Catena 178 (2019) 307–312
Contents lists available at ScienceDirect
Catena
journal homepage: www.elsevier.com/locate/catena
The combined effect of termite bioturbation and water erosion on soil T
nutrient stocks along a tropical forest catena in Ghana
Jeppe Aagaard Kristensena,⁎, Susan Helene Boëtiusb, Mark Abekoec, Theodore W. Awadzic,
Henrik Breuning-Madsenb,1
a Department of Physical Geography and Ecosystem Science, Lund University, Sölvegatan 12, 223 62 Lund, Sweden
bDepartment of Geoscience and Natural Resource Management, Faculty of Science, University of Copenhagen, Øster Voldgade 10, DK-1350 Copenhagen K., Denmark
c Department of Soil Science, University of Ghana, Legon, P.O. Box LG59, Accra, Ghana
A R T I C L E I N F O A B S T R A C T
Keywords: In the tropical moist semi-deciduous forests of West Africa, soil catenas with extremely gravel-rich soil horizons
Nutrient stock depletion at the summits and upper slopes and largely gravel-free profiles at the lower slope are common. Previous in-
Termite bioturbation vestigations have suggested that these gravel layers are the result of macro-invertebrates mining of fine-grained
Tropical forest catena soil material from the subsoil leaving behind the gravel, to build galleries at the surface subsequently exposing it
West Africa to water erosion transport downslope. We examined the indirect effect of this process on the distribution along a
Ecosystem functioning
Soil erosion soil catena of crucial base cations (Ca
2+, Mg2+, K+) and plant available phosphorous (P), which is often growth-
limiting in these tropical ecosystems. We found that the export of fine-grained soil material at the top of the
catena reduces the soil stocks (to 1m) of these elements by up to 60%, while the soil fertility downslope did not
change significantly. This important long-term (100–1000 yr scale) reduction in soil fertility at the top of slopes
resulting from bioturbation and water erosion is overlooked in contemporary literature, which primarily focus
on the beneficial impact termites and ants have on ecosystem functioning in more level savannah landscapes. As
the type of catena studied is widespread across tropical environments, this effect is likely ecologically substantial.
Future research should aim at understanding such long-term consequences of bioturbation on landscape ecology
as well as soil heterogeneity and fertility, so we do not overlook potential negative ecosystem effects.
1. Introduction et al., 2002). However, other authors have in parallel suggested that
long periods of soil faunal activity, especially that of termites and ants,
In tropical West Africa, gravelly soil horizons are common at the drive a relative accumulation of coarse fragments at a certain depth in
summits and upper slopes of the landscape, while the soils at the lower the soil solum, because of selectively mining of bulk soil matrix finer
part of the slope and at the bottom are almost gravel free throughout than 1–2mm at this depth for building structures at the surface (Nye,
the profile (Nye, 1954; Dijkerman and Miedema, 1988; Adu, 1992). The 1955; Fritsch et al., 2002). Thus, in areas with more upward transport
gravelly layers are exposed at the surface or covered by a shallow of fine-grained particles than removal from the surface (e.g. by ero-
gravel-free soil layer. They are typically 0.5–1m thick and rarely> sion), a gravel free soil layer might develop superimposing a gravel rich
1.5 m (Dijkerman and Miedema, 1988; Adu, 1992; Awadzi et al., layer.
2004). The gravel consists of primarily iron nodules (pisolites) mixed Our own studies in Ghana support the faunal theory (Awadzi et al.,
with rock fragments, mainly quartz (Breuning-Madsen et al., 2007; 2004; Breuning-Madsen et al., 2004, 2007, 2017; Kristensen et al.,
Elberling et al., 2013). 2015). Based on sediment deposition dating (OSL), we suggested that
The genesis of the gravelly layers is a matter of scientific dispute. A on timescales of 1–5 ka, termites can cause redistribution of soil ma-
group of geological/geochemical models repeatedly emphasise the terial finer than 2mm that is relevant for soil and landscape formation
importance of great age, climatic changes and tectonic activity for (Breuning-Madsen et al., 2017; Kristensen et al., 2015). This conclusion
formation and composition of continuous and discontinuous ironstone is in line with several reviews (e.g. Lobry de Bruyn and Conacher, 1990;
layers (e.g. Alexander et al., 1956; Nahon, 1986; Thomas, 1994; Fritsch Wilkinson et al., 2009; Jouquet et al., 2016), and recent methodological
⁎ Corresponding author.
E-mail address: jeppe.aa.kristensen@gmail.com (J.A. Kristensen).
1 Deceased.
https://doi.org/10.1016/j.catena.2019.03.032
Received 9 October 2018; Received in revised form 28 February 2019; Accepted 20 March 2019
Available online 27 March 2019
0341-8162/ © 2019 Elsevier B.V. All rights reserved.
J.A. Kristensen, et al. Catena 178 (2019) 307–312
developments have allowed scientists to quantify horizontal (e.g. re- mid-July, and secondarily from September to November. There is ty-
mote sensing and geostatistics; Funch, 2015; Mujinya et al., 2014; Obi pically a dry period from December to February. The annual potential
and Ogunkunle, 2009), and vertical redistribution of matter (e.g. se- evapotranspiration is about 1400mm, while the annual actual evapo-
diment dating; Kristensen et al., 2015; Jouquet et al., 2017). transpiration is about 1200mm (Christensen and Awadzi, 2000).
Recently, authors have applied the concept of ecosystem services to Christensen and Awadzi (2000) estimated the surface run off to be
evaluate the importance of termites for humans (Jouquet et al., 2011; about 15% of the precipitation during the rainy season posing a risk of
Kaiser et al., 2017). Yet, most research has hitherto focused on the surface soil erosion, which resulted in small scale fluvial patterns in
improvement of soil conditions, due to the direct accumulation of soil surface sediments along the catena (e.g. micro-scale alluvial fans and
organic matter (SOM) and nutrients in the faunal structures at or near braided sediments). The mean annual temperature is about 27–28 °C
the surface (mounds, nests, fungus combs) (e.g. Lobry de Bruyn and with little seasonal variation (Christensen and Awadzi, 2000).
Conacher, 1990; Sarcinelli et al., 2009, 2013), and erosion control re- The landscape is gently rolling and the soils on the slopes form one
sulting from the creation of soil macropores in relatively level savannah of the most common catenas in the tropical moist semi-deciduous forest
landscapes. In contrast, the indirect long-term (100–1000 year) effects zone in Ghana, which is the Bekwai/Nzima-Oda association (Fig. 2)
on the soil fertility due to termite mediated sorting of soil particle sizes according to the Ghanaian soil classification system (Ahn, 1970;
– vertically in the soil profile, and horizontally across the landscape – Owusu-Bennoah et al., 2000). The soil parent materials are almost ex-
have received less attention, although the vast majority of the desirable clusively Pre-Cambrian rocks, predominantly consisting of phyllites and
soil properties are related to the fine particles (< 0.2mm), e.g. water schists (Adu, 1992). The catena is about 500m long with an average
and nutrient retention. Hence, the indirect termite impact on soil slope of approximately 5%. The soils at the upper part of the slope,
quality also varies substantially across the landscape. Here we asked the called Bekwai and Nzima soil series, are red-brown or brown, concre-
question: What is the effect of the faunal and water mediated down- tionary, acid, well-drained kaolinitic clay soils formed in phyllites with
slope transport of fine-grained material on the soil stocks of the im- intrusions of quartz as the main constituents (Wills, 1962; Owusu-
portant nutrients phosphorus (P), calcium (Ca), potassium (K) and Bennoah et al., 2000). The soils are well drained, yet weathered rock is
magnesium (Mg) along a typical West African forest catena? found at a depth of 150 to 200 cm, which occasionally impede the
drainage in the wet season, resulting in the formation of temporary
groundwaters producing pseudogley features near the base of the
2. Materials and methods solum. According to the World Reference Base they are mainly classi-
fied as Acrisols (IUSS Working Group WRB, 2006). At the middle and
2.1. The study site lower slope, the Kokofu soil series are slightly-to-very acid, yellowish-
brown clay loams developed in gravel-free colluvium deposited as a
The study site is at the University of Ghana's research station in result of soil erosion upslope. The uppermost meter of the soil was
Kade (6.10°N 0.84°W, Fig. 1). The climate is humid tropical with an deposited during the last 4–5 ka (Breuning-Madsen et al., 2017). The
average annual rainfall of about 1400mm, mainly falling from March to soils at the middle slope are mainly Acrisols. At the catena bottom, the
soils are imperfectly drained greyish clay loams to sandy loams typical
for the Oda soil series, which are usually flooded in the peak rainy
season. The parent material is in situ weathered rock. They are classi-
fied as Gleysols (IUSS Working Group WRB, 2006). Due to the inter-
ference between nutrients coming from these seasonal floodings and the
groundwater in general, the lowest catena member (Oda) is not used to
infer any conclusions on the effects of upslope processes on soil nutrient
stocks. At the lower end of the catena a small tributary channel called
the Kadepon stream (‘stream’ in Fig. 2) is situated, which is only active
in the main rainy season when the bottom is flooded (Christensen and
Awadzi, 2000).
The study was conducted in a uniform and undisturbed (i.e. no dead
wood has been removed since it was protected in the 1940s) semi-de-
ciduous moist tropical forest. The dense vegetation belongs to the
Antiaris-Chlorophora association (Lawson et al., 1970). Termites of the
families Kalotermitidae (Cryptotermes sp.), Rhinotermitidae (Copto-
termes intermedius and Coptotermes sp.) and Termitidae (Microtermes
subhyalinus) dominate the macro-invertebrate community (Awadzi
et al., 2004). These tree-eating termites do not build mounds like the
savannah termites (primarily Macrotermes), but build galleries of soil
material in dead logs and trees (see pictures of soil filled logs in sup-
plementary information of Breuning-Madsen et al. (2017)). When the
logs decompose, the gallery material forms a gravel-free horizon on top
of the existing soil surface. When the dead trees collapse, the soil ma-
terial in these can form passive dome-shaped heaps of considerable
dimensions, which is subsequently levelled by erosion adding a fine-
grained soil layer to the surroundings. Awadzi et al. (2004) estimated
one of the larger mounds to be 1.6 m high with a diameter of 8m
yielding an approximate volume of 50m3 and weight of about 75Mg,
and they concluded that the soil was primarily mined from the subsoil
below the A-horizon. Such heaps are visible on approximately the upper
half of the catena. For a more thorough description of these processes
and properties of the soil material in logs and heaps, we refer to Awadzi
Fig. 1. The location of Kade in Ghana and the associated agroecological zones. et al. (2004) and Breuning-Madsen et al. (2007). We cannot exclude the
308
J.A. Kristensen, et al. Catena 178 (2019) 307–312
Fig. 2. Model of the genesis of the catena (a), pictures of the four soil profiles studied (b), and their particle size distributions (c).
possibility that some of the soil transport could be due to other animals were taken in triplicate from three depths within the uppermost meter.
or processes, but judging from our numerous visits at the site and A horizontal surface was excavated at the chosen depth and three
previous studies mentioned above, we deem it is most likely that ter- 100 cm3 soil sampling steel rings were gently hammered vertically into
mites are the main upward soil transporting agents, while water erosion each soil layer of interest, to avoid soil compaction. The soil samples
is the main downslope transport agent. were cut to fit the volume of the ring, before covering the top and
bottom with plastic lids. In the two soil pits at the top of the slope, soil
2.2. Soil sampling samples were taken from 4 depths. The procedure was the same as for
the two profiles downslope, except for the gravelly horizons, where it
Four soil profile pits were inspected to represent the four dominant was not possible to hammer the rings into the soil. Instead, the proce-
soil series along the slope: Bekwai and Nzima at the upper slope, Kokofu dure described by Grossman and Reinsch (2002) was used in triplicates.
from the mid-lower slope and Oda from the bottom of the catena Similar to the other horizons, a horizontal surface was excavated into
(Fig. 2). We do not have samples from replicated soil pits, but from the profile wall at the layer of interest. The gravel-rich soil sample was
three decades of research along the catena and excavation of several carefully taken from the surface by hand to create a hole without dis-
soil pits at all segments, we carefully selected these four pits as good turbing the surrounding soil. The volume of the removed soil was
representatives of the four dominant soil types. The soils were described measured by carefully lining the whole with a plastic bag and filling it
in the field according to the FAO guidelines (FAO, 1990) and the colour with water until the water surface was at same level as the exposed
according to Munsell Soil Colour Chart. From the soils without gravelly horizontal soil surface. The weight of water in the bag was converted to
3
horizons, undisturbed soil samples for bulk density determinations volume of soil (g water= cm soil). Additionally, disturbed bulk soil
309
J.A. Kristensen, et al. Catena 178 (2019) 307–312
samples from all horizons were collected for chemical analyses. Com- concentrations than Nzima but higher than Bekwai. The Oda soil at the
posite samples of minimum three subsamples were used to get a good bottom close to the stream showed relatively high nutrient concentra-
representation of average soil conditions in each horizon. tions. Overall, plant available P concentrations were very low along the
catena; the ratios to the other nutrients were approximately 1:8 for K,
2.3. Laboratory analyses 1:20 for Mg and 1:70 for Ca except for the valley bottom, where the
concentrations of Ca and Mg were relatively high.
We dried the soil samples for determination of bulk density at
105 °C for 24 h. Based on the weight of dry soil and the volume of 3.3. Nutrient stocks
sample, the bulk density was calculated. The aggregates in the dis-
turbed soil samples were crushed and sieved through a 2mm mesh. The Considering only the fine-earth fraction (< 2mm), Nzima was more
weight of particles (gravel) > 2mm was determined. The particles nutrient rich than the Bekwai, especially according to Ca and Mg
finer than 2mm was used for chemical and physical properties as fol- (Table 3). Kokofu showed intermediate stocks of Ca, Mg and P com-
lows: pared to Nzima and Bekwai. The Oda soil at the valley bottom had a
The texture was determined by the hydrometer method for the silt relatively high content of Ca and Mg compared to the other soils. In-
and clay fractions (Day, 1965), and dry sieving of the sand and gravel cluding the gravel changed the nutrient stock substantially. The nu-
fractions. Soil pH was determined potentiometrically in a suspension of trient stock increased downslope from the Bekwai at the top to the Oda
soil and 0.01M CaCl2 at a soil-liquid ratio of 1:2.5 (Thomas, 1996). soil at the bottom of the catena. The nutrient stock in the highly eroded
Total organic carbon (TOC) was analyzed using the dry combustion Bekwai was reduced by 60% of a similar soil without gravel, while the
method at 1.250 °C in oxygen on an Eltra SC-500 analyzer, with an reduction in nutrient stocks in the Nzima was only about 45%, due to
accuracy of± 0.2% (ELTRA, 1995). As no carbonates were present in the 20 cm thick termite soil layer upon the gravelly horizon. At the
the samples, the total carbon content is equivalent to the total organic colluvium, the reduction in nutrient stocks due to gravel inclusion was
carbon content. negligible. Similarly, the Oda soil at the valley bottom showed very
The exchangeable Ca2+, Mg2+ and K+ ions were extracted with 1M little difference between including/excluding gravel.
ammonium acetate (NH4C2H3O2) at pH 7 (Chapman, 1965) and the
concentrations were determined by Atom Absorption Spectro- 4. Discussion
photometry (AAS) using a Perkin Elmer Analyst 400 (Perkin Elmer,
Waltham, MA, US). Plant available P (POlsen) was extracted with 0.5M 4.1. The distribution of nutrients along the catena
sodium bicarbonate. After extraction, the phosphorus content was de-
termined spectrophotometrically by the molybdenum-blue method The soil at the top of the catena showed a 45–60% reduction in soil
(Murphy and Riley, 1962). The exchangeable cations and P were ana- nutrient stocks due to the upward termite transport of fine material and
lyzed in laboratory triplicates. Therefore, variation in Table 2 represent the subsequent downslope transport, which led to a relative increase in
analysis variation. chemically inert pisolite gravel in the solum. We believe that the dif-
ference between the two upper-slope soil members in terms of nutrient
2.4. Stock calculations stocks was primarily due to varying degrees of erosion. This is inferred
from the following two assumptions: 1) the thicknesses of the gravel
The nutrient stock (kg/ha) was calculated to the depth of 1m using layer is proportional to the upward movement of the fine-grained soil
the formula: matrix between the gravel, and 2) if the upward transport is similar,
then the thickness of the gravel-free termite deposited layer on top of
Nutrient content100 = (Thi 10000 BDi Ci) the gravel horizon indicates whether down-slope removal of soil ex-
i=100 ceeds the upward transport by termites. The thickness of the gravel
Th: thickness of the soil layer (m) layers (Table 1) suggests that the upward movement of soil is similar or
BD: bulk density (kg/m3) slightly higher at the Bekwai (~75 cm) compared to the Nzima
C: content of nutrient in %/100 (~65 cm). Hence, the different thicknesses of the termite deposited
layers (largely absent at Bekwai, ~20 cm at Nzima), suggests that the
3. Results down-slope removal of soil is higher than the upward transport at the
Bekwai compared to the Nzima, where a gravel-free and relatively
3.1. Soil physical and chemical characteristics fertile (based on cations) layer has accumulated on top of the gravel
horizon. The Nzima also had higher stocks of Ca and Mg than Bekwai in
Gravelly horizons dominate the two profiles at the upper slope, the fine-earth fraction (< 2mm, Table 3), while it was almost equal for
especially at the top of the solum (Table 1; Fig. 2c). In the reddish- K and P. This suggests that the Bekwai is also slightly stronger weath-
brown Bekwai, the gravelly layer began at the surface, while a 20 cm ered than Nzima, which is consistent with previous studies (Owusu-
thick gravel-free layer superimposed the gravelly horizons in the yel- Bennoah et al., 2000).
lowish-brown Nzima. The Kokofu soil was acid, gravel-free, yet less The Kokofu soil on the colluvial middle part of the slope was gravel-
clayey than the termite deposited material on top of the Nzima. The soil free as suggested by the conceptual model (Fig. 2). Yet, it was less
at the catena bottom was poorly drained and only slightly acid at the clayey than the termite sediment at the top of the Nzima (Table 1). We
top solum. Despite the wetness, the carbon content was low compared speculate that this is because the finer material containing most of the
to the soils upslope. An auger boring to 5m depth showed soft rock nutrients was transported all the way to the stream at the bottom of the
(saprolite) with quartz veins, and the 7% gravel below 37 cm depth catena during the rainy season, when most of the surface erosion occurs.
exclusively consisted of quartz. Quantifying the erosion over the year would be needed to check this
assumption. The Kokofu soil had intermediate contents of Ca, Mg and P
3.2. Nutrient concentrations in the fine-earth fraction compared to Nzima and Bekwai, which might
be expected, as it is formed by mixed material from these two upslope
For all four studied nutrients (Ca, Mg, K, P), the concentrations members. Thus, although the nutrient stocks including gravel was
decreased with depth (Table 2). At the top of the catena, Nzima had larger in the Kokofu than in the upslope soils, this was primarily caused
higher concentrations of Ca and Mg than Bekwai, while it was almost by the reduction in stocks due to termite activity and erosion upslope
equal for K and P. Kokofu at the colluvium generally showed lower rather than a positive impact downslope.
310
J.A. Kristensen, et al. Catena 178 (2019) 307–312
Table 1
Textures, organic matter contents, pH (H2O) and bulk densities for all horizons.
Profile Horizon⁎ Depth cm Clay <2 μm Silt 2–50 μm Sand 50–2000 μm Gravel > 2000 μm Organic matter % pH H2O Bulk density kgm−3
Bekwai Bτ(+A) 0–6 17 22 18 43 7.2 5.1 920
Eτ1 6–40 7 9 14 70 1.4 5.3 1090
Eτ2 40–75 7 10 13 70 1.3 5.2 890
B+C 75–100 51 35 14 2 0.8 4.8 1390
Nzima Bτ 0–20 31 41 25 3 2.9 5.8 1380
A+Eτ 20–55 12 13 11 64 1.6 5.9 1410
Eτ 55–83 15 8 13 64 1.1 5.8 1380
B+C 83–100 52 22 20 6 1.1 5.5 1370
Kokofu A 0–12 19 48 33 0 3.1 5.3 1400
E+B 12–49 21 46 33 0 1.0 5.2 1570
Bt 49–100 28 40 29 3 0.8 5.6 1660
Oda A 0–13 13 39 47 1 2.3 6.3 1290
Cg1 13–37 16 40 41 3 0.3 6.4 1680
Cg2 37–100 23 34 36 7 0.2 7.1 1760
⁎ According to Breuning-Madsen et al. (2004).
Table 2 The Oda soil at the catena bottom seems to have received negligible
Concentrations (‰) of calcium (Ca), magnesium (Mg), potassium (K) and amounts of particulate soil material from upslope. If this was the case,
phosphorus (P) for all horizons. AVG= average, SE= standard error on the relative clay and SOM accumulation would be expected, which con-
analysis (n= 3). tradicts our results (Table 1). Instead, we think the most likely ex-
Soil Horizon Chemical analyses, < 2000 μm planation for this net loss of fine material from the catena is that most of
the fine-grained material from upslope washed all the way to the stream
Ca, ‰ Mg, ‰ K, ‰ P, ‰ during the rainy season when the water erosion was highest and the
valley floor flooded, i.e. was flushed directly out of the system. How-
AVG SE AVG SE AVG SE AVG SE
ever, the higher content of nutrients in the Oda compared to upslope
Bekwai Bτ(+A) 1.634 0.049 0.401 0.002 0.108 0.028 0.015 0.001 might be due to the addition of dissolved nutrients from upslope, but
Eτ1 0.299 0.002 0.141 0.002 0.091 0.004 0.009 0.001 might also be due to nutrient enrichment during flooding.
Eτ2 0.290 0.040 0.118 0.009 0.094 0.005 0.009 0.001
B+C 0.181 0.023 0.071 0.007 0.049 0.007 0.006 0.002
Nzima Bτ 1.262 0.026 0.258 0.001 0.087 0.019 0.013 0.000
A+Eτ 0.841 0.035 0.298 0.006 0.083 0.023 0.010 0.001 4.2. Ecosystem consequences
Eτ 0.716 0.023 0.286 0.004 0.060 0.016 0.010 0.000
B+C 0.690 0.009 0.306 0.003 0.051 0.008 0.009 0.001 The ecosystem consequences of termite and ant activity are in-
Kokofu A 0.920 0.017 0.187 0.001 0.073 0.009 0.014 0.001
E+B 0.416 0.025 0.115 0.004 0.051 0.020 0.009 0.001 creasingly recognised in contemporary literature. Most studies conclude
Bt 0.668 0.018 0.172 0.002 0.055 0.012 0.009 0.000 that the effect of termites is positive, i.e. they provide valuable eco-
Oda A 1.336 0.026 0.349 0.002 0.042 0.007 0.021 – system services to humans (Jouquet et al., 2011; Kaiser et al., 2017).
Cg1 0.725 0.018 0.408 0.008 0.055 0.005 0.024 – Many such conclusions have been reached in savannah ecosystems,
Cg2 1.094 0.062 0.972 0.015 0.083 0.013 0.014 – where termites construct productivity hotspots through concentration
of nutrients and moisture in and around their nests and other physical
structures (e.g. Lobry de Bruyn and Conacher, 1990). Similarly, the
Table 3
Nutrient stocks to the depth of 1m. The left part shows the nutrient stocks positive effect the termite created soil macropores can have on soil
excluding the gravel (> 2mm), while the right part shows the nutrient stocks erosion control in such environments has been mentioned repeatedly.
including the gravel. In contrast, our results demonstrate that in undulating landscapes
macro-invertebrate activity may have a negative effect on landscape
Excluding gravel Including gravel scale soil fertility. P is of particular interest here, as it is often found to
kgm−2 kgm−2 be the limiting nutrient in tropical ecosystems (Reich and Oleksyn,
2004). We hypothesise that the 45–60% reduction in stocks due to the
Ca Mg K P Ca Mg K P termite mediated export of the fine-earth fraction from the catena, we
Bekwai 0.564 0.221 0.143 0.014 0.246 0.092 0.057 0.006 found here, results in an ecologically substantial reduction in plant
Upper slope available P at the landscape level in sloping environments, as these
Strongly catenas are widely described across the tropics (Ahn, 1970). This hy-
eroded pothesis should be tested in other such environments with considerable
Nzima 1.539 0.525 0.131 0.019 0.867 0.276 0.069 0.010 bioturbation, and should be accounted for in future evaluations of the
Upper slope
Slowly eroded ecological implications of termite activity. The increasing density of
Kokofu 0.968 0.246 0.089 0.015 0.951 0.241 0.088 0.015 termite mounds and decreasing soil depth with increasing elevation
Mid-lower observed along a catena in Brazil (Sarcinelli et al., 2009) and at an
slope inselberg in Ghana where we conducted a study previously (Kristensen
Colluvium et al., 2015) suggests that this process is widespread across tropical
Oda 1.762 1.328 0.124 0.010 1.664 1.245 0.116 0.009
Catena catenas. Moreover, bioturbation should be considered across taxa, as
bottom also vertebrate and floral bioturbation can result in similar downslope
In situ movement of soil material (Heimsath et al., 2002; Stockmann et al.,
weathering 2013).
311
J.A. Kristensen, et al. Catena 178 (2019) 307–312
5. Conclusion and redoximorphic features in a faulted landscape near Manaus, Brazil. Eur. J. Soil
Sci. 53, 203–217.
We assessed the combined effect of termite bioturbation and Funch, R.R., 2015. Termite mounds as dominant land forms in semiarid northeasternBrazil. J. Arid Environ. 122, 27–29.
downslope water erosion on soil nutrient stocks along a catena in a Grossman, R.B., Reinsch, T.G., 2002. Bulk Density and Linear Extensibility. In: Dane, J.H.,
moist semi-deciduous tropical forest in Ghana. The results suggested Topp, G.C. (Eds.), Methods of Soil Analysis: Physical Methods, Part 4. Soil Science
that this process reduced the nutrient stocks by up to 60% at the upper Society of America, Madison, pp. 201–228.Heimsath, A.M., Chappell, J., Spooner, N.A., Questiaux, D.G., 2002. Creeping soil.
slope due to relative accumulation of chemically inert gravel, while the Geology 30 (2), 111–114.
effect was negligible further downslope. This is most likely because IUSS Working Group WRB, 2006. World reference base for soil resources. In: World Soil
water erosion transported the majority of the fine-grained material Resources Reports No. 103. FAO, Rome.Jouquet, P. et al., 2011. Influence of termites on ecosystem functioning. Ecosystem ser-
directly to the stream at the base of the catena during the major rainy vices provided by termites. Eur. J. Soil Biol. 47: 215–222.
season, as the surface erosion peak coincided with flooding of the ca- Jouquet, Pascal, Bottinelli, Nicolas, Shanbhag, Rashmi R., Bourguignon, Thomas, Traoré,
tena base. Thus, the landscape scale net effect on soil nutrient stocks of Saran, Abbasi, Shahid Abbas, 2016. Termites: the neglected soil engineers of tropical
soils. Soil Sci. 181, 157–165.
this termite mediated downslope transport of fine-earth material Jouquet, P., Caner, L., Bottinellia, N., Chaudhary, E., Cheik, S., Riotte, J., 2017. Where do
(< 2mm) was negative and should be considered in future ecosystem South-Indian termite mound soils come from? Appl. Soil Ecol. 117–118, 190–195.
services evaluations. Kaiser, D., Lepage, M., Konaté, S., Linsenmair, K.E., 2017. Ecosystem services of termites
(Blattoidea: Termitoidae) in the traditional soil restoration and cropping system Zaï
in northern Burkina Faso (West Africa). Agric. Ecosyst. Environ. 236, 198–211.
Acknowledgements Kristensen, J.A., Thomsen, K.J., Murray, A.S., Buylaert, J.-P., Jain, M., Breuning-Madsen,
H., 2015. Quantification of termite bioturbation in a savannah ecosystem: application
This work was funded by the Danish development cooperation, of OSL dating. Quat. Geochronol. 30, 334–341.Lawson, G.W., Armstrong-Mensah, K.O., Hall, J.B., 1970. A catena in the tropical moist
DANIDA. We want to warmly acknowledge the contribution to this semi-deciduous forest near Kade, Ghana. J. Ecol. 58, 371–398.
work from our late colleague Professor H. Breuning-Madsen, who has Lobry de Bruyn, L.A., Conacher, A.J., 1990. The role of termites and ants in soil mod-
been moving Ghanaian soil science forward for more than three dec- ification: a review. Aust. J. Soil Res. 28, 55–93.Mujinya, B.B., Adam, M., Mees, F., Bogaert, J., Vranken, I., Erens, H., Baert, G., Ngongo,
ades. M., Van Ranst, E., 2014. Spatial patterns and morphology of termite (Macrotermes
falciger) mounds in the Upper Katanga, D.R. Congo. Catena 114, 97–106.
References Murphy, J., Riley, J.P., 1962. A modified single solution method for the determination ofphosphate in natural waters. Anal. Chim. Acta 27, 31–36.
Nahon, D.B., 1986. Evolution of iron crusts in tropical landscapes. In: Colman, S.M.,
Adu, S.V., 1992. Soils of the Kumasi Region, Ashanti Region, Ghana. Ghana Ministry of Dethier, D.P. (Eds.), Rates of Chemical Weathering of Rocks and Minerals. Academic
Food and Agriculture Scientific Services Division. Soil and Land-use Survey Branch. Press, Orlando.
Memoir 8, Kumasi. Nye, P.H., 1954. Some soil-forming processes in the humid tropics. A field study of a
Ahn, P.M., 1970. West African Soils. Oxford University Press, London. catena in the West African forest. J. Soil Sci. 5, 7–21.
Alexander, L.T., Cady, J.G., Whittig, L.D., Dever, R.F., 1956. Mineralogy and chemical Nye, P.H., 1955. Some soil-forming processes in the humid tropics. The action of the soil
changes in the hardening of laterite. In: Transactions of 6th International Congress of fauna. J. Soil Sci. 6, 73–83.
Soil Science, Paris. V, pp. 67–73. Obi, J.C., Ogunkunle, A.O., 2009. Influence of termite infestation on the spatial varia-
Awadzi, T.W., Cobblar, M.A., Breuning-Madsen, H., 2004. The role of termites in soil bility of soil properties in the Guinea savanna region of Nigeria. Geoderma 148,
formation in the tropical semi-deciduous forest zone, Ghana. Dan. J. Geogr. 104 (2), 357–363.
27–34. Owusu-Bennoah, E., Awadzi, T.W., Boateng, E., Krogh, L., Breuning-Madsen, H.,
Breuning-Madsen, H., Awadzi, T.W., Mount, H.R., 2004. Classification of soils modified Borggaard, O.K., 2000. Soil properties of a toposequence in the moist semi-deciduous
by termite activity in tropical moist semi-deciduous forests of West Africa. Soil Surv. forest zone, Ghana. West Afr. J. Appl. Ecol. 1, 1–10.
Horizon 45, 111–120. Reich, P.B., Oleksyn, J., 2004. Global patterns of plant leaf N and P in relation to tem-
Breuning-Madsen, H., Awadzi, T.W., Koch, C.B., Borggaard, O.K., 2007. Characteristics perature and latitude. Proc. Natl. Acad. Sci. 101 (30), 11001–11006.
and genesis of pisolitic soil layers in a tropical moist semi-deciduous forest of Ghana. Sarcinelli, T.S., Schaefer, C.E.G.R., Lynch, L.S., Arato, H.D., Viana, J.H.M., Filho, M.R.A.,
Geoderma 141 (1–2), 130–138. Gonçalves, T.T., 2009. Chemical, physical and micromorphological properties of
Breuning-Madsen, H., Kristensen, J.A., Awadzi, T.W., Murray, A.S., 2017. Early cultiva- termite mounds and adjacent soils along a toposequence in Zona da Mata, Minas
tion and bioturbation cause high long-term soil erosion rates in tropical forests: OSL Gerais State, Brazil. Catena 76, 107–113.
based evidence from Ghana. Catena 151, 130–136. Sarcinelli, T.S., Schaefer, C.E.G.R., Filho, E.I.F., Mafia, R.G., Neri, A.V., 2013. Soil mod-
Chapman, H.D., 1965. Cation exchange capacity. In: Black, C.A., Evans, D.D., White, J.L., ification by termites in a sandy-soil vegetation in the Brazilian Atlantic rain forest. J.
Ensminger, L.E., Clark, F.E. (Eds.), Methods of Soil Analysis. Part 2. Chemical and Trop. Ecol. 29, 439–448.
Microbiologial Properties. ASA, Madison, pp. 891–901. Stockmann, U., Minasny, B., Pietsch, T.J., McBratney, A.B., 2013. Quantifying processes
Christensen, E., Awadzi, T.W., 2000. Water balance in a moist semi-deciduous forest of of pedogenesis using optically stimulated luminescence. Eur. J. Soil Sci. 64 (1),
Ghana. West Afr. J. Appl. Ecol. 1, 11–22. 145–160.
Day, P.R., 1965. Particle fractionation and particle-size analysis. In: Black, C.A., Evans, Thomas, M.F., 1994. Geomorphology in the Tropics. A Study of Weathering and
D.D., White, J.L., Ensminger, L.E., Clark, F.E. (Eds.), Methods of Soil Analysis. Denudation in Low Latitudes. John Wiley and Sons, Chichester.
Agronomy 9. American Society of Agronomy, Madison, Wisconsin, pp. 545–567. Thomas, G.W., 1996. Soil pH and soil acidity. In: Sparks, D.L., Page, A.L., Helmke, P.A.,
Dijkerman, J.C., Miedema, R., 1988. An Ustult-Aquult-Tropept catena in Sierra Leone, Loeppert, R.H., Soltanpour, P.N., Tabatabai, M.A., Johnston, C.T., Sumner, M.E.
West Africa, I. Characteristics, genesis and classification. Geoderma 42, 1–27. (Eds.), Methods of Soil Analysis. Part 3. Chemical Methods. Soil Science Society of
Elberling, B., Breuning-Madsen, H., Knicker, H., 2013. Carbon sequestration in iron-no- America and American Society of Agronomy, Madison, pp. 475–490.
dules in moist semideciduous tropical forest soil. Geoderma 200–201, 202–207. Wilkinson, M.T., Richards, P.J., Humphreys, G.S., 2009. Breaking ground: pedological,
ELTRA, 1995. CS500 Simultaneous Carbon/Sulfur Determinator. ELTRA GmbH, Neuss. geological, and ecological implications of soil bioturbation. Earth Sci. Rev. 97 (1–4),
FAO, 1990. Guidelines for Soil Description, 3rd edition. Food and Agriculture 257–272.
Organization of United Nations, Rome. Wills, J.B., 1962. Agriculture and Land Use in Ghana. Ghana Ministry of Food and
Fritsch, E., Montes-Lauar, C.R., Boulet, R., Melfi, A.J., Balan, E., Magat, P., 2002. Lateritic Agriculture. Oxford University Press, London.
312