples taken outside of the forest fragment. In soil systems, increasing mite diversity with increasing soil microhabitat complexity is well known (Anderson 1977). The substrate outside of the forest fragment is completed degraded and consists of compact clay with minimal interstitial spaces and little debris accumulation. Heavy impacts from grazing and general human use are evident and particularly pronounced in Zahar. Habitat heterogeneity and complexity is one explanation for the observations that we recorded in this pilot study, a feature that has conclusively been shown to shape oribatid mite communities (Lindo and Winchester 2006). The ability of forests to provide microhabitats which have a positive association with increased oribatid mite diversity and in general microarthropod diversity has been documented (Lindo and Winchester 2007).

Species richness estimators support the observation that our sampling program, although conservative in nature was sufficient to account for oribatid mite diversity contained in these forest fragments. However, I treat this result with caution and coax it in the context of spatial (fragment size) and temporal (seasonal) constraints. Interestingly, each site had a large number of rare species and species that were site specific. This supports the assertion that forest fragments act as repositories for arthropod biodiversity which is likely linked, among other things, to plant diversity. In addition, I suspect that this trend becomes more pronounced with increasing fragment size. The number of unique forest fragment species suggests that the oribatid mite community associated with these forests is distinctly different from the surrounding disturbed habitat and varies across fragments. I suspect that comparisons of oribatid mite richness between the canopy and ground would be significant but would require an extensive canopy sampling program that includes sampling at heights greater than 10 metres (e.g., 20-45m)… graduate student possibilities!!

In conclusion, well-developed soils still exist in the Church forests of Ethiopia and support a rich, diverse microarthropod community as exemplified by the oribatid mites documented in this study. Forest fragments likely vary in their species composition, further supporting the role that Ethiopia’s forests have in maintaining local biodiversity. Conservation initiatives encompassing all forest fragments are needed to maintain the unique and spectacular diversity that is intimately associated with Ethiopian afro-montane forests.

Literature cited

Alemayehu, W.E., Sterck, F.J. and Bongers, F. 2010. Species and structural diversity of church forests in a fragmented Ethiopian Highland landscape. Journal of Vegetation Science 21 (5), pp. 938 – 948.

Anderson, J.M., 1977. The organization of soil animal communities. Ecology Bulletin 25, pp. 15-23.

Colwell, R.K., 2005. Estimate S: Statistical estimation of species richness and shared species from samples. Version 7.5. Persistent URL <>.

Lindo, Z. and Winchester N.N. 2006. A comparison of microarthropod assemblages with emphasis on oribatid mites in canopy suspended soils and forest floors with ancient western redcedar trees. Pedobiologia 50, pp. 31-41.

Lindo, Z. and Winchester N.N. 2007. Local-regional boundary shifts in oribatid mite (Acari: Oribatida) communities: species-area relationships in arboreal habitat islands of a coastal temperate rain forest, Vancouver Island, Canada. Journal of Biogeography 34, pp. 1611-1621.

Turner, I.M. and Corlett, R.T. 1996. The conservation value of small, isolated fragments of lowland tropical rain forest. Trends in Ecology and Evolution Vol. 11 (8), pp. 330-333.

Figure 1. Ethiopian Church forest, Debresna study site. Sample trees and distance from the edge are highlighted and indicate position where Berlese samples were collected.

Figure 1. Ethiopian Church forest, Debresna study site. Sample trees and distance from the edge are highlighted and indicate position where Berlese samples were collected.

Figure 2. Neville Winchester, sampling the suspended soils in Zahar.

Figure 2. Neville Winchester, sampling the suspended soils in Zahar.

Table 1. Observed and theoretical total oribatid mite species richness for Zahar and Debresna study sites.
Zahar Debresena
Observed 31 37
ICE 36.56 (± 3.31) 43.47 (± 12.74)
Chao 1 31.24 (± 0.79) 37.61 (± 2.05)
Chao 2 31.06 (± 1.03) 38.14 (± 2.75)
Jackknife 1st order 35.44 (± 4.44) 37.88 (± 0.88)
Jackknife 2nd order 34.96 (± 4.29) 34.5 (± 12.64)

Connect with CanopyMeg