top of page


Temporal and spatial abundance and distribution data for individual butterfly species will be detailed for individual species. This was recorded from the sampling of species making up the fruit-feeding butterfly assemblage of Semuliki National Park, using two differing monitoring methodologies (long-term and snapshot). Each methodology was compared to evaluate its efficiency and effectiveness in capturing the totality of a species subset (mostly forest-restricted species) making up the key insect functional group - the fruit-feeding butterfly assemblage.  

All information has been distilled from field data collected for the PhD 'The ecology and conservation of a fruit-feeding butterfly assemblage in a
Ugandan lowland rainforest
'. A paper 'Evaluating survey methodology efficiency for an Afrotropical lowland forest fruit-feeding butterfly assemblage' was submitted to 'The Journal of Insect Conservation' which presented the key findings of the PhD. This paper was rejected, but in retrospect, with the flawed quality of this journal's editorial process as evidenced in a previously published paper where Bebearia laetitioides was described as a key indicator species for near primary-forest and in the same paper was also recorded as the most abundant species in a sample of both primary and secondary forest in the Cameroon, where it doesn't exist (the nearest distributional record being in the eastern DRC - around 2400 km away), I now accept as an honour and thank you!

Methodologies protocol

For both methodologies two transects were sampled at two locations (Kirumya trail (KR) and Red Monkey trail (RM)) which were a direct distance of around 7.5kms apart and around 1.5km from the park road boundary (to remove the influence of invasive and exotic species that were located within the park margins). Total length of both transects was 500m. All traps used were of the standard Van Someren‐Rydon design and fermented banana bait was replaced every 48 hours or in the dry season every 24 hours depending on bait state. All trapped butterflies were marked with a felt tip pen (small dot to margin of hindwing, colour dependent on species) and any re-trapped were excluded from dataset count.

Intensive monitoring methodology (Snapshot): Seasonal sampling (dry-Feb and wet-Oct) was undertaken twice in each season for a total of four sampling replicates. Sampling was undertaken on six consecutive days from ten traps, 50m apart at midstorey (around 50cm to 1m above forest floor).

Long-term monitoring methodology: Traps were sampled approximately every 14 days for two years. Five spatial sampling stations were created 100m apart. At each trap station three traps were laid - one on the forest floor; one in the understory and one in the canopy.

Overall general result summary

Intensive monitoring methodology - 480 trap days: sampling abundance 5181, species richness 64 (excluding Bicyclus) 87 including Bicyclus.

Long-term monitoring methodology - 1455 trap days: sampling abundance 8938, species richness 71 (excluding Bicyclus).

KR site - Total abundance combining methodologies: 5885 specimens sampled. Species richness 61 long-term monitoring (excluding Bicyclus). Species richness 73 for intensive monitoring (including Bicyclus), 51 excluding Bicyclus.

RM site - Total abundance combining methodologies: 8234 specimens sampled. Species richness 63 long-term monitoring (excluding Bicyclus). Species richness 72 for intensive monitoring (including Bicyclus), 50 excluding Bicyclus.

Individual Species

The following genera from the tribe Adoliadini (Bebearia, Euphaedra, Euriphene - Cymothoe (incertae sedis) contributing to the assemblage in both components of diversity and abundance and being important indicators of near primary forest were the focus of individual species data analysis. Species within these genera have been the subject of various publications and have become a popular choice as bioindicators for measuring the functional integrity; aspects of anthropogenic disturbance and edge effect in African tropical forests The sensitivity to habitat change seen in some species within the four stated genera (requiring a closed canopy with a shady understorey) - (Birket‐Smith, 1970, Larsen, 2005) make this group potentially excellent indicators for monitoring habitat change and with species richness within these genera can provide the relative health of the ecosystem being sampled.

Family Nymphalidae: Subfamily Limenitidinae: Tribe Adoliadini.

Genus Bebearia

14 Bebearia species were trapped over the long-term monitoring period, while 13 were trapped for the combined intensive periods. Bebearia plistonax although present was never trapped. At both sites 13 Bebearia species were trapped during the long-term monitoring programme this was in comparison to only eight species at KR and 11 species at RM from the combined intensive periods.

The two most abundant species sampled in the traps at both sites were from the genus Bebearia (B. laetitioides and B. brunhilda) and these two species comprised 38% (KT) and 29% (RM) trapped total for both sites and combining methodologies respectively. Bebearia laetitioides was nearly twice as common at both sites (31% KR and 15% RM) as the second most abundant species B. brunhilda (18% KR and 14% RM).


Bebearia cocalia.jpeg
Bebearia cocalia.jpeg
Bebearia brunhilda.jpeg
Bebearia laetitioides.jpeg

Bebearia laetitioides (female, upperside)

Bebearia brunhilda (female, upperside)

Bebearia cocalia (female and male, upperside)

Bebearia abesa.jpeg
Beberia abesa.jpeg
Bebearia absolon.jpeg
Bebearia zonara.jpeg

Bebearia abesa (male, upperside)

Bebearia abesa, B. absolon and B. zonara (females, upperside)

Bebearia cocalia was a significant component of total abundance for both methodologies at 4% for each methodology. However, this was a site-specific species: for example only two specimens were trapped at Kirumya during the intensive monitoring periods.


Many tropical butterfly species are rare and this was the case with the 12 other species. Bebearia plistonax was trapped only once and this was outwith the monitoring periods. Bebearia mandinga and Bebearia flaminia were trapped rarely and were very scarce within the park while all the remainder of Bebearia species recorded at Semuliki were trapped infrequently and were never abundant. For seasonality assemblage composition of Bebearia species for the two different methodologies and for both sites please see PhD.

The question of whether there is a bias towards the frequency of rare or intermediate species abundances and the models that have described species distribution abundances will obviously then influence the focus of any biodiversity research or targeted species conservation. Of course, describing global species abundance distributions to assess a species rarity or commonness is dependent on sampling intensity. As species sampling effort is increased through time, the underlying shape of individual species abundance distributions can only then be accurately described (Callaghan et al., 2023). However with butterfly species abundance and diversity being greatest in the tropics and with minimal long-term sampling of lowland forest species in Uganda, species abundances can't yet be estimated for changes and especially not the using of presence-only data of tropical insects as a proxy for abundance.


Five Bebearia species recorded from Semuliki National Park (Bebearia barce, B. plistonax, B. seeldrayersi, B. laetitioides and B. partita) have been categorised as Vulnerable (IUCN Red List categories and criteria, 2012) in a publication 'Nationally Threatened Species for Uganda' (Wildlife Conservation Society, 2016). Vulnerable being defined as the 'best available evidence' and indicating the categorisation meets some criteria suggesting a 'continuing decline, observed, inferred or projected can be met'. This categorisation for these five species is completely unjustified. There has been no widespread, long-term or standardised monitoring data available for each of these species (apart from the PhD sampling data) to suggest any species decline of either area of occupancy; area/extent of distributions and or decline in habitat quality of the Semuliki National Park habitat or a decline in the location number. There has also been no suggestion when categorising these species of addressing the difficulty of evaluating the natural short-term stochastic variation in population sizes (seen extensively in Bebearia laetitiodes) for these species and in some cases inter-site variation within a specific forest reserve or park. 

The occurrence and abundance data from the PhD, can only provide a baseline for any future comparative assessments of species of the fruit-feeding assemblage, either from Semuliki or from other lowland forest reserves or forest patches throughout Africa.  . 

Genus Euriphene

Three Euriphene species (Euriphene saphirina, E.atossa and E. ribensis) were trapped over the long-term monitoring period and three were also trapped for the combined intensive periods. Euriphene amaranta and E. doriclea have been recorded at Semuliki but were never trapped. At both sites three Euriphene species were trapped during the long-term monitoring programme, this was in comparison to only one species at KR (E. saphirina) and three species at RM from the combined intensive periods.







The three Euriphene species recorded during the two applied monitoring methods were never trapped in abundance  with a total of only 48 from the combined monitoring intensive periods and 172 during the long-term monitoring period.

One Euriphene species recorded from Semuliki National Park (Euriphene atossa) has been categorised as Vulnerable (IUCN Red List categories and criteria, 2012) in a publication 'Nationally Threatened Species for Uganda' (Wildlife Conservation Society, 2016), again without justification - see explanation above for Bebearia.

Euriphene atossa.jpeg

Euriphene atossa (male, underside)

Genus Euphaedra

13 Euphaedra species (Euphaedra alacris, E. albofasciata, E. caerulescens, E. eberti, E. edwardsii, E. eleus, E. harpalyce, E. diffusa, E. imitans, E. medon, E. rattrayi , E. ruspina and E. zadacchi) were trapped over the long-term monitoring period while 12 were trapped for the combined intensive periods. Four species sampled during the long-term monitoring programme were not trapped during the combined intensive periods (E. hybrida, E. eberti,  E.ruspina and E. zaddachi) while two species (E. preussi and E. viridicaerulea) were trapped during the intensive periods but not during the long-term monitoring.

Euphaedra medon was the most common Euphaedra species sampled. It was the third most abundant species sampled during the long-term monitoring programme after B. laetitioides and B. brunhilda. Euphaedra species richness between sites was similar for both monitoring methodologies. A few species were only sampled in one of the transect sites but with the rarity of many Euphaedra species encountered at Semuliki: E. imitans, E. viridicaerulea, E. zadacchi and E. ruspina for example, this was not a surprise. 

A number of Euphaedra species: E. christyi, E. procera and E. mirabalis for example were sampled from traps which were randomly located within the forest or during post-PhD period sampling.

One Euphaedra species recorded from Semuliki National Park; Euphaedra rattrayi  has been categorised as Vulnerable  (IUCN Red List categories and criteria, 2012) in a publication 'Nationally Threatened Species for Uganda' (Wildlife Conservation Society, 2016), this is again unjustifiable for reasons described above. While, apparently according to this publication, it is also not present at Semuliki. Another Euphaedra species recorded from Semuliki, E. alacris which is detailed within the list is again supposedly absent from Semuliki. It is also recorded as being an Albertine Rift endemic, which is not the case.

Genus Cymothoe

Similar species richness for this genus was found at both sites and for both monitoring programmes although abundance was 37% greater during the long-term monitoring programme. 


Taking the combined intensive periods, site abundance of Cymothoe species was 2.5 times greater at Kirumya than at Red Monkey. This was in contrast for Cymothoe abundance during the long-term monitoring programme where abundance was 1.6 times greater at Red Monkey than Kirumya. Cymothoe ochreata and C. cyclades abundance was found to be significantly more frequent at Kirumya than Red Monkey. A number of species were very uncommon and sampled rarely including Cymothoe reginaeelisabethae, C. colmanti and C. lurida.

Three Cymothoe species recorded from Semuliki National Park; Cymothoe confusa, C. cyclades and C. sangaris have been designated as Vulnerable (IUCN Red List categories and criteria, 2012) in a publication 'Nationally Threatened Species for Uganda' (Wildlife Conservation Society, 2016). See the discussion above for the non-justification of this category.

Cymothoe beckeri (male)

Specific genus site links, references and bibliography:

Birket‐Smith, J., (1970). The behaviour of Euphaedra in relation to temperature, relative humidity and light (Lep. Rhopalocera). Entomologica Scandinavia 1: pp.123‐126.

Callaghan, C.T., Borda-de-Água, L., van Klink, R., Rozzi, R. and Pereira, H.M., (2023). Unveiling global species abundance distributions. Nature, Ecology and Evolution:

Fox, R.Harrower, C.A., Bell, J.R. et al., (2019). Insect population trends and the IUCN Red List process. Journal of Insect Conservation 23: 269-278.

Larsen, T., (2005). Butterflies of West Africa, Apollo Books, Stenstrup.

Reşit Akçakaya, H., Hochkirch, A., Bried, J.T. et al., (2021). Calculating population reductions of invertebrate species for IUCN Red List assessments. Journal of Insect Conservation 25: 377-382.

Wildlife Conservation Society (2018). Nationally threatened species for Uganda (National red list for Uganda for the following
taxa: mammals, birds, reptiles, amphibians, butterflies, dragonflies and vascular plants). Akite, P., Prinsloo, S., Ayebare, S., Toloa, E. and Clausnitzer, V.. Proposed list of threatened butterflies.

bottom of page