Effects of the Louisiana crayfish invasion and other human impacts on the African clawless otter in the Ewaso Ng’iro ecosystem.
An Interim Project Report.
Mordecai O. Ogada, BSc., MSc.
The introduction and subsequent spread of non-indigenous species into new habitats has always been shown to have far-reaching ecological effects. This is especially true within the confines of inland aquatic ecosystems. An example is the effect of the Nile perch (Lates niloticus) on the cichlid species inhabiting the Lake Victoria ecosystem. This study examined the effect of the exotic Louisiana red swamp crayfish (Procambarus clarkii Girard) on two indigenous aquatic species in the Ewaso Ng’iro ecosystem in central Kenya. These are the African clawless otter (Aonyx capensis Lichtenstein) and the indigenous freshwater crab (Potamonautes neumannii), which are predators and competitors to the crayfish, respectively. This study covered both the rainy and dry seasons in order to cover the seasonal variations in the aquatic community.
The central Laikipia plateau is a semi-arid region where rivers and other water sources are heavily exploited. The study also shed some light on the status of the African clawless otter, which is unknown in much of its range. It is a seldom seen animal, and most respondents in the study area were found to be unaware of its existence there.
Intensive sampling suggests that crayfish have supplanted indigenous crabs in much of the lower Ewaso Ng’iro river north of the equator. This was confirmed by experimental ‘competition’ between captive crabs and crayfish.
The trap sampling also revealed that crayfish are excluded from the upper Ewaso Ng’iro south of the equator by the low water temperature there. Collection of otter spraints (faeces) within the study area revealed seasonal variation in the territorial marking behaviour of the otters. Laboratory analysis of faeces also revealed that crayfish was the primary food source used by the otters in the lower Ewaso Ng’iro, and that the availability of this resource varied between seasons. A relationship was found between the seasonal variation in otter territorial marking behaviour of the otters and the variation in availability of crayfish. This appeared to be ‘crayfish-driven’ because seasonal variation in otter behaviour was not observed in the upper Ewaso Ng’iro where the water temperature was too low for crayfish. This study recommends ways in which people can exploit the Ewaso Ngi’ro river with minimum effect on aquatic fauna. These methods need to be applied urgently because currently in Kenya, there is no regulation on water extraction, pollution by ‘small’ (non-industrial) polluters is largely unchecked, and sport fishermen continue to introduce or ‘stock’ aquatic ecosystems with exotic species without any thought to the ecological impacts of these activities.
The earliest recorded sighting of the African clawless otter in the Laikipia plateau was in 1883 (Thomson,1885), so it can be assumed that they preceded the Louisiana red swamp crayfish. It is the most important predator in the upper Ewaso Ng’iro river ecosystem, with other prominent predators being the marsh mongoose (Atilax paludinosus), fish eagle (Haliatus vocifer), darter (Anhinga rufa), goliath heron (Ardea goliath), monitor lizard (Varanus niloticus) and predatory fish e.g. barbels (Barbus paludinosus). According to Kingdon (1989), the marsh mongoose is the nearest ecological competitor of the clawless otter, although it forages mostly in marshes and on the riverbank whereas the otter forages in deep water. Otters’ preferred hunting depth is 0.5 to 1.5 meters (Somers, 2000). This is an important factor in determining otter distribution in shallow rivers such as the Ewaso Ng’iro. The clawless otter is a large animal measuring 1 to 1.6 metres in length and weighing, on average, 10 to 18 kg (Kingdon,1997). It is mainly crepuscular but activity has been recorded at all times of day or night.
Diet of the African clawless otter in the study area.
The most frequent evidence of Aonyx is the dung deposits that consist mainly of crushed crabshells and catfish bones (Kingdon, 1997), although they also eat molluscs, small mammals and birds (Stuart & Stuart 1997). Copley (1950) noted these otters eating young maize cobs and they may therefore also eat fallen fruit that are frequently plentiful along riverine habitats. The most important factor in their dietary choices here seems to be availability of the particular food item. Therefore, crayfish are the most important part of the diet of Aonyx in the lower Ewaso Ng’iro ecosystem as are crabs in the upper Ewaso Ng’iro. especially because hunting is mostly by touch in turbid water. Fish is less prominent in their diet because their clawless digits make fish very difficult to catch. Consequently, analysis of the composition of otter faeces over time can be a reliable indicator of aquatic community dynamics.
The Louisiana Red Swamp Crayfish
THE STUDY AREA
The Ewaso Ng’iro river flows from the Aberdares and Mt. Kenya through Laikipia and Samburu Districts of central Kenya, ending in the Lorian swamp (Fig. 1). However, this study was based at the Mpala Research Centre in Laikipia. The Laikipia area is dissected by the Ewaso Ng’iro river and its tributaries which arise in the Aberdare Mountains and Mount Kenya (Fig1). This river system is an important component of the Laikipia and Samburu ecosystems because it is a semi-arid area receiving an average of 600mm of rainfall per annum (Paton & Ogada, 2001). It is the only permanent water source for numerous herds of livestock reared by ranchers and pastoralists in the area as well as a large wildlife population, particularly in the Samburu National Reserve. The river water levels fluctuate greatly between seasons from a minimum of 0.4 meters deep and a maximum of 3.5 meters deep measured at a gauge mounted at Mpala Research Centre (pers. obs.).
Crayfish and crabs were sampled using funnel-mouth fish traps. These are made using ¾ inch size wire mesh reinforced with binding wire. These were weighted with stones, placed on the bottom of the river, baited with meat and secured to the river bank using coloured nylon string. Sites will be every 1 km along the Ewaso Ng’iro and Ewaso Narok rivers where they border Mpala Ranch (about 20 sites). Traps were emptied, rebaited (if necessary) and reset every 24 hours during the trapping period, which was every fourth week. The study area was divided into three ‘blocks’.
a) The Ewaso Ng’iro river upstream of the junction with the Nanyuki river.
b) The Nanyuki river and the Ewaso Ng’iro downstream from the junction of the two rivers.
c) The Ewaso Narok river (another tributary of the Ewaso Ng’iro).
Crayfish caught in the funnel-mouthed traps were weighed and measured to establish the age distribution. Weight of the crayfish and diameter of their antennule and antennule bases were also noted and used to derive a mathematical relationship between the 2 variables. Antennules were chosen for measurement because being situated on the cephalothorax, they are always eaten by the otter (unlike claws or legs, which are sometimes discarded) and are usually retrieved intact from otter faeces. The fact that there are only 2 antennules per crayfish, allows for the deduction of numbers eaten to a high degree of accuracy. This allowed for the deduction of the mass of crayfish consumed at each study site from measurement of antennule bases retrieved from faecal samples. The composition, abundance, location, and frequency of the faecal samples in the study area were used as an additional indicator of fluctuations in the distribution of the otters and their prey.
Otter population densities were estimated using a combination of direct observation and mapping of territories. Frequency of spraint collection was used to deduce temporal distribution of otters within the study area. Territorial marking sites were assessed for various abiotic factors including vegetation, substrate composition, water speed, and water depth to deduce the possible effects of these factors on crustacean distribution.
Otters were live-trapped in Tomahawk single-door, rigid live traps baited with fish (Wilson et al. 1996), and otter spraints from other territories (Michael Somers, pers. comm), marked and released. Trapped animals were anaesthetized using a pole syringe, using etorphine hydrochloride. The anaesthesia was reversed using Diprenorphine.
More emphasis was placed on spraint collection and camera trap photography as indicators than trapping and marking of otters. This is because I found that repeated trapping at latrine sites caused the otters to abandon them, thus depriving the study of its most valuable data source. Both sexes of the clawless otter establish territories on which they remain for long periods (Kingdon, 1989). Otter distribution, along the same stretch of the Ewaso Ng’iro, was initially investigated by direct observation (between 06.00h & 10.00h, and between 16.00h & 19.00h when the animals appear most active and can be observed through binoculars) and by finding faecal deposits (spraints). Sightings of otters and locations of spraints will be recorded using a Global Positioning System (GPS). Spraints were collected and returned to the laboratory where they were dried, weighed, sorted and their contents identified under a dissecting microscope. Again using antennule base size, the size distribution of prey was compared to the size distribution obtained from trapping, to see if otters were selective in the sizes of prey they took. The mathematical relationship between the antennule base size and overall weight of live caught crayfish was used to deduce the mass of crayfish being taken by otters over time.
The three motion sensor cameras (Cam Trakker) will be moved from site to site in order to confirm and extend our estimates of otter home range sizes obtained from observation and trapping. The Cam Trakker is a fully automated 35mm camera combined with a passive infra red motion detector, powered by 4 alkaline batteries, which can be recharged.
1. Weights of crayfish caught in the wire mesh traps were regressed against the diameters of their antennule bases and a linear relationship was established between the 2 variables:
W=(-x + yD)
-Where ‘W’ is weight and ‘D’ is antennule diameter.
This will then be used to deduce the amount (weight) of crayfish being consumed by otters from various points in the study area, as well as the otters’ preferred prey size.
1. Density of otter spraints (# spraints per unit length of river) within seasons was regressed against proportion of crayfish (by mass) in the spraints to determine whether availability of crayfish is influencing the otters’ use of the habitat. Seasons in this context were defined by the river water level and not precipitation within the study area.
African clawless otters mark their territories with spraints and aggregations of faecal deposits on rocks along the river banks (Kingdon,1997). It was possible to map otter territories by placing a spraint on rocks along the river up to a point where it gets removed by the otters and replaced with another spraint. This identifies the point at which the next territory begins. Three territories were thus identified and marked A, B, and C (see map). This method was preferred to telemetry because I noticed that repeated trapping of otters caused them to abandon their holts, which were an important source of long-term data. However, the otters that were trapped were freeze marked to facilitate recognition of individuals during direct observation.
During the dry seasons when the Ewaso Ngiro river level dropped below 35cm, Territorial behaviour of otters in the study area broke down, i.e. placement of ‘alien’ spraints on rocks did not elicit any reaction from resident otters. During these periods, otter faeces were found randomly scattered along the riverbanks and not aggregated at holts. This is illustrated in Graph 2 by plotting faeces collection numbers against a function derived from the GPS coordinates at which they are collected, so that faeces aggregations appear as rows of points.
Each of the otter territories was found to include a man-made water reservoir. These are small dams with walls 1.2 –3 meters high.
Direct observation revealed that each territory is occupied by a family group consisting of 4-6 individuals. Population of otters within the study area marked on the map can therefore be estimated to be at least 12 individuals.
Analysis of all otter faeces collected during the study period gave mean proportions of 64% crayfish remains, 34.85% fish remains and 1.15% crab remains in terms of weight. This indicates that the exotic crayfish has become a very important part of the diet of the African clawless otter in the study area.
Crayfish caught in sampling traps throughout the study area were weighed and measured. Analysis of weight and diameter of antennule bases gave the following mathematical relationship between the 2 variables;
y= -36.981+ (58.87* x)
Where y is the weight of the crayfish and x is the diameter of the antennule base.
Using the above formula and by counting the number of antennules in each faecal sample, it was possible to calculate the number of crayfish eaten and the amount of crayfish biomass removal by otters per faecal sample. Crayfish biomass removed per faecal sample was positively correlated to the Ewaso Ng’iro river level (corr.=0.41, p<0.0001, n=120). This trend is also illustrated in graph 3.
Graph 1. River Level trends 30th Jan 2003- 1st Oct
2003
Graph 2. Analysis of Faeces Aggregation in the study area.
Graph 3. Crayfish biomass in faeces vs. River Level
Graph 4. Mean crayfish weights per faecal sample
Graph 5. Otter faeces collections at different river levels.
INTERIM CONCLUSIONS
1. The Freshwater crab (Potamonautes neumannii)
The freshwater crab has been supplanted by the exotic crayfish (Procambarus clarkii) within the study area and is probably approaching local extinction. Trap sampling over the study period yielded a total of 406 crayfish and five crabs. This is further borne out by the analysis of otter faeces over the same period. Only 1.15% (by weight) of all otter faeces collected during this period consisted of crab remains. This is compared to faeces collected from the Ewaso Ng’iro river in 1990 that consisted ‘almost entirely’ of crab remains (Kruuk, 1990). I therefore conclude that P. neumannii is an inferior competitor to P. clarkii in shared habitats.
2. The Louisiana crayfish (Procambarus clarkii)
This invasive species has become a primary food source for otters in the Ewaso Ng’iro ecosystem. Its success in this ecosystem is likely to have had an effect on other aquatic organisms as well, and this needs to be investigated further. The spread of the crayfish has been restricted by water temperature because crayfish cannot survive in water below 14 deg. Celsius (Covich, 1978). No crayfish have been trapped south of the point on the Nanyuki river marked “14’C” on the map. Otter faeces collected from this zone were also found not to contain any crayfish remains.
3. The African clawless otter (Aonyx capensis)
This predator’s dependency on the Louisiana crayfish as a food source has made it susceptible to environmental factors affecting its prey base. This is exacerbated by the effect of the Louisiana crayfish on other prey species like the freshwater crab. This has been a negative human impact on the clawless otter.
The periods of low river levels are characterized by lower percentages of crayfish in otter faeces. Otter faeces collected during these periods had higher percentages of bony fish remains. I considered high percentages of fish remains in faeces a sign of food stress, because the otters’ vestigial claws made it very difficult for them to catch fish. Aggregation of faeces was diminished during the dry seasons and the territorial reactions to ‘alien’ spraints placed within known otter territories were not observed.
During the surveys of the riverbanks during the dry season, we found crayfish remains in the faeces of monitor lizards (Varanus niloticus), genets (Genetta genetta), and olive baboons (Papio anubis). Three monitor lizards were accidentally caught in the crayfish sampling traps and examination of their stomach contents revealed that they were feeding on crayfish. Crayfish remains were also recovered beneath known roosts of goliath herons (Ardea goliath) grey herons (Ardea cinerea) and black-headed herons (Ardea melanocephala). Close examination of the foraging techniques used by the above species reveals that all of them are ‘dippers’ (as opposed to divers). This means that they would not be able to catch crayfish (which are bottom feeders) at a depth of >50 cm in turbid water.
I therefore conclude that there is increased competition for crayfish during the dry season from species that would not otherwise be able to access crayfish. The otters are therefore forced to forage over larger stretches of river in order to fulfill their nutritional needs, and cannot afford to defend territories. During the dry season, the man-made water reservoirs provide refuges for crayfish. The recovery of the crayfish population during the wet season starts from these reservoirs, hence their inclusion in the territories defended by the otters in the wet season. This has been a positive human impact on the otter population
4. Interspecific Competition for Crayfish
Analysis of river depth against faeces collection indicated a threshold river level of 38cm beyond which there was a sharp rise in the number of otter faecal samples collected (Graph 5). It appears that the low availability of crayfish to otters at river levels below 38cm are a result of interspecific competition. The sharp rise beyond 38cm level was followed by a gradual decline at levels > 50cm. The graph is also an indicator of fluctuations in territorial behaviour because the holts are now well known to us and we are only likely to miss the single scats deposited away from the holts. The capabilities of the various competitors were examined closely in order to explain the 39cm threshold. These included the 3 species of herons found within the study area (Young & Wamiti 2003).
a) The nile monitor lizard is a versatile and powerful swimmer, which can stay submerged for up to 20 minutes (Spawls et. al., 2002). This makes them the least vulnerable predator to river level fluctuations at this low level. They forage on land and in the water taking both fish and invertebrate prey. These factors discounted the possibility of the monitor being the key competitor for crayfish.
b) The goliath heron is a large bird, standing up to 1.5 meters tall (Stevenson & Fanshawe, 2002). It is most common in wetlands and swamps (Zimmerman et. al., 1996) so it is not found at high density in river ecosystems without much cover like the Ewaso Ng’iro. Its legs are an average 60cm long (NMK collection), enabling it to forage in water over 50cm deep, so it would not be affected by the 39cm depth threshold.
c) The grey heron is a much smaller bird than the goliath heron, standing 1 meter tall (Stevenson & Fanshawe, 2002). Its legs are average 45cm long (NMK collection) so its foraging is likely to be affected by the 39cm depth threshold. However, this species uncommon in Laikipia because it favours alkaline lakes and coastal flats (Zimmerman et. al., 1996). It is therefore unlikely to be the key competitor for crayfish.
d) The black-headed heron is slightly smaller than the grey heron standing 92cm tall (Zimmerman et. al., 1996). Its legs are 42cm long (NMK collection) and this places it in the bracket affected by the 39cm threshold. It is the commonest and most widespread East African heron. It is known to prefer drier habitats (Stevenson & Fanshawe, 2002) and is consequently most abundant species of heron in Laikipia. In addition to this, the leg length is closest to 39cm. This species is therefore the most likely to be the driving force behind the 39cm threshold.
1. We need to do a more intensive survey of the upper Nanyuki river south of the 14 degree mark where there are no crayfish. This would be to determine whether the variation in otter behaviour downstream is crayfish-driven. This is because the exposure of crabs to predators does not change with water depth because they voluntarily crawl out of the water onto the riverbank rocks. In that case, otters would defend and mark their territories throughout the year.
2. We also need to intensively survey the lower Ewaso Ng’iro just north of Laikipia district. In this region, the river dries up during the dry season. It is important to know how otters utilize this part of the river and how the drying up affects their ecology. The duration and geographical extent of the drying up is greatly influenced by the extraction of water by upstream users.
Alderman, D. J. (1996). Geographical spread of bacterial and fungal disease of crustaceans. Rev. Sci. Tech. O I E (Off. Int. Epizoot.), 15, 603-632.
Begon, M., Harper, J. L., &Townsend, C. R., (1996). Ecology; Individuals, Populations, and Communities (third edition). Blackwell Science, Oxford.
Caughley, G. (1977). Analysis of Vertebrate Populations. Wiley, London.
Cleveland, W. S., (1981). LOWESS- A program for smoothing scatter plots by robust locally weighted regression. The American Statistician 35: 54
Copley, H. (1950). Small Mammals of Kenya. Highway Nature Guides, Highway Press, Nairobi.
Day, G.I., Schemnitz, S.D. & Taber, R.D. (1980). Capturing and marking mammals. In S.D. Schemnitz (ed.) Wildlife Management Techniques Manual. The Wildlife Society, Washington.
DeLury, D.B. (1947). On the estimation of biological populations. Biometrics, 3, 145-167.
Digby, P.G.N. & Kempton, R.A. (1987). Multivariate Analysis of Ecological Communities. Chapman & Hall.
Hobbs, H. H., Jass, J.P. & Huner J.V. (1989). A review of global crayfish introductions with particular emphasis on two North American species (Decapoda, Cambaridae). Crustaceana, 56, 299-316.
Holdich, D. M. (1999). The negative effects of established crayfish introductions. In: F. Gherardi and D. M. Holdich, (eds) Crayfish in Europe as alien species. How to make the best of a bad situation? 31-47, A.A. Balkema, Rotterdam and Brookfield.
Kingdon, J. (1989). East African Mammals. IIIA Carnivores. University of Chicago Press.
Kingdon, J. (1997). The Kingdon Field Guide to African Mammals. Academic Press, London.
Kruuk, H. & Goudswaard, P.C. (1990). Effects of fish populations in Lake Victoria on the food of otters (Lutra maculicollis Schinz and Aonyx capensis Lichtenstein). African Journal of Ecology, 28, 322-329.
Lodge, D. M. & Hill, A. M. (1991). Factors governing species composition, population size, and productivity of cool-water crayfishes. Nordic Journal of Freshwater Research, 69, 111-136.
Lodge, D.M., Taylor, C.A., Holdich, D.M. & Skurdal, J. (2000). Nonindigenous crayfishes threaten North American freshwater biodiversity: lessons from Europe. Fisheries, 25, 7-20.
Maxwell, G. (1960). Ring of Bright Water. Longmans, London.
Mkoji, G.M., Hofkin, B.V., Mungai, B.N., Kihara, J.H., Mungai, F.K., Ouma, J.H., Koech, D.K. & Loker, E.S. (1992). Control of natural populations of schistosome transmitting snails by the crayfish, Procambarus clarkii, in temporary man-made ponds in Kenya. In: Aquaculture and Schistosomiasis. pp. 264-271, National Academy Press, Washington, DC.
Noblitt, S. B., & Payne, J. F. (1995). A comparative study of selected chemical aspects of the eggs of the crayfish Procambarus clarkii and P. zonangulus. Crustaceana 68(6)
Paton, S., & Ogada, M. O., (2001). 1998-2001 Meteorological and hydrological summary for Mpala Research Centre. Smithsonian Institution, Washington, D.C.
Pielou, E.C. (1969). An Introduction to Mathematical Ecology. Wiley, New York.
Romesberg, H.C. (1984). Cluster Analysis for Researchers. Lifetime Learning Publications, Belmont, California.
Rowe-Rowe, D. T. (1977). Food ecology of the African clawless otter in Natal, South Africa. Oikos Vol. 28 (210-219)
Somers, M. J. (2000). Foraging behaviour of Cape clawless otters (Aonyx capensis) in a marine habitat. J. Zool., 252: 473-480. Zoological Society of London.
Stevenson, T., Fanshawe, J. (2002) Field Guide to the Birds of East Africa. T & AD Poyser, London.
Stuart, C. & Stuart, T. (1997). The Larger Mammals of Africa. Struik.
Thorp, J. H. & Covich, A. P. (2001). Ecology and Classification of North American Invertebrates. Academic Press.
Thomson, J. (1885). Through Maasai Land (1968 Edition). Hodder and Stoughton.
Young, T. P., Wamiti, W. (2003) Checklist of the birds of Mpala & Segera. University of California, Davis.
Zimmerman, D. A., Turner, D. A., & Pearson, D. J. (1996) Birds of Kenya and Northern Tanzania. Princeton University Press. Princeton, NJ.