Content uploaded by Arne Erpenbach
Author content
All content in this area was uploaded by Arne Erpenbach on Jan 17, 2017
Content may be subject to copyright.
35
Flora et Vegetatio Sudano-Sambesica 19, 35-51 Frankfurt, December 2016
Termites and savannas –
an overview on history and recent scientific progress
with particular respect to West Africa and to the genus Macrotermes
Received: 2016-10-31; revised: 2012-12-14; accepted: 2012-12-16
Arne Erpenbach1 & Rüdiger Wittig1,2*
1 Institute of Ecology, Evolution and Diversity, Goethe University Frankfurt am Main, Max-von-Laue Str. 13, 60438 Frank-
furt am Main, Germany
2 Biodiversity and Climate Research Centre (LOEWE BiK-F), Georg-Voigt-Straße 16-18, D-60325 Frankfurt/Main, Ger-
many
* Corresponding author; email: r.wittig@bio.uni-frankfurt.de
Summary: Particularly in savannas, termites are ecosystem engineers and a keystone group in ecology. For the understand-
ing of the savanna vegetation, mound building termites are of particular interest. Due to their special soil chemistry and
physical structure, termite mounds often host other plants than the surrounding savanna. As our knowledge of the specific
contribution of mound-building termites to overall savanna diversity and ecosystem dynamics doubtlessly is not complete,
this paper summarises the state of the art in order to stimulate further research. According to the research interest of the au-
thors, focus is laid on the West African savanna and on the genus Macrotermes.
Key words: ecosystem engineers, Macrotermes, savanna, termite biology, termite mounds, vegetation pattern, West Africa
Termites et savanes - une vue d'ensemble de l'histoire et des progrès de la recherche avec
un accent particulier sur l'Afrique de l'Ouest et du genre Macrotermes
Résumé: Les termites sont des ingénieurs écosystémiques et un groupe “clé de voûte” en écologie, notamment dans les sa-
vanes. Pour comprendre la végétation des savanes, les termites à termitière présentent un intérêt particulier. En effet, de par
la spécificité de leurs caractéristiques de sol et de leur structure physique, les termitières hébergent fréquemment vegetation
differente de la savane environnante. Cependant, le rôle des termites à termitière dans la diversité [végétale] des savanes et la
dynamique de ces écosystèmes n’est que partiellement connu. Cet article propose donc un état des lieux des connaissances et
vise à stimuler la recherche sur cette thématique. L’accent est mis sur la savane d’Afrique de l’Ouest et sur le genre Macro-
termes, qui sont au cœur des travaux de recherche des auteurs.
Mots clés: ingénieurs de l'écosystème, Macrotermes, savane, termitières, Afrique de l'Ouest
Termiten und Savannen – eine Übersicht zu Geschichte und Fortschritten der Forschung un-
ter besonderer Berücksichtigung von Westafrika und der Gattung Macrotermes
Zusammenfassung: Insbesondere in Savannen fungieren Termiten als Ökosytem-Ingenieure und stellen damit eine ökolo-
gische Schlüsselgruppe dar. Besonders wichtig für Zusammensetzung und Struktur der Savannenvegetation hügelbauende
Termiten, denn die Termitenhügel beherbergen aufgrund ihrer besonderen Bodeneigenschaften oft andere Pflanzen als die
umgebende Savanne. Trotz einer Vielzahl publizierter Forschungsergebnisse sind die Kenntnisse zweifellos noch nicht aus-
reichend, um den Beitrag der hügelbauenden Termiten zu Diversität und Dynamik von Savannenökosystemen vollständig zu
verstehen. Der vorliegende Übersichtsartikel soll als Basis für weitere Untersuchungen dienen. Gemäß den Forschungsinter-
essen der Autoren wird der Schwerpunkt auf Westafrika und die Gattung Macrotermes gelegt.
Schlagworte: Ökosystemingenieure, Macrotermes, Savanne, Termitenbiologie, Termitenhügel, Vegetationsmuster, Westafri-
ka
1 Introduction
Savannas are heterogeneous and species-rich ecosystems that cover a large proportion of the global landmass, characteristi-
cally composed of a patch-mosaic landscape of trees and shrubs in a matrix dominated by C4 grasses. Termites, which are
particularly abundant in savannas ( 1971), are ecosystem engineers and a keystone group in savanna ecology
( 1990; et al. 1998). Even though their mounds cover only a small percentage of the savanna surface,
termites and their mounds provide a variety of direct and indirect resources and ecosystem services to various organisms,
including plants and animals, but also the local human populations. In addition to influencing key processes such as nutri-
ent flux and carbon cycling, the mound-building activity of certain termite taxa provides a prominent structural feature of
savanna landscapes. Termite mounds often support different plants than the surrounding savanna matrix and are explicitly re-
ferred to as a specific habitat of many plant taxa in floristic works. Due to their special soil properties, physical structure, and
36
Flora et Vegetatio Sudano-Sambesica 19
in the tropics to the beneficial activities of temperate earth-
worms, inspired by Darwin’s work on the latter (
1881). Drummond’s work, however, was largely ignored
and disregarded, besides some rare contributions from the
field of vegetation ecology. Nearly six decades later, a re-
view by (1943) called for further research on the
question whether termite activity is detrimental or beneficial
to the ecosystem on a longer timescale, and addressed open
questions. He particularly emphasised the need for research
re-evaluating the effects of termites on soil fertility and
soil erosion, and he remarked that large abandoned termite
mounds may show increased fertility.
With the availability of new analytical methods and their ap-
plication in the field, numerous studies on the influence of
termites on soil have been published since then. In the sec-
ond half of the 20th century, research articles started to pro-
vide pedological data on the composition of termite mound
soils ( 1955; 1955; 1962). These have
emphasised the relevance of termites for ecosystem func-
tions, and also confirmed (1943) conclusions
that different groups of termites may have very different ef-
fects on their environment due to their differing ecology and
behaviour.
2.2 Termite ecology and classification
2.2.1 Termite diversity
Termites are a group of insects currently comprising approx-
imately 2900 described species ( et al. 2013), with
an expected 500 to 1000 species remaining to be described
( 2011). However, given that a recent study found
20 putative new termite species in Pendjari National Park,
West Africa, alone ( et al. 2011), this estimate
is likely conservative. Termites are currently ranked either
as infraorder Isoptera or epifamily Termitoidae, but are phy-
logenetically securely placed within the order Blattodea,
more commonly known as cockroaches (-
2013; et al. 2013). The genus Macrotermes
belongs to the monophyletic Macrotermitinae. However, for
our understanding of the impact of termites, and in particu-
lar Macrotermes, on the savanna ecosystem, the established
functional classification is more relevant than the intra-or-
der taxonomy.
2.2.2 Feeding groups
Termites show diverse nesting behaviour and feeding hab-
its. Functionally, they are separated into different feeding
guilds. The most recent categorisation, supported by mor-
phological and phylogenetic data ( et al. 2001),
distinguishes between (phylogenetically) lower-level wood-
feeding termites (group I), grass-, wood-, litter-, and lichen-
feeding termites (group II, including the fungus-cultivating
species), soil/wood-interface-feeding termites (group III)
and true soil-feeding termites (group IV). It should be noted
that different feeding groups have different impacts on their
environment, but also that the vast majority of termites live
in the soil environment and provide similar ecosystem ser-
vices ( et al. 2006, 2011; et al. 2015).
The nutritional habits of termites are probably a major driv-
characteristic vegetation, termite mounds are thus important
drivers of environmental heterogeneity, which is considered
a major driver of species diversity worldwide ( et al.
2004; et al. 2014).
Despite the acknowledged importance of termites as eco-
system engineers, doubtlessly our knowledge of the specific
contribution of mound-building termites to overall savanna
diversity and ecosystem dynamics is not complete. There-
fore, in order to stimulate further research, this paper sum-
marises the state of the art.
Literature concerning the influence of termites on ecosys-
tems often integrates results and generalises conclusions
across systems (even across biomes), and also across termite
species or broader termite groups. As termites are taxonomi-
cally and ecologically diverse, we start with a short intro-
duction to termite biology (section 2), then generally treat
the influence of termites on savannas (section 3), focus on
the influence of Macrotermes and the vegetation of the West
African savanna (section 4) and summaries some own re-
sults concerning the influence of termite-induced hetero-
geneity on savanna vegetation along a climatic gradient in
West Africa (section 5).
2 Termite biology – a short overview
2.1 Research history
Ever since the early naturalists’ explorations during the 18th
century, termites have occupied a prominent place in writ-
ings devoted to tropical environments. The respective sci-
entific literature can be divided into two major branches
( 2005). One focusses on the short-
term consequences of termite activity, which are mostly re-
garded as detrimental to agriculture, silviculture, and human
constructions. This branch, widely concerned with termite
control and mitigation of termite damage, produced a large
body of publications reaching back to colonial times. More
relevant to this paper is the second branch of literature,
which focusses on the longer-term perspective of natural
history, including the function of termites in tropical ecosys-
tems. While their often destructive impact on human efforts
is acknowledged, in this branch of the literature termites are
also seen as fascinating study objects. This view already
found expression in the first scientific accounts of termites
by (1779) and (1781), both concerning
Macrotermes. These ground-breaking publications dealt
with the description, classification and systematics of ter-
mites, and also delivered vivid and insightful accounts of
their natural history, including the intricate construction of
termites’ nests, and the often complex social organisation of
their colonies. The admiration of the early naturalists for ter-
mites becomes evident in a comparison of the architectural
achievement of erecting large termitaria as surpassing the
human effort to build the Great Pyramid of Giza (-
1781).
(1781) also commented on beneficial effects
of termites, such as the removal of dead biomass, but these
were thereafter mainly ignored or dismissed for a century
until (1886) compared the influence of termites
37
Erpenbach & Wittig
the soil, in living plants, or in dead wood. Some species’
nests are as simple as a few tunnels inside a trunk, but oth-
ers’ reach a degree of sophistication and size unrivalled by
other constructions in the animal kingdom. Many species re-
inforce or cover their nests with plant-derived carton, while
others use soil cemented by saliva and faeces, or a combina-
tion of soil and carton. Nests that are covered by soil and
protrude above the soil surface are called termite mounds
or termitaria. Several other synonyms exist, especially in
older literature (see, e.g., 1979). The most impos-
ing mounds are the large termitaria of the savannas, with the
largest and most elaborate mounds built by the genus Mac-
rotermes, which we focus on in this paper.
2.3.2 Mound functions
It has been argued that termite mounds constitute a part of
a colony superorganism in so far as they are a functional
unit and no part can survive without the others (
2011). Besides protection against predators, termite mounds
protect the colony from direct intrusion of rainwater and
flooding, which would otherwise have catastrophic conse-
quences. Damages by, e.g., rain or other organisms, are con-
stantly repaired. The major challenge for a termite colony,
however, is minimising variations in temperature and hu-
midity. Especially in the fungus-cultivating genera, the most
important function of the mound is creating and maintaining
homoeostasis for the colony and their symbionts. Termites
in this monophyletic group have an obligate mutualistic
relationship with basidiomycete fungi of the genus Termi-
tomyces ( et al. 2002; 2005),
cultivating these fungi inside the mound. Macrotermes cre-
ate "fungus combs" with a high surface area out of partly-
digested plant material in special fungus chambers of their
mounds and inoculate them with fungal spores in a high
density. These fungus gardens are then constantly nursed,
preventing infection by other biota. While the termite work-
ers already can take advantage of easily digestible parts of
the plant material, as e.g. short-chained carbohydrates, the
colony depends on the fungus regarding further nutritional
needs, especially considering continuous nitrogen supply
throughout the year.
Homoeostasis of the mound environment might be more
important for the fungus than for the termites themselves
( 2003; 2005), which because
of their thin cuticles are very prone to dessication, but can
cope by relocating quickly. In fact, termite workers actively
transport water from the groundwater table into the mound,
travelling as far as 50 m below ground ( et al. 1974);
it is however mainly the architecture of the mound which
ensures a high and constant air humidity. Two possible and
not mutually exclusive hypotheses are that macropores cre-
ated by termites around the mound increase infiltration and
allow percolation of water towards the mound, and that
higher clay and silt content in the mound increase the soil
water matrix potential to wick up and hold humidity from
the surrounding soil ( 2006).
Constant temperature and a continuous gas exchange in the
mound are also of particular importance for Macrotermes,
since Termitomyces has a narrow temperature tolerance and
ing force behind their evolution. Termites have co-evolved
with protists, bacteria, and fungi to digest lignocellulose,
an extremely abundant biopolymer which very few animals
can take advantage of. In cases where both phylogenies
are known, the phylogeny of their symbionts matches the
termite phylogeny very well (see, e.g., et al. 2002;
et al. 2002; 2005;
2011). Macrotermes, like all other Macro-
termitinae, are fungus cultivators which co-evolved with the
fungus genus Termitomyces. The mutualistic symbiosis is
obligate for the termites, and Macrotermes mounds are an
evolutionary result of this symbiosis (
2005).
2.2.3 Caste differentiation
In contrast to most other Blattodea, all termites are eusocial
insects. They live in nests with communities of up to several
million individuals, which are usually all descendants of
one founding couple, also called the royal couple, consist-
ing of queen and king. The queen of some species can reach
a remarkable age of more than a decade, continuously pro-
ducing offspring. The vast majority of termite species have
castes fulfilling specific tasks in the colony. Different types
of workers construct galleries and enlarge the nest, feed the
other castes, tend offspring, and, in the case of fungus-culti-
vating species like Macrotermes, tend the fungus. Depend-
ing on the species, workers sometimes undergo multiple
stages in their lifetime, fulfilling different tasks at different
time periods in their development. The reproductive alates
are dispersal units, turning into kings and queens when mat-
ing to found new colonies. Alates are only produced after
the nest has reached a certain size, which makes large in-
vestments in the protection of the nest and especially the
royal couple adaptive from an evolutionary viewpoint. The
soldier caste is concerned with aggressive defence. Soldiers
often have enforced head capsules and well-developed man-
dibles. Some clades have chemical defences instead, with
soldiers spraying defensive substances on attackers (-
1984). Interestingly, soldiers show altruistic behav-
iour, attacking intruders to stall them while workers close
off passages into the nest’s interior, preventing both intru-
ders and the defending soldiers from entering the nest. Also,
in at least one species, senescent workers literally explode
to distribute toxic chemicals to stop intruders (
et al. 2012). While their colonies are thus not completely
defenceless, termites are still comparably easy prey. Their
main mode of defence is keeping predators out of the nest
and ensuring the persistence of the colony by protecting the
often completely immobile queen. The nest is therefore an
integral part of the termite community, which is especially
compelling in case of the mound building genus Macro-
termes.
2.3 Termite mounds
2.3.1 Nests and mounds
Most, but not all termites construct complicated structures
to enclose their nest as Macrotermes does. Many species’
nests are inconspicuous to the human observer, hidden in
38
Flora et Vegetatio Sudano-Sambesica 19
its metabolism is inhibited by CO2 ( 2011). While aer-
ation, gas- and temperature flux within mounds are not yet
fully understood, it is clear that the mounds’ internal struc-
ture of galleries and tunnels is functionally highly adaptive
( 2003). Generally, internal and external architecture
evolved certain patterns within each species, derived from
self-organisation processes. Further self-organisation pro-
cesses ensure that both internal and external architecture can
also be modified in adaptation to environmental conditions,
and change over the lifetime of a colony (-
1998b, 1999; 2011).
In summary, termites are a moderately large but understud-
ied group of social insects. The focus group in the context
of this paper, the genus Macrotermes in West Africa, con-
structs large mounds. The main function of Macrotermes
mounds is to provide a protected and stable environment
for the termite nest, including the fungus gardens. Termites
transport and modify large amounts of soil over the lifetime
of a colony to build and maintain their mounds, continuous-
ly engineering the soil environment of savanna landscapes
in the process.
2.4 Focus species
2.4.1 Species distribution
This paper focusses on two species of the fungus-cultivating
genus Macrotermes. Macrotermes bellicosus (Smeathman
1781) and Macrotermes subhyalinus (Rambur 1842) are
widely distributed in West and East Africa. While M. sub-
hyalinus occurs through-out West African savannas, M. bel-
licosus seems to reach the northern limit of its distribution
within the North Sudanian vegetation zone (sensu
1983; 1984) and only rarely is found in the Sahel
vegetation zone.
2.4.2 Variability of mound types
It is difficult to identify termite species in the field, and ter-
mite individuals, preferentially from the solider caste, are
needed for species determination. In general, the two Mac-
rotermes species present in West Africa build differently
shaped mounds. According to (2011), M. subhyalinus
generally builds mounds featuring openings for ventila-
tion, while M. bellicosus always builds completely closed
mounds. In our study area, M. subhyalinus mostly builds
relatively flat hillocks with thick walls, while M. bellicosus
mounds are more diverse and range from domes to "cathe-
dral" mounds with ridges and turrets, reaching heights of six
meters and more.
However, mound architecture of the two species can over-
lap, as each species shows regional as well as local variabil-
ity, with mounds being adapted to local environmental con-
ditions ( 1979; 1998b; et
al. 2009a). Mounds also can be re-colonised multiple times
after the death of a colony, and by different species. Further-
more, the majority of vegetated mounds are uninhabited and
heavily eroded. Thus, it is often difficult to identify the spe-
cies that originally built or at least occupied each mound.
Therefore, in the following, we do not differentiate between
M. subhyalinus and M. bellicosus.
2.4.3 Soil composition of Macrotermes mounds
While most termite species from all feeding groups manipu-
late the soil environment, the extent and effects of their ac-
tions differ. Between-group differences in effects on the soil
environment are generally poorly understood. However, the
availability of data for Macrotermes is comparably good,
and M. bellicosus and M. subhyalinus in particular are com-
parably well studied.
One feature generally attributed to Macrotermes colonies is
an enrichment in nutrients in the mound. Macrotermes are
one of the few organisms which, with the help of their fun-
gal partners, can completely decompose plant biomass, in-
cluding the recalcitrant lignocellulose, reportedly leading to
nutrient accumulation in mound soil ( et al. 2012).
Estimates in the literature indicate that termites may process
40 to 100% of dead wood in the ecosystem ( et al.
2013), with Macrotermes having the greatest impact.
Macrotermes also move large quantities of soil for mound-
building and construction of soil sheetings to protect for-
aging galleries, and therefore contribute to soil turnover.
Both sheetings and mound material are enriched in clay
( et al. 2010), and certain types of clay have shown
to be actively selected by termites ( et al. 2002a,
2004, 2007). Some evidence even exists that termites may
be able to change clay mineralogy, thus changing its swell-
ing and shrinking characteristics ( 1982; et
al. 2002b, 2007).
In addition to structural adaptation to local microclimate
( 1998a,b), recent studies have shown
that composition of mound soil varies within populations of
M. bellicosus according to local topography and hydrology
( et al. 2009a,b). Further influence of local geological
and pedological conditions on composition of mound soil
has been reported for the closely related congeneric species
M. falciger (Gestäcker 1891) in Central Africa ( et
al. 2010, 2013, 2014; et al. 2015a), supporting the
idea that the impact of termites on soil turnover and soil
composition depends on the surrounding environment.
3 Interrelationship between termites and
the savanna ecosystem
3.1 Research history
The first descriptions of the association of certain plant
species with termite mounds date from the beginning of
the 20th century ( 1908; 1913;
1915). These works also included speculations that chang-
es in soil conditions due to termite activity were the reason
for these associations. The influence of termites and their
mounds on the vegetation at a landscape scale was then re-
cognised in an essay by the geographer (1936). Troll
was among the first to describe the influence of Macroter-
mitinae on the vegetation structure of East African savan-
nas. The importance he attributed to these insects is obvious
from the fact that he labelled the corresponding ecosystems
"termite savannas". Incidentally, his description of distinct
clusters of termite mound vegetation defining the character-
istics of a whole landscape predated his seminal coining of
39
Erpenbach & Wittig
the term "landscape ecology" ( 1939). Further stud-
ies of landscape ecology in African savannas in the 1940s
( 1943; 1947; et al. 1948) ex-
plored relationships between soils and vegetation, including
the influence of termites. Later detailed studies of savanna
vegetation were specifically interested in the synecology of
plant communities on termite mounds ( 1952; -
1954; 1963; 1965;
1968). The species lists and community descrip-
tions in these works, as well as occasional remarks like
"[growing] on termite mounds" in regional floras provide a
baseline for further ecological research. In a parallel branch
of investigations, pedological studies analysed chemico-
physical composition of termite mound soils ( 1955;
1955; 1962). Most recently, information from
different scientific fields began to get integrated to address
the role of termites in the savanna ecosystem (
1971; 1978; 1988; et al. 2000;
et al. 2010; et al. 2011), and the current
scientific debate tries to identify the processes behind pat-
terns in the savanna ecosystem which are caused by termites
and their mounds (e.g., et al. 1998;
et al. 2010a; et al. 2011; 2012a,b;
et al. 2015). However, even today, termites can
be seen as “neglected” soil and ecosystem engineers (-
et al. 2016).
3.2 Importance of termite mounds for the savanna eco-
system
Termites affect the savanna ecosystem both directly and
indirectly. In a direct way, termites influence biomass turn-
over in the savanna by processing large amounts of organic
matter. It is estimated that termites consume and metabo-
lize organic material at the same rate as herbivores of the
savanna regions ( 1971; 1978;
et al. 2000), thus contributing greatly to dead biomass
removal and biomass mineralisation. Termites also affect
the savanna system through their mound-building activity,
as their mounds create heterogeneity in availability of wa-
ter, nutrients and minerals. By transporting and manipulat-
ing clay-enriched soil for mound and soil sheeting construc-
tions from the subsoil to the surface, they contribute to soil
turnover and mineral redistribution in the savanna. The con-
struction of mounds also can trigger cascades of effects and
feedback loops in the soil environment, based on abiotic and
biotic environmental processes ( et al. 2006; -
et al. 2015). For example, higher air humidity and
better aeration changes microbial activity in mounds, affect-
ing nitrogen availability to plants ( et al. 2006;
et al. 2010). Interactions of other organisms with
termite mounds, especially plants, sum up to ecosystem
engineering ( et al. 1997; et al. 1998;
et al. 2006, 2016). Plants and animals, including
human populations, greatly profit from termites and their
mounds, which makes it particularly important to better un-
derstand the ecology of mounds systems.
3.2.1 Water availability
Water availability on termite mounds is usually thought to
be higher than in the surrounding savanna, as a result of
increased water percolation towards mounds, wick effects,
and increased water-holding capacity in mounds due to clay
enrichment ( 2006). Soil macropores created by ter-
mites can lead to highly increased infiltration rates, but the
effect seems to be variable between species and to depend
on termite activity patterns ( -
1990; 2001; et al. 2004;
et al. 2013; 2014). Despite the known ef-
fects of termites on soil moisture, only one study has em-
pirically demonstrated the effect of higher water availabil-
ity for plants on mounds as compared to the surrounding
savanna ( et al. 1999). The latter paper, which sug-
gests higher leaf-shedding to be responsible for the higher
water availability on mounds is widely cited, but is derived
from a very small number of samples: five mounds for leaf-
shedding differences and only one mound for water poten-
tial measurements.
The mound’s pediment, or halo, has been shown to have low
infiltrability due to surface sealing by silt and clay eroded
from the mound surface, and water transported towards the
mound’s centre from around and below the mound bypasses
the pediment ( et al. 2015; 2006).
3.2.3 Nutrient enrichment
There is much more data on nutrient content of mounds,
which is often higher than in the surrounding soils. Termites
are generally thought to create nutrient hotspots through ac-
cumulation and mineralisation of biomass in their nest and
through soil transport ( 1977; et al. 1983;
et al. 1988; 1988; et al. 2009;
et al. 2008; et al. 2010). The evidence for en-
richment of mound soil with organic matter, nitrogen, and
phosphorus is ambiguous ( et al. 2004; et
al. 2010) and varies between species and studies from dif-
ferent localities. Nutrients are slowly released from mounds
by leaching and erosion, making termite mounds important
for plants both on and off mounds. Some evidence exists
that elevated nutrient concentrations, including potassium,
in mound soil are of organic origin ( et al. 2008), or at
least subsidised by mineralisation of organic matter. How-
ever, like most detailed studies of mound material, the re-
sults of et al. (2008), that higher potassium content in
mounds is due to mineralisation of plant matter by termites,
were derived from a small sample size (four mounds), and
probably do not reflect broader environmental heterogeneity
of termite mounds and their savanna matrix .Other studies
assume that nutrients derived from plant material transport-
ed to the mound by termites might be retained in the mound
centre and thus partly protected from leaching ( et al.
2015a; et al. 2012). However, until recently, it was
not yet known whether nutrient input by termites or clay
content is more important for soil fertility of termite mounds
( et al. 2010).
3.2.4 Clay enrichment
An enrichment in clay and silt compared to the surround-
ing savanna topsoil has been reported by most studies of
termite mound soil ( 1955; 1955; 1962,
1975; 1978; 1988; et al.
40
Flora et Vegetatio Sudano-Sambesica 19
2009, 2010), which might strongly influence nutrient avail-
ability to plants. Termites select clay and silt particles from
the subsoil and transport them to the surface to construct
their mounds and foraging gallery sheetings ( et
al. 2002a; et al. 2009b, 2012), and are probably able
to alter clay mineralogy, creating highly expandable clays
( 1982; et al. 2002b). Studies often remark
on higher levels of exchangeable bases and higher pH in
termite mound soil, which is directly related to cation ex-
change processes involving high-surface clay particles.
Higher clay content may strongly slow leaching of nutri-
ents. Thus, concentration and manipulation of clay particles
by termites can influence nutrient availability to plants on
mounds. Plants in the surrounding savanna may profit from
prolonged leaching as well as from slow and continuous
erosion of mound material.
As a combined results of three studies ( et
al. 2013, 2014, 2016) described in section 5) clay enrich-
ment can be identified as the most influential factor for the
mound–savanna contrast across savanna types. It is, how-
ever, important to emphasise that only the interplay of clay
with other factors, including the plant macronutrients po-
tassium and phosphorus, was able to explain a large part of
the vegetational contrast. Selective transport of subsoil clay
particles, mineralization of plant biomass, and possibly also
manipulation of clay mineralogy by termites are important
for changes in the availability of major and minor cations in
termite mound soil. In addition to macronutrients and ma-
jor cations, plant micronutrients, including trace elements,
rare earth elements, and heavy metal elements, have been
found in higher concentrations in termite mound soil than
in the surrounding topsoil ( et al. 2008, et al.
2009, et al. 2014). Thus, higher cation exchange
capacity on mounds is a direct result of termite activity,
and increased levels of cations might continue to be avail-
able from the mound soil for a prolonged time ( et al.
2015a; et al. 2013).
3.2.5 Soil turnover
For West Africa, the contribution of termite mounds to sa-
vanna soil turnover by erosion has been estimated between
0.75 and 1 mm ha-1 mound material per year, equivalent to
about 7.5 to 10 m3ha-1 ( 1984). Assuming that ero-
sion continues undisturbed, (1984) estimated that
it would take 20 to 25 years to completely erode a single
large mound of M. bellicosus. However, termite mounds can
last for several decades, or even centuries ( 1967;
et al. 2010; et al. 2015b), depending on re-
colonisation and local environmental conditions. Besides
their contribution to soil turnover by erosion, termites and
their mounds can cause larger-scale physical changes to
landscape structure.
3.2.6 Ecosystem engineering and effects on vegetation
Termites have been described as ecosystem engineers due
to their ability to induce changes in whole landscapes via
multiple feedback loops, emphasising plant–soil feedbacks
on mounds ( et al. 1997; et al. 2006; -
et al. 2012). Plant communities on mounds have often
been reported to be different from the surrounding savannas
and to show higher diversities on mounds (e.g. et al.
2009), and this is mostly attributed to higher fertility of ter-
mite mound soil ( et al. 2010).
A widely cited paper by et al. (1998), the key-
stone to a series of studies from the Okavango delta (
et al. 1993; et al. 1993, 1998, 2012), showed
that the construction of termite mounds leads to the forma-
tion of vegetated islands: the combined effects of particular
changes to soil composition and vegetation lead to further
soil changes due to physico-chemical soil precipitation pro-
cesses. This, in turn, contributes to enlarging the island and
allows further plants to colonise, which further stabilise the
emerging microhabitat with their roots, and cause further
precipitation processes due to evapotranspiration.
While the local environmental conditions in the Okavango
delta are very specific, similar processes seem to take place
to a varying degree in other localities as well, and precipi-
tates in termite mound soil have not only been found in the
Okavango. Most commonly, studies have reported mounds
to contain carbonates, with mounds sometimes developing
localised nodules or crusts even in otherwise carbonate-
free surroundings ( 1947; 1955; 1962,
1974; 1979; et al. 2007; et al. 2011).
Furthermore, the formation of sesquioxides in mounds has
been observed in strongly seasonal localities with a high
groundwater table ( et al. 2011, 2013, 2014), indi-
cating the influence of local environmental conditions on
mound composition.
Very important for the savanna ecosystem are effects of
ecosystem engineering related to vegetation. The particular
soil conditions of mounds are thought to promote the es-
tablishment of specific species and functional groups, such
as trees. For example, it has been suggested that termite
mounds are focal points of forest species establishment in
savanna ecosystems ( 1963; et al. 2008b), and
that they offer microhabitats different from the surrounding
savanna for plant colonisation ( et al. 2009; et
al. 2010).
In addition to promoting plant establishment, termite
mounds are focal sites for woody plant regeneration
( 2008; et al. 2008a, 2015) and
help plants to survive adverse environmental conditions
like fire, drought and inundation ( et al. 2008b).
Microclimatic effects like shading and protection from fire
and wind by already established vegetation further promote
plant survival on mounds ( 1967;
et al. 1973). Termite mounds often harbour succulents, xe-
rophytes and evergreen plant species (e.g., 1943;
1975; 1975-
1976), which are especially sensitive to fire. Succulents are
most often found growing in the centre of mound thickets,
on top of the mounds ( et al. 2012), where they
are most protected from fire. Microtopography of mounds
and the concretion of the outer soil layer of mounds lead to
high runoff rates and low infiltration rates on the mounds
themselves, which might be an environmental filter favour-
ing xerophytic adaptions. However, under wetter habitat
conditions, microtopographic elevation above the sur-
41
Erpenbach & Wittig
rounding area and the mounds’ internal tunnels and gal-
leries might help plants to escape root asphyxiation in the
case of flooding ( 1988; et al. 1998).
Vegetation can also benefit from higher water availability in
mound soil, as foraging tunnels around the mounds lead to
higher infiltration and percolation towards the mound, and
clay enrichment leads to higher field capacity of mound
soil ( 2006). As mentioned in the case of soil pre-
cipitation processes, vegetation can cause further feedback
loops, for example by acting as collectors of aerosols and
enhancing soil structural integrity with their roots (-
et al. 2012). Further feedback may happen due to re-
mineralisation of plant biomass contributing to mound soil
fertility, which has been reported to be already elevated as
compared to the surrounding savanna due to termite activity
( et al. 2006, 2011; et al. 2010). Termites
from the original mound building species as well as other
species can be involved in mineralising plant biomass grow-
ing on mounds, thus profiting from the effects of ecosystem
engineering themselves.
Published germination experiments on mound soil are
scarce, but mound topsoil has been shown to have a high
mechanical impedance to root penetration ( et al.
2007; et al. 1999), hampering seedling establish-
ment. Unfortunately, handling of the soil in experimental
studies leads to further soil compaction, sometimes limiting
the significance of experimental results for our understand-
ing of the processes involved in seedling establishment un-
der natural conditions ( et al. 2007; et
al. 2012; 2010; 2013). Still,
soil compaction on mounds is likely to favour certain plant
phenotypes.
Regarding the rhizosphere, it has moreover been reported
that Borassus palms in West Africa direct their root foraging
towards nutrient-rich patches resulting from termite mounds
( et al. 1996), and that greater rooting depths oc-
cur on mounds in Ghana ( et al. 1973). Seedling es-
tablishment and survival thus depend both on the soil condi-
tions of mounds and species-specific phenological and mor-
phological traits.
Recent results suggest that disturbances by fire and herbi-
vores interact to maintain community diversity both in the
savanna and on mounds, but the differences in community
patterns between mounds and the surrounding savannas
have been attributed to termite-induced soil changes (-
& 2012a; et al. 2013; et al. 2015,
et al. 2016).
3.2.7 Effects on animals and their feedbacks with vege-
tation
Termite mounds, whether active or vegetated and eroded,
are attractive for a wide range of mammals (
2003; 2004; et al.
2013). Large herbivores have been discussed as maintain-
ers of termite-induced heterogeneity in the savanna as con-
sumers of biomass ( 2012a,b; et al.
2013), but they might also play a considerable role in seed
dispersal ( et al. 2015; et al. 1999). Also Pri-
mates may be important for seed dispersal to mounds, but
there is a lack of systematic observations regarding their
utilisation of mounds and mound vegetation. In West Afri-
can savannas, the genus Papio in particular is known to be
an effective dispersal agent for a number of species occur-
ring on or even characteristic for termite mounds, like Dio-
spyros mespiliformis, several species of Grewia, and Tama-
rindus indica ( 2008; et al.
1979). Further ethological as well as ecological studies on
primates’ relations to termite mounds are needed. Consider-
ing small mammals, some studies mention that rodents use
abandoned termite mounds for seed hoarding (
2002; et al. 2002). For Macrotermes, the focus
genus of this paper, the possible role of termites as disper-
sal vectors, as suggested by et al. (2015), can quite
safely be dismissed because, unlike many species of ants,
Macrotermes do not transport and store seeds in their nests.
Termite mounds also affect other animals, both directly and
indirectly through effects on the vegetation. Mounds them-
selves, and the often more dense vegetation on mounds
compared to the surrounding savanna, provide food resourc-
es and sites for shelter, burrowing, foraging, thermoregula-
tion, roosting, and breeding, as has been found for reptiles
( et al. 2010), birds ( et al. 1999; et al.
2011), and mammals ( 2003). Several
mammal species feed on termites, and a variety of animal
species perch on mounds, take refuge in their cavities, di-
rectly exploit the mound material for minerals through ge-
ophagy, and prefer browsing or grazing on plants growing
on termite mounds (e.g., 1994;
2004; 2004; et al.
2005, 2006; et al. 2010;
et al. 2010b; but also see et al. 2013;
2013; et al. 2013). The interactions of
mounds, plants, and animals are suspected to cause feed-
back loops, similar to those observed for large trees (
et al. 1999). A summarising conceptual system analysis of
this view, focussing on feedbacks creating and maintain-
ing habitat heterogeneity of the savanna landscape, can be
found in et al. (2010). Thus, both for animals and
plants, termite mounds can be described as keystone struc-
tures sensu et al. (2004).
3.2.8 Benefits to humans – the example of West Africa
Several plants used by local communities grow preferen-
tially or exclusively on termite mounds. Some of these
plants, and also the mounds themselves, are sometimes
of significance in traditional belief systems ( et al.
2009), while others are harvested for medicinal purposes or
as foodstuff. It is also common to harvest termites direct-
ly from the mound as poultry fodder, and to capture large
amounts of alates on their nuptial flight for animal as well as
human consumption, depending on local tradition (
et al. 2009). Under certain seasonal climatic circumstances,
the basidiomycete fungi cultivated by the subfamily Macro-
termitinae can produce fruit bodies, which are an important
source of protein and are also used in traditional medicine
( et al. 2009; et al. 2011, 2013). Also, in sev-
eral parts of Africa humans, especially children and preg-
nant women, are known to consume termite mound soil; this
may have medical benefits as a result of its elevated mineral
42
Flora et Vegetatio Sudano-Sambesica 19
contents, however research so far is inconclusive (
1997; 2003).
Clay-rich termite mound soil is also often used for construc-
tion or production of clay bricks throughout Africa, and
is also used to amend soils by spreading mound material,
which has been reported to increase soil fertility and sup-
press crop infection with parasitic plants ( 1977;
et al. 2007; et al. 2009; et
al. 2012). Furthermore, humans can actively promote eco-
system engineering by termites through agricultural prac-
tice, and termites have successfully been used in soil resto-
ration, e.g. in case of the traditional Zaï system in Burkina
Faso ( et al. 2008; 2011;
2014). Termite mounds have also been used for prospecting
of gold and other minerals. Already Herodot (2500 BP) re-
lates in his ’History’ that according to the Persians most of
the gold collected in India would be carried up with sand
by ’ants’ dwelling under ground. Even though it is doubt-
ful if this description of gold exploitation indeed refers to
termites, contemporary research has shown that analysis
of termite mound soil can be used to detect gold anomalies
( 1972; et al. 2009).
Although it is clear that termites provide many benefits
to humans, the contribution of termites, termite mounds,
and other species depending on termite mounds to house-
hold income and human sustenance is yet to be quantified
( et al. 2009). As far as we are aware, ecological con-
sequences of human utilisation of termites, termite mounds,
and their products (including, e.g., NTFPs) have rarely been
studied so far. Considering the magnitude and speed of
land-use change in most parts of Africa, we suggest that sus-
tainable and locally adapted management of the resources
provided by termites would be necessary, and would thus
welcome further research informing resource management
decisions.
4 Termite mounds and vegetation patterns
4.1 Differences between mounds and the surrounding
savanna
Most of the differences between the vegetation on mounds
and of the surrounding savanna have been mentioned in the
frame of treating the interrelationship between termites and
the savanna ecosystem (section 3; particularly 3.2.6 and
3.2.7). Some aspects not mentioned above are:
The majority of characteristic mound species in West
Africa is woody, which is in accordance with the no-
tion that termite mounds are focal sites for woody
plant regeneration as well as refugia for large trees to
reach senescence ( et al. 2011, 2013; et
al. 2008b).
Other characteristic mound species are geophytes,
suggesting that water availability on mounds is not
uniformly high, especially not in the compact topsoil
layer. In addition to having higher water-storage capac-
ity, geophytes in their dormant stage are likely more
resistant to cracking/swelling cycles between dry and
wet season, which might be another advantage in the
face of higher expansible clay content in mound and
pediment soil ( et al. 2002a,b, 2004;
et al. 2013).
Studies analyzing the spatial distribution of species
on mounds have shown spatial differentiation within
mound microhabitats ( 1967; -
et al. 2012; 2013), offering soil prop-
erties, microclimatic effects and protection from fire
as possible explanations for the spatial distribution of
species on mounds.
Up to date, only very few studies ( et al.
2013, et al. 2014) have specifically addressed
the fact that climatic gradients play a role in the de-
gree of difference between the vegetation of termite
mounds and the surrounding savanna. Also, landscape-
scale gradients, e.g. based on underlying pedology,
hydrology or microclimatic effects have rarely been
addressed. Exceptions includes studies on mound com-
position and structure ( 1998a,b;
et al. 2009a,b), vegetation composition in relation
to soil changes mediated by termites or land use/graz-
ing history ( et al. 2016; et al. 2016;
et al 2016). It is of interest that some of
these studies deliver evidence that the notion of termite
mounds as being richer in nutrients than the matrix,
thus being more densely vegetated by woody plants
and more attractive to large herbivores does not hold
true in all habitat contexts ( et al. 2013;
2013; et al. 2016; et al
2016).
4.2 Creation of vegetation mosaics
Termite mounds have long been suspected to be responsi-
ble for various patterns including savanna landscapes dot-
ted with thickets, striped with with woody vegetation, or
dotted with bare patches, and in some cases research has
shown clear causal relationships. soil heterogeneity intro-
duced by termites in feedback with woody plant growth has
been shown to induce the formation of islands in the Oka-
vango delta ( et al. 1998; et al.
1998, 2012), leading to a mosaic grassland dotted with veg-
etated islands. Similar grass-dominated landscapes dotted
with thickets were described as "termite savannas" (
1936), also implying causal relationships. Several studies
since have confirmed that termite mounds cause or heav-
ily contribute to the formation of thicket clumps in savan-
nas (e.g., 2008; 1997; et al. 2009).
In regions of high aridity, savannas often show patterns of
dots, bands, or stripes of vegetation, separated by bare areas.
These patterns are known collectively as "tigerbush" (
1956; 1950), and different factors
including termite mounds have been hypothesised to be re-
sponsible for their creation. Initially, pattern-generation was
attributed to low and spatially highly stochastic precipita-
tion and local runoff processes from bare areas, which can,
e.g., be initiated by termite mounds ( 1956;
1950; 1970, 1971). However, spatially
explicit modelling studies have shown that vegetation pat-
terns can be self-organising as a result of competition and
43
Erpenbach & Wittig
facilitation of plants ( 2001; -
1997; 1999; et al.
1995). While these mechanistic modelling approaches did
not consider spatial heterogeneity explicitly in modelling
banded, striped or dotted patterns, they did not rule out the
possibility that termite mounds might have facilitative or in-
hibitive influences and could be involved in pattern genesis
at the landscape scale.
4.3 Desertification
The generation of vegetation patterns by termite mounds
has been connected to desertification processes, but recent
evidence contradicts this idea. Dotted vegetation patterns
have been associated by several authors (e.g., et
al. 2004; et al. 2009) with catastrophic shifts in
vegetation cover that lead to desertification of arid environ-
ments. This view has been opposed by a recent modelling
study ( et al. 2015), which demonstrated that
ecosystem resilience to drought would be higher and recov-
ery after drought would be faster under local-scale influ-
ence of termite mounds. By including only spatial changes
in water availability and water use efficiency introduced by
regularly spaced mounds in their model, et al.
(2015) were able to reproduce vegetation patterns at sev-
eral spatial scales, both on and off mounds, which matched
patterns observed in situ. These results corroborate the hy-
pothesis of et al. (2000) that termite mounds, as
ecological buffers, can prevent catastrophic shifts towards
either encroachment or desertification, and hence help sa-
vanna systems to persist. Termite mounds may therefore
play an important role in avoiding desertification in arid
areas. The role of termite mounds is particularly important
because many regions worldwide are threatened by changes
in the amount and temporal distribution of precipitation due
to global climate change. However, since the timeframes in
regard both to the persistence of termite mounds themselves
as well as of colonisation, establishment and senescence of
termite mound vegetation are unknown, further research
throughout Africa in these regards would be needed when
addressing the potential role of termite mounds in the savan-
na ecosystem under global change.
4.4 Landscape restoration
Besides being important for ecosystem resilience, termites
might even be used in landscape restoration after severe
droughts. The ecosystem engineering effects of termites on
environmental heterogeneity and vegetation have already
successfully been used to restore degraded and laterite-
encrusted sites in West Africa. Landscape restoration has
been a subject of considerable interest in the Sahel region.
While some regional-scale vegetation models predict a fu-
ture ’greening’ of the Sahel based on precipitation chang-
es, mainly involving herbaceous vegetation ( et al.
2011, 2013), the same authors emphasise that human im-
pact, in particular deforestation, might trigger landscape
degradation and desertification in West Africa. Results
from a series of studies in the Sahel suggest that rehabilita-
tion of degraded landscapes can be facilitated by mulching
and actively promoting termite activity ( 1997a,b;
1999; et al. 1999;
1997; 1999). The tech-
niques used in these rehabilitation studies are adapted from
traditional farming, like the traditional Zaï system, and have
been shown to be highly effective ( et al. 2008;
2011, 2014, et al. 2016). In line
with findings of improved resilience to drought (-
et al. 2015), the cited rehabilitation studies have shown
that termites indeed play the role of an ecological buffer in
a dynamic savanna landscape. Furthermore, long-term man-
agement experience in promoting termite activity in a Zaï
system in Burkina Faso has shown that termites enable even
the formation of a closed-canopy forest on formerly highly
degraded land under the current precipitation regime (-
2014, et al. 2017).
In summary, evidence from multiple sub-disciplines of ecol-
ogy indicates that termite mounds are important elements of
the savanna system across scales, ranging from local scale
facilitation/inhibition processes to landscape-scale patch dy-
namics, pattern generation and ecological buffering mecha-
nisms, to regional - or even continental - scale dynamics
of tree–grass coexistence, forest–savanna boundaries, and
desertification. Active promotion of termite activity has
been successfully used for ecosystem restoration, and it has
been shown that termite mounds can enhance ecosystem re-
silience in the savanna system. However, little progress has
been made so far to integrate the effects of termite mounds
across scales. In our own projects, shortly presented in the
following section, which described and analysed commu-
nity and diversity patterns, and their drivers from local to
regional scales, we tried to deliver a starting point to close
this gap.
5 Vegetation of termite mounds across
different scales and along a climatic
gradient – the West African example
With respect to the effects of termite mound on various eco-
logical parameters, in particular on soil characteristics, it is
suggestive to think of "termite mound vegetation" as similar
across large areas of a continent, as similar observations re-
garding soil modification by termites and the prevalence of
woody plant thickets on mounds (as compared to the sur-
rounding savanna matrix) have been made in many different
locations. However, most of the studies mentioned above
have examined termite mound vegetation only at a local
scale or even only at a single site, leaving larger-scale pat-
terns largely unknown. In particular, effects of regional and
landscape-scale biotic and abiotic environmental variability
on termite mound–savanna dynamics have been mostly ig-
nored. Only recently, progress has been made to generalise
the effects of termite mounds, at least at the landscape scale,
using modern remote sensing and modelling techniques
( et al. 2010a; et al. 2015; et al.
2016). In addition, former studies from Africa have mainly
focussed on East African and South African savanna sys-
tems. However, our own recent projects ( et al.
2013, 2014, 2016) tried to fill the West African gap. Two
of these projects (see 5.1 and 5.2) quantified differences in
plant species diversity and community composition between
mounds and the surrounding savanna matrix at regional and
landscape scales. The third one (5.3) additionally followed
44
Flora et Vegetatio Sudano-Sambesica 19
individual seedlings and saplings to address the processes
driving the patterns quantified in the first two projects by in-
vestigating the influence of termite mounds on colonisation,
growth and mortality of plants.
5.1 The influence of termite-induced heterogeneity on
savanna vegetation along a climatic gradient in West
Africa
5.1.1 Gradient analysis of communities
In a regional study, ranging from the Sahel to the southern
limit of the Northern Sudanian vegetation zone, -
et al. (2013) showed that termite mound vegetation
differs strongly from the surrounding savanna along the
entire length of a strong bioclimatic gradient. Interestingly,
mound and savanna vegetation pronouncedly changed along
the gradient, and with a similar magnitude of change. Thus,
mounds do not represent azonal elements in the savanna,
but are influenced by climate as strongly as the surround-
ing savannas. At a regional scale, mean annual precipitation
was the most influential factor for differences in vegetation
along the transect, both for savannas and for mounds. Dif-
ferences in plant community composition and plant diver-
sity between mound and savanna vegetation were related to
differences in soil composition, particularly in base satura-
tion and cation exchange capacity.
5.1.2 Diversity analysis
The analysed measures of diversity were often, but not con-
sistently higher on mounds than in the surrounding savan-
nas, whether all vegetation layers were considered together
or whether woody plants and herbaceous plants were con-
sidered separately. Diversity analysis showed a complicated
pattern, very likely reflecting variation in local environmen-
tal conditions, disturbance regime and protection status of
the research sites. The community contrast between mound
and savanna quantified using DCA was more pronounced
with increasing precipitation. Species accumulation curves
showed that species richness generally increased with in-
creasing precipitation. Both a comparison of species accu-
mulation curves and response ratios of diversity measures
indicated that the contribution of mounds to local phytodi-
versity also increased with increasing precipitation. This ef-
fect was stronger for woody plants. et al. (2013)
conclude that communities of mounds and savannas are
well separated throughout the whole study area and climate
and soil help explain the magnitude of this separation.
5.1.3 Indicator species analysis
Several species were identified as characteristic mound spe-
cies, as they frequently and abundantly occur on termite
mounds. Along the entire transect, all in all eleven species
were identified as characteristic for termite mounds. Ac-
cording to their position within the transect, the particular
plots revealed an increase in the number of characteris-
tic species with increasing precipitation. et al.
(2013) argue that this increase might be related to an overall
larger species pool in the more humid areas. Notably, sev-
eral functional traits reoccurred among characteristic mound
species: all were woody or at least subligneous, severals
howed scrambling or climbing growth forms, and all, except
Wissadula amplissima (L.) R.E.Fr. and Combretum aculea-
tum Vent.,featured diaspores with adaptations to zoochory.
Four species of Malvaceae (Wissadula amplissimia and
three species of Grewia) as well as three Capparaceae spe-
cies (Capparis sepiaria L., Cadaba farinosa Forssk., and
Maerua oblongifolia (Forssk.) A. Rich.) were among the
characteristic species. According to the authors cited above,
this might suggest further shared functional traits or pread-
aptations leading to selection against these species in the
surrounding savanna or a competitive advantage of these
characteristic species in the mound microhabitat.
5.2 The contribution of Macrotermes mounds to lands-
cape-scale variation in vegetation in Pendjari National
Park (North Benin)
5.2.1 Soil gradient analysis of communities
In a landscape-scale study in Pendjari National Park (North
Benin), strong systematic variation of savanna vegeta-
tion was observed ( et al. 2016). In a DCA of
termite mound plots, a stronger grouping according to the
surrounding savanna types was found than expected. Cor-
relation analysis of DCA results indicated that different soil
parameters were related to vegetation differences between
savanna plots of different savanna types than to vegeta-
tion differences between mound plots of different savanna
types. Further correlation analysis of CCA results for each
savanna type with soil parameters showed that only some
soil parameters (pH, conductivity, available potassium)
were correlated with the difference between mound com-
munity and savanna community across all savanna types;
correlations with further soil parameters differed between
savanna types. While mound communities were always dis-
tinct from the surrounding savanna, they varied to different
extents depending on the surrounding savanna type. In par-
ticular, both savanna vegetation and mound vegetation were
more variable in areas with moderately deep alluvial soils
than in areas with more extreme soil conditions. Results
from both correlative analyses combined showed that differ-
ences in mound vegetation between savanna types and the
contrast between mound and savanna vegetation in each sa-
vanna type were influenced by variability in soil parameters.
Additional variability in plant communities of mounds and
savannas, which differed in extent between savanna types,
suggests the importance of further factors, such as local en-
vironmental filtering by flooding and available local-scale
species pools, for mound and savanna community composi-
tion.
5.2.2 Analysis of community distances in response to soil
Clay enrichment and related changes in potential cation ex-
change capacity, pH, and concentrations of magnesium and
calcium explained 22% of the variation in just one latent
variable. Concentration of potassium cations was identified
to be an important factor for three further latent variables,
and additional inclusion of potassium concentration, plant
available phosphorus, and sand content increased the varia-
tion explained to 62%. These results were interpreted as
45
Erpenbach & Wittig
confirming the hypothesis that clay enrichment is the most
important driving factor for the development of termite
mound vegetation, but emphasise the importance of interac-
tions between clay content and plant macronutrients.
5.2.3 Indicator species analysis
et al. (2016) found ten characteristic species for
the termite mound microhabitat. Each of the four sampled
savanna types had its own characteristic mound species
set. Both Detarium and Crossopteryx savannas had several
indicator species, whereas only one and two characteris-
tic mound species, respectively, were detected for mounds
in the temporally waterlogged Mitragyna savannas and in
Terminalia savannas. Characteristic species included six
species in the herbaceous layer (i.e., plants below 50 cm
height), five of which were geophytes capable of subterra-
nean dormancy. These results were interpreted as support
for a hypothesised environmental filtering by protection
from fire ( 2008; et al. 2008a) and
by differences in water availability between mounds and the
surrounding savanna.
5.3 Dynamics of juvenile woody plant communities on
termite mounds in a WestAfrican savanna landscape
5.3.1 Analysis of diversity patterns
In a local-scale monitoring study conducted over the course
of two years ( et al 2014), mound communi-
ties were more species rich and more diverse than savanna
communities, while savanna communities showed higher
species evenness. While the most frequent species occurred
both on mounds and in savannas, species that were restricted
to one plot type occurred particularly infrequently. Although
mound and savanna plots shared nearly half of their spe-
cies pools, their communities were distinct. In a multivari-
ate analysis of community distances, mound and savanna
communities showed a clear separation for both non-woody
and juvenile woody plants. Mounds showed a slightly lower
non-woody plants. Partly in accordance with expectations,
the results showed that higher overall diversity on mounds
towards more uniform juvenile woody plant vegetation on
mounds than in the surrounding savanna.
5.3.2 Analysis of population dynamics
Basic population metrics of juvenile woody plant commu-
nity structure (basal area, number of individuals and num-
ber of stems) showed significantly higher values on ter-
mite mounds than in savannas at each sampling time, and
increased over time with one minor exception. Individual
turnover between samplings was high, with only about one
third of 859 sampled individuals being present at all sam-
pling occasions. No significant differences could be found
between mounds and savannas for metrics of juvenile
woody plant community dynamics (individual mortality,
colonisation and turnover). Further analysis of causal driv-
ers of population dynamics and resulting community pat-
terns was impeded by high turnover and die-back rates of
woody juvenile plants, and overall low frequency and abun-
dances of species. Only four woody species were present at
more than four sampled plots at all sampling times.
5.3.3 Analysis of soil composition
In the study of et al. (2014), Termite mound
topsoil was significantly different from the surrounding sa-
vanna topsoil. On average, mound soil contained more than
twice the amount of clay and plant available potassium than
the surrounding savanna, and potassium cation concentra-
tion was more than four times higher. Base concentration,
base saturation, concentrations of magnesium and calcium
cations, and pH were also significantly higher on mounds,
and the C/N ratio was significantly lower. These results
were in accordance with expectations and support the hy-
pothesis that mounds are favourable microhabitats for plant
colonisation.
5.3.4 Analysis of plant dispersal modes
A comparison of dispersal modes showed a significantly
larger number of zoochorous species within the community
of juvenile woody plants on mounds than in the commu-
nity of juvenile woody plants in the surrounding savanna.
Neither anemochorous species nor species with mixed dis-
persal mode showed differences in frequency of occurrence
between mound sand savannas. One can conclude that dis-
persal of zoochorous species is directed towards mounds,
which is congruent with the hypothesis that mounds are
keystone structures for both animals and plants in the sa-
vanna landscape.
In summary, our studies showed that mound communi-
ties are more uniform than savanna vegetation on the local
scale, but variation of mound vegetation in relation to the
surrounding habitat is pronounced at the landscape scale.
Furthermore, community composition was related to cli-
matic parameters at the regional scale, where turnover oc-
cured in the same order of magnitude as for the surrounding
savannas. Juvenile plant communities on mounds differed
in several parameters from juvenile communities in the sur-
rounding savannas, however, individual turnover between
seasons is too pronounced to allow final conclusions about
colonisation, recruiting and survival processes responsible
for the pronounced contrast between termite mound and
savanna vegetation. From our findings, and the body of
literature discussed above, it can be concluded that termite
mounds are an integral part of the savanna ecosystem across
scales, and that they are not only of considerable importance
for particular characteristic species, but also for maintain-
ing patterns and processes of the savanna over longer time
frames.
We suggest that future progress in understanding the role of
termite mounds in the savanna ecosystem will depend on
an integration of data across scales, while recognising the
variability of termite mound vegetation. Currently, much of
the actual patterns and processes might still be obscured by
unclear definitions of terms, uncertainties in regard to eco-
logical context, and methodological problems regarding
comparability of data and suitable analyses. Recent discus-
46
Flora et Vegetatio Sudano-Sambesica 19
sions have highlighted several problems to address, e.g. the
nutrient status of the matrix soil ( et al. 2013;
2013), or even the definition of “matrix” in-
cluding the distance to which mound influence occurs, and
may be detected also in regard to statistical tools (
2012, 2016). In addition, age and size of
termite mounds as well as the identity of their builders are
sometimes not well defined in studies ( 2016).
It is of considerable importance to address such issues be-
fore integrating data and aiming at a synthesis. However, it
is already clear that termite mounds are and important fea-
ture of savanna landscapes throughout Africa, and beyond.
The role of termite mounds for savanna vegetation at the
landscape scale will likely affect local human population
the most, and land use change will certainly affect processes
and patterns in regard to termite mounds locally and at the
landscape scale. However, the effects of climate change are
likely to have an impact on larger-scale patterns, where the
role of termite mounds is unclear. Since population dynam-
ics and processes at the local scale are neither fully known
nor understood for both termites and termite mound vegeta-
tion, further local studies with well-defined protocols re-
garding species, sizes and possibly ages of mounds could
help to inform larger-scale modelling and synthetic ap-
proaches.
References
(2005): Fungus-growing termites
originated in African rain forest. Current Biol 15: 851-855.
-
(2002): The
evolution of fungus-growing termites and their mutualistic
fungal symbionts. Proc National Acad Sci 99: 14887-14892.
-
(2012): Soil particle accumulation in termite (Ma-
crotermes bellicosus) mounds and the implications for soil
particle dynamics. Ecol Res 27: 219-227
(2009a): Physi-
cochemical and morphological properties of termite (Ma-
crotermes bellicosus) mounds and surrounding pedons on a
toposequence of an inland valley in the southern Guinea sa-
vanna zone of Nigeria. Soil Sci Plant Nutrition 55: 514-522.
(2009b): Soil-particle
selection by the mound-building termite Macrotermes bel-
licosus on a sandy loam soil catena in a Nigerian tropical
savanna. J Trop Ecol 25: 449-452.
(2000, eds.): Termites:
Evolution, sociality, symbioses, ecology. Kluwer, Dor-
drecht, The Netherlands.
(2007): The impact of mound-building
termites on surface soil properties in a secondary forest of
Central Amazonia. Appl Soil Ecol 37: 267-276.
(1943): Termites and the fertility of soils. Trop
Agric (Trinidad) 20(6): 107-112.
-
(2015): Seed dispersal by
ungulates as an ecological filter: a trait-based meta-analysis.
Oikos 124: 1109–1120.
(2002): Initiation à l’agroforesterie en zone
sahelienne: Les arbres des champs du plateau central au
Burkina Faso. IRD Éditions/Karthala, Paris, France.
(2007): Biological control of
Striga hermonthica by Cubitermes termite mound powder
amendment in sorghum culture. Appl Soil Ecol 37: 175-183.
(2002): Rongeurs fouisseurs et régénération naturelle dans
une forêt classée du Burkina Faso. Bois forêts tropiques
271: 104-106.
(2013): Order Blattodea. Zo-
otaxa 3703: 46-48.
(1913): Notes biologiques sur quelques fourmis
et termites du Congo Belge. Rev Zool Africaine 2: 396-431
(2011, eds.): Biology of
termites: A modern synthesis. Springer, Heidelberg, Germa-
ny.
(2008): Thicket clumps: A characteristic feature
of the Kagera savanna landscape, East Africa. J Veg Sci 19:
31-44.
(2015):
Termite mounds can increase the robustness of dryland eco-
systems to climatic change. Sci 347: 651-655.
(2015): Why is the influence of
soil macrofauna on soil structure only considered by soil
ecologists? Soil and Tillage Research 146: 118–124.
(1982): Some aspects of the action of termites on
soil composed of clays. Clay Minerals 17: 453-462.
(2010):
Termites, vertebrate herbivores, and the fruiting success of
Acacia drepanolobium. Ecol 91: 399-407.
(1956): Étude sur photographies aéri-
ennes d’une formation vegetale sahelienne: labrousse tigrée.
Bull Institut Française Afrique Noire. Sér A Sci Naturelles
18: 677-684.
(1963): Vegetation and geomorphology in Nort-
hern Rhodesia: An aspect of the distribution of the savanna
of Central Africa. Geogr J 129: 290-305.
(1979): The nests of Macrotermes bellicosus
(Smeathman) from Mokwa, Nigeria. Insectes Sociaux 26:
240-246.
(1975): Remarques
sur l’écomophologie de la flore termetophile du Haut-Shaba
(Zaïre). Contribution à l’étude de l’écosysteme forêt claire
(Miombo), Note 20. Bull Soc Royale Bot 108: 167-181.
(2001): Periodic spotted patterns
in semi-arid vegetation explained by a propagation-inhibi-
tion model. J Ecol 89: 616-628.
(1988): Nutrient
cycling by mound-building termites in low-fertility soils of
semi-arid tropical Australia. Australian J Soil Res 26: 375-
390.
(1998):
The mound-building termite Macrotermes michaelseni as an
ecosystem engineer. J Trop Ecol 14: 507-520.
47
Erpenbach & Wittig
(1885): The Formation of Vegetable Mould,
Through the Action of Worms: With Observations on Their
Habits. John Murray, London, United Kingdom
(2014): Spatial variability and abi-
otic determinants of termite mounds throughout a savanna
catchment. Ecogr 37: 852-862.
(2016): Termite
mounds alter the spatial distribution of African savanna tree
species. J Biogeogr 43: 301-313.
(1999): Large trees, fer-
tile islands, and birds in arid savanna. J Arid Environm 4:
61-78.
(2001) The effect of a soil-feeding termite, Cubi-
termes fungifaber (Isoptera: Termitidae) on soil properties:
termites may be an important source of soil microhabitat he-
terogeneity in tropical forests. Pedobiol 45: 1-11.
(1886): On the termite as the tropical analo-
gue of the earth-worm. Proc Royal Soc Edinburgh 13: 137-
146.
(2005): Petite histoire des con-
naissances acquises sur les termites et leur role agroécolo-
gique. Étude Gestion Sols 12: 153-163.
-
(2006): Fluore-
scent pseudomonads occuring in Macrotermes subhyalinus
mound structures decrease Cd toxicity and improve its ac-
cumulation in sorghum plants. Sci Total Environment 370:
391–400.
(1947): Observations on the ecology of the
Budongo rain forest, Uganda. J Ecol 34: 20-87.
(2011): An introduction to termites. In -
NT (eds): Biology of termites:
A modern synthesis, 1-26. Springer.
(1993): Plant-
distribution in the islands of the Okavango delta, Botswana:
determinants and feedback interactions. African J Ecol 31:
118-134.
(2015a): The origin and im-
plications of variations in soil-related properties within
Macrotermes falciger mounds. Geoderma 249/250: 40-50.
(2015b):
The age of large termite mounds – radiocarbon dating of
Macrotermes falciger mounds of the miombo woodland of
Katanga, DR Congo. Palaeogeogr Palaeoclimatol Palaeo-
ecol 435: 265–271.
-
(2013): The influence of termite-in-
duced heterogeneity on savanna vegetation along a climatic
gradient in West Africa. J Trop Ecol 29: 11-23.
(2014): Dynamics of
juvenile woody plant communities on termite mounds in a
West African savanna landscape. Flora Vegetatio Sudano-
Sambesica 17: 28–41.
(2016): The contribution of Macrotermes mounds
to landscape-scale variation in vegetation in a West African
National Park. J Veg Sci (early view) doi:10.1111/jvs.12463
(1968): The vegetation of Zambian termitaria.
Kirkia 6: 169-179.
(1908): Ruwenzori. An account of the expedition
of H.R.H. Prince Luigi Amadeo of Savoy, Duke of the Ab-
ruzzi. Dutton, New York, USA.
(2003): Miombo woodland
termite mounds: resource islands for small vertebrates? J
Zool 259: 161-168.
(1915): Termite economy. South African J Sci 12:
60–64.
(2010):
Multi-scaled habitat considerations for conserving urban bi-
odiversity: native reptiles and small mammals in Brisbane,
Australia. Landscape Ecol 25: 1013-1028.
(2012): Ef-
fects of erosion from mounds of different termite genera on
distinct functional grassland types in an African savannah.
Ecosystems 15: 128-139.
(2006): The importance of nu-
trient hot-spots in the conservation and management of
large wild mammalian herbivores in semi-arid savannas.
Biol Conservation 130: 426-437.
(2010): Termite mediated heterogeneity of
soil and vegetation patterns in a semi-arid savanna ecosys-
tem in Namibia. PhD thesis, Univ Würzburg, Würzburg,
Germany.
(1984): La végétation de la Haute-Volta. PhD
thesis, Université de Bordeaux III, Bordeaux, France.
(2011):
Uncovering cryptic species diversity of a termite communi-
ty in a West African savanna. Molecular Phylogenetics Evol
6: 964-969.
(2500 BP): Thalia. In The History § 102 - § 105.
(1955): A chemical and physical study of the soils
of termite mounds in East-Africa. J Ecol 43: 449-461.
K (2011): Modelling biome shifts and tree cover change for
2050 in West Africa. J Biogeogr 38: 2248-2258.
K (2013): The projected impact of climate and landuse
change on plant diversity: An example from West Africa. J
Arid Environm 96: 48-54.
(2004): Termite mounds
as nutrient-rich food patches for elephants. Biotropica 36:
231-239.
(1997): Fruchtmerkmale, endozoochore Sa-
menausbreitung und ihre Bedeutung für die Zusammenset-
zung der Pflanzengemeinschaft: Untersuchungen im Wald-
Savannenmosaik des Comoé Nationalparks, Elfenbeinküste.
PhD thesis, Univ Würzburg, Würzburg, Germany.
(1988): The ecology of African floodplain fo-
rests in semi-arid and arid zones: A review. J Biogeogr 15:
127-140.
(1965): The flora and fauna of
Lolui Island, Lake Victoria - a study of vegetationmen and
monkeys. J Ecol 53: 573-597.
(2000): Ecological buf-
fering mechanisms in savannas: a unifying theory of long-
term tree-grass coexistence. Plant Ecol 150: 161-171.
48
Flora et Vegetatio Sudano-Sambesica 19
(1990): Termites, soil fertility and carbon cycling
in dry tropical Africa - a hypothesis. J Trop Ecol 6: 291-305.
(1997): Positive and
negative effects of organisms as physical ecosystem engi-
neers. Ecol, 78: 1946–1957.
(2011): Large termitaria act as
refugia for tall trees, deadwood and cavity-using birds in a
miombo woodland. Landscape Ecol 26: 439-448.
(2013): Escaping the flames: large termi-
taria as refugia from fire in miombo woodland. Landscape
Ecol 28: 1505-1516.
(2002a): Termite soil
preferences and particle selections: strategies related to eco-
logical requirements. Insectes Sociaux 49: 1-7.
(2002b): Ef-
fect of termites on clay minerals in tropical soils: fungus-
growing termites as weathering agents. European J Soil Sci
53: 521-527.
(2004) The soil structu-
ral stability of termite nests: role of clays in Macrotermes
bellicosus (Isoptera, Macrotermitinae) mound soils. Europe-
an J Soil Biol 40: 23-29.
(2006): Soil invertebrates as ecosystem engineers: intended
and accidental effects on soil and feedback loops. Appl Soil
Ecol 32: 153-164.
S (2007): Role of the fungus-growing termite Pseudacan-
thotermes spiniger (Isoptera, Macrotermitinae) in the dyna-
mic of clay and soil organic matter content. An experimental
analysis. Geoderma 139: 127-133.
DE (2011): Influence of termites on ecosystem functioning.
Ecosystem services provided by termites. European J Soil
Biol 47: 215-222.
(2016): Termites: The neglected
soil engineers of tropical soils. Soil Sci 181: 157-165.
(2014): Termites and ants in Burkina Faso (West
Africa): taxonomic and functional diversity along land-use
gradients; ecosystem services of termites in the traditional
Zaï system. PhD thesis, Univ Würzburg, Würzburg, Ger-
many.
(2017):
Ecosystem services of termites (Blattoidea: Termitoidae) in
the traditional soil restoration and cropping system Zaï in
northern Burkina Faso (West Africa). Agriculture Ecosys-
tems Environment 236:198-211.
-
(2012): Biodiversity islands in the savanna - analysis
of the phytodiversity on termite mounds in Northern Benin.
Flora Veg Sudano-Sambesica 15: 3-14.
(1999): In-
fluence of large termitaria on soil characteristics, soil water
regime, and tree leaf shedding pattern in a West African sa-
vanna. Plant Soil 206: 47-60.
KE (2011): Environmental and biological determinants of
Termitomyces species seasonal fructification in central and
southern Côte d’Ivoire. Insectes Sociaux 58: 371-382.
(2013): So-
cio-economical aspects of the exploitation of Termitomyces
fruit bodies in central and southern Côte d’Ivoire: raising
awareness for their sustainable use. J Appl Biosci 70: 5580-
5590.
(1779): Naturgeschichte der sogenannten weißen
Ameise. Beschäftigungen Berlinische Ges naturforschender
Freunde 4: 1-28.
(2003): Thermoregulation and ventilation of termite
mounds. Naturwiss 90: 212-219.
(2011): Termite mound architecture, from function
to construction. In Bignell DE, Roisin Y & Lo NT (eds.):
Biology of termites: A modern synthesis, 349-374. Springer,
Heidelberg.
(1998a): Experimental heating
of Macrotermes bellicosus (Isoptera, Macrotermitinae)
mounds: what role does microclimate play in influencing
mound architecture? Insectes Sociaux 45: 335-342.
(1998b): The effects of tempe-
rature on the architecture and distribution of Macrotermes
bellicosus (Isoptera, Macrotermitinae) mounds in different
habitats of a West African Guinea savanna. Insectes Sociaux
45: 51-65.
(1999): Reproductive success of
Macrotermes bellicosus (Isoptera, Macrotermitinae) in two
neighbouring habitats. Oecol 118: 183-191.
(2016): No evidence for an elephant-termite feedback loop
in Sand Forest, South Africa. Biol Conserv 203: 125-133.
-
DE (2010a): Regional insight into savanna hydrogeomor-
phology from termite mounds. Nature Comm 1(6).
(2010b): The spatial extent of termite influences on herbivo-
re browsing in an African savanna. Biol Conserv 143: 2462-
2467.
(1979): Seed dispersal by baboons in the Shai Hills, Ghana.
Ecol 60: 65-75.
(2007): Calcium
carbonate in termite galleries - biomineralization or upward
transport? Biogeochem 82: 241-250.
(2011): Termite phylogenetics and
co-cladogenesis with symbionts. In
(eds.): Biology of termites: A modern synthesis 2,
7-50. Springer, Dordrecht, The Netherlands
(1990): The role of
termites and ants in soil modification - a review. Australian
J Soil Res 28: 55-93.
(2004): Termitaria as browsing
hotspots for African megaherbivores in miombo woodland.
J Trop Ecol 20: 337-343.
(1950): Vegetation patterns in the semi-
desert plains of British Somaliland. Geogr J, 116(4/6): 199-
211.
(1975-1976): De l’origine de la flore terme-
tophile du Haute-Shaba (Zaïre). In (ed): Comptes
rendus de la VIIIe réunion de l’AETFAT Genève 16-21 sep-
tembre 1974. Boissiera, Conservatoire bot 24b.
49
Erpenbach & Wittig
(1997a): Effect of termites and mulch on the phy-
sical rehabilitation of structurally crusted soils in the Sahel.
Land Degradation Development 8: 269-278.
(1997b): The impact of termites and mulch on the
water balance of crusted Sahelian soil. Soil Technol 1: 121-
138.
(1999): Contribution of termites
to the breakdown of straw under Sahelian conditions. Biol
Fertility Soils 29: 332-334.
(1997): Termite-induced change in
soil structure after mulching degraded (crusted) soil in the
Sahel. Appl Soil Ecol 6: 241-249.
(1999): The biological and
physical role of mulch in the rehabilitation of crusted soil in
the Sahel. Soil Use Management 15: 123-127.
(1999): Ter-
mite- and mulch-mediated rehabilitation of vegetation on
crusted soil in West Africa. Restoration Ecol 7: 33-41.
(1998):
The role of biota in the initiation and growth of islands on
the floodplain of the Okavango alluvial fan, Botswana.
Earth Surface Processes Landforms 23: 291-316.
(1993): Vegetati-
on-induced, subsurface precipitation of carbonateas an ag-
gradational process in the permanent swamps of the Oka-
vango (delta) fan, Botswana. Chemi Geol 107: 111-131.
(2012): Island forming processes in the
Okavango delta, Botswana. Geomorphol 179: 249-257.
(2013): The
impact of subterranean termite activity on water infiltration
and topsoil properties in Burkina Faso. Ecohydrol 6: 324-
331.
(1947): A soil reconnaissance journey through
parts of Tanganyika Territory December 1935 to February
1936. J Ecol 35: 192-265.
-
(2009): Fungus culturing, nutrient mining and
geophagy: a geochemical investigation of Macrotermes and
Trinervitermes mounds in southern Africa. J Zool 278: 24-
35.
(2005): Termitaria are
focal feeding sites for large ungulates in Lake Mburo Natio-
nal Park, Uganda. J Zool 267: 97-102.
(2009): Mound build-
ing termites contribute to savanna vegetation heterogeneity.
Plant Ecol 202: 31-40.
(1996): Root fora-
ging strategies and soil patchiness in a humid savanna. Plant
and Soil, 182:171–176.
(1948): Tro-
pical soil-vegetation catenas and mosaics -a study in the
south-western part of the Anglo-Egyptian Sudan. J Ecol 36:
1-84.
(2010): Termite bioturbation effects on elec-
tro-chemical properties of Ferralsols in the Upper Katanga
(D.R. Congo). Geoderma 158: 233-241.
E (2011): The origin of carbonates in termite mounds of the
Lubumbashi area, D.R. Congo. Geoderma 165: 95-105.
-
(2013): Clay compo-
sition and properties in termite mounds of the Lubumbashi
area, D.R. Congo. Geoderma 192: 304-315
(2014): Spatial
patterns and morphology of termite (Macrotermes falciger)
mounds in the Upper Katanga, D.R. Congo. Catena 114: 97-
106.
(1954): La végétation de Kaniama (en-
tre Lubishi-Lubilash, Congo Belge). Publ Institut national
l’étude agronomique Congo belge, Sér sci 61, 499 pp.
(2013): Direct and indirect effects of
termites on savanna tree-seedling growth. Plant Ecol 214:
811-819.
(2013): Termite
mounds may not be foraging hotspots for megaherbivores in
a nutrient-rich matrix. J Tropl Ecol 29: 551-558.
(1955): Some soil forming processes in the humid
tropics. J Soil Sci 6: 73-83.
(2013): Termite mounds as browsing hot-
spots: an exception to the rule. J Veg Sci 24: 211-213.
(1973): Root distri-
bution under a thicket clump on Accra Plains, Ghana - its
relevance to clump localization and water relations. J Ecol
61: 439-454.
(2012a): Large herbivores maintain
termite-caused differences in herbaceous species diversity
patterns. Ecol 93: 2095-2103.
(2012b): Termite activity, not grazing,
is the main determinant of spatial variation in savanna her-
baceous vegetation. J Ecol 100: 232-241.
(2013): Termites, large
herbivores, and herbaceous plant dominance structure small
mammal communities in savannahs. Ecosystems 16: 1002-
1012.
(2009): Termite species
variations and their importance for termitaria biogeochemi-
stry: towards a robust media approach for mineral explora-
tion. Geochem Exploration Environment Analysis 9: 257-
266.
(1984): Defense mechanisms of termites.
Annual Rev Entomol 29: 201-232.
(1979): Termite hills in Africa: their characteri-
stics and evolution. Catena 6: 267-291.
J (2004): Self-organized patchiness and catastrophic shifts
in ecosystems. Sci 305(5692): 1926-1929.
(1999): Suppres-
sion of plant growth on the mounds of the termite Copto-
termes lacteus Froggatt (Isoptera, Rhinotermitidae). Insectes
Sociaux 46: 366-371.
(2002): Phylogenetic relationships in Termitomyces (fami-
ly Agaricaceae) based on the nucleotide sequence of ITS:
a first approach to elucidate the evolutionary history of the
symbiosis between fungus-growing termites and their fungi.
Molecular Phylogenetics Evol 22: 423-429.
50
Flora et Vegetatio Sudano-Sambesica 19
(1994): Utilization of termitarium
soils by éléphants and its ecological implications. African J
Ecol 32: 222-232.
(1983): Termitaria-nu-
trient patchiness in nutrient-deficient rain forests. Biotropica
15: 1-7.
(2011): Using soil and water conservation
techniques to rehabilitate degraded lands in northwestern
Burkina Faso. Internat J Agric Sustainability 9: 120-128.
(2008):
Restoring soil potentialities using Zaï and compost in Yaten-
ga (Burkina Faso). Biotechnol Agronomie Société Environ-
nement 12: 279-290.
-
(2009): Early-warning signals for critical tran-
sitions. Nature 461(7260): 53-59.
(1963): Aperçu sur les groupements végétaux du
Katanga. Bull Soc Royale Bot Belgique 96: 233-447.
(2008):
Impact of termite activity on soil environment: a perspective
from their soluble chemical components. International J En-
vironm Sci Technol 5: 431-444.
(2014): Do the large termite
mounds of Macrotermes concentrate micronutrients in ad-
dition to macronutrients in nutrient-poor African savannas?
Soil Biol Biochem 68: 95-105.
(2016): Woody species
composition in an African savanna: determined by centuries
of termite activity but modulated by 50 years of ungulate
herbivory. J Veg Sci 27: 824–833.
(2012): Application of distance-
decay models for inferences about termite mound-induced
patterns in dryland ecosystems. J Arid Environm 77: 138-
148.
(2016): Termite mounds alter the spatial dis-
tribution of African savanna tree species: artefacts and real
patterns. J Biogeogr (early view) doi:10.1111/jbi.12865
(2009): Integrating ethno-eco-
logical and scientific knowledge of termites for sustainable
termite management and human welfare in Africa. Ecol So-
ciety 14: Art. 48.
(2010): Termite-induced heterogeneity in African savanna
vegetation: mechanisms and patterns. J Veg Sci 21: 923–
937.
(1781): Some account of the termites, which
are found in Africa and other hot climates. In a letter from
Mr. Henry Smeathman, of Clement’s Inn, to Sir Joseph
Banks, Bart. P. R. S. Philosoph Transact Royal Soc London
71: 139-192
-
(2012): Explosive backpacks in
old termite workers. Sci 337: 436.
(2014): Environmental
heterogeneity as a universal driver of species richness across
taxa, biomes and spatial scales. Ecol Letters 17: 866-880.
(2013): Termites
facilitate and ungulates limit savanna tree regeneration. Oe-
col 172: 1085-1093.
(2004): Animal species
diversity driven by habitat heterogeneity/diversity: the im-
portance of keystone structures. J Biogeogr 31: 79-92.
(1995): A model si-
mulating the genesis of banded vegetation patterns in Niger.
J Ecol 83: 497-507.
(1943): The vegetation of the Karamoja district,
Uganda: an illustration of biological factors in tropical eco-
logy. J Ecol 31: 149-177.
(2012): Quantifying the
masses of Macrotermes subhyalinus mounds and evaluating
their use as a soil amendment. Agric, Ecosystems Environ-
ment 157: 54-59.
(2008): Effects of controlled live-
stock grazing and annual prescribed fire on epigeal termite
mounds in a savannah woodland in Burkina Faso. Insectes
Sociaux 55: 183–189.
(2008a): Im-
pact of Macrotermes termitaria as a source of heterogeneity
on tree diversity and structure in a Sudanian savannah un-
der controlled grazing and annual prescribed fire (Burkina
Faso). Forest Ecol Management 255: 2337-2346.
(2008b) Macrotermes mounds as sites for
tree regeneration in a Sudanian woodland (Burkina Faso).
Plant Ecol 198: 285-295.
(2015): Long-term effects of Macrotermes
termites, herbivores and annual early fire on woody under-
growth community in Sudanian woodland, Burkina Faso.
Flora 211: 40-50.
(1936): Termitensavannen. In
W (eds): Festschrift zur Vollendung des sechzigsten Lebens-
jahres Norbert Krebs, 275-312. J. Engelhorn, Stuttgart, Ger-
many.
(1939): Luftbildplan und ökologische Bodenfor-
schung: ihr zweckmäßiger Einsatz für die wissenschaftliche
Erforschung und praktische Erschließung wenig bekannter
Länder. Z Ges Erdkunde 139: 241-298.
(2006): Termites as mediators of the wa-
ter economy of arid savanna ecosystems. In
(eds): Dryland Ecohydrology, 303-313.
Springer, Dordrecht, The Netherlands
(2013): Functional traits of trees on and off termite
mounds: understanding the origin of biotically-driven hete-
rogeneity in savannas. J Veg Sci 24: 227-238.
(1962): Soil below a termite mound. J Soil Sci
13: 46-51.
(1967): A termite mound in an iron age burial
ground in Rhodesia. J Ecol 55: 663-669.
(1972): Distribution of gold in termite mounds
and soils at a gold anomaly in Kalahari sand. Soil Sci 113:
317-321.
(1974): Calcium-carbonate in termite mounds.
Nature 247: 74-74.
51
Erpenbach & Wittig
(1975): Composition of termite (Macrotermes
spp.) mounds on soil derived from basic rock in 3 rainfall
zones of Rhodesia. Geoderma 14: 147-158.
(1977): Use of mounds of termite Macrotermes
falciger (Gerstacker) as a soil amendment. J Soil Sci 28:
664-672.
(1983): The vegetation of Africa: A descriptive
memoir to accompany the UNESCO/AETFAT/UNSO vege-
tation map of Africa. Natural Ressources Research 20. UN-
ESCO, Paris, France.
(1970): Brousse-tigrée patterns in Southern
Niger. J Ecol 58: 549-553.
(1971): Vegetation stripes on sheet wash sur-
faces. J Ecol 59: 615-622.
(1952): The vegetation of southern Rhodesian ter-
mitaria. Rhodesia Agric J 49: 288-292.
(2003): Clay mineralogical and related characte-
ristics of geophagic materials. J chem Ecol 29: 1525-1547.
(1988): Termites and the soil environment. Biol
Fertility Soils 6: 228-236.
(1978): The role of termites in eco-
systems. In Brian M (ed): Production ecology of ants and
termites. Internat Biol Program Vol 13, 245-292. Cambridge
University Press, Cambridge, UK.
(2013): Variation in savanna vegetation on
termite mounds in north-eastern Namibia. J Trop Ecol 29:
559-562.
(1997): Editorial: Geophagy: a vestige of pa-
laeonutrition? Trop Med Internat Health 2: 609-611.
