FEATURE
creased grazing intensity – Decreasers
– as well as Increaser species that prolif-
erated under chronic heavy grazing are
shown in Figure 1.
We were also able to quantify the sen-
sitivity of forbs to grazing based on the
relative positions of their ‘centres’ of
distribution along the grazing gradi-
ents. These grazing sensitivity weights
(GSWs) were scaled from 0 to 10 to re-
flect an increasing negative response to
increased grazing intensity. Forbs with a
GSW of more than five were Decreas-
ers, whereas those with a low GSW were
Increasers. Forbs with low GSWs were
generally prostrate or low-growing
whereas the grazing-sensitive forbs with
a high GSW tended to be taller, upright
species with growing points vulnerable
to defoliation.
The Decreaser/Increaser response cate-
gorisation has been commonly applied
to grasses but not to forbs. Although
all or most grasses in a grassland have
been thus classified the response of
many has not been determined and not
all grasses are consistently influenced
by grazing.
Our research showed that most (88-
92%) forbs are similarly unaffected by
grazing or too rare to precisely deter-
mine their grazing response pattern.
Therefore, we reduced the list of around
370-400 forb species in each grassland
to a small subset of the most responsive
forbs: 24 and 32 species in Mistbelt and
Sandstone grassland, respectively.
The pattern of distribution of just these
indicator forbs closely matches the pat-
tern of all forbs along the grazing gradi-
ent. In both grasslands there were more
Decreaser than Increaser forbs, with
the latter including alien species such
as Richardia brasiliensis (Figure 1) and
some native forbs (e.g. Spermacoce
natalensis, Aristea abyssinica; Figure
1). The potential key grazing indicator
forbs, plus a few somewhat less sensi-
tive alternatives, were common and lo-
cally abundant enough to be potential
key species for assessing and monitor-
ing and assessing the ecological state
and condition of mesic grassland. The
two grasslands shared 13 indicator spe-
cies but each had unique indicators,
emphasising the need to conduct graz-
ing impact studies in different grassland
types.
Like the weighted grass species meth-
od developed as a quick means of mon-
itoring important changes in the grass
species composition of Highland Sour-
veld (Hurt and Hardy 1989), we propose
that key indicator forbs could be used
to more efficiently and effectively survey
the diverse forb communities that are
an integral part of mesic grassland.
A site score – which we call the forb con-
dition score (FCS) - is simply the sum of
the relative abundances of key indica-
tor species found at a site weighted
(multiplied) by their respective GSWs.
The weighed sum is divided by 10 so
the maximum FCS would tend to 100
if most of the key species present are
Decreasers with high GSWs. A low FCS
would indicate that heavy to severe
grazing had reduced the abundance of
or eradicated grazing-sensitive species
and promoted more grazing-resistant
forbs (Figure 2).
Figure 2: The weighted indicator graz-
ing forb method applied to a fence-line
contrast in mesic grassland (FCS = forb
condition score).
When a survey of the abundance of all
forbs present at a site is undertaken, or
such data are already available, in Mist-
belt or Sandstone grassland, the cal-
culated FCS would contribute towards
understanding the ecological condition
at the site by indexing its state of over-
grazing. A more rapid survey just look-
ing for key indicator forbs would also
provide a condition index for assess-
ment and monitoring.
The FCS method could be used togeth-
er with other important indicators such
as basal cover, ratings of soil erosion,
age and state of the grass sward, as well
as standard veld condition assessment
(VCA) methods to provide a deeper un-
derstanding of the integrity and health
of mesic grassland.
Our previous research (Scott-Shaw and
Morris 2015) showed that the VC score
alone is not a good predictor of forb
species richness (which is laborious to
determine). In contrast, the FCS is sig-
nificantly positively correlated with total
indigenous forb species richness (Mist-
belt: r = 0.793; Sandstone: r = 0.885) and
can be used to adequately predict the
number of indigenous forbs likely to be
found at a site. Thus the few key species
present at a site can tell a lot about the
state and diversity of the whole grass-
land forb community.
We envisage that agricultural, botani-
cal and conservation professionals and
citizen scientist could include the FCS
method in their current toolbox for rou-
tinely assessing and monitoring grass-
land sites. The method does need field
testing to determine which key forb in-
dicators could be confused with other
species (which may or may not have the
same grazing response) and the list of
grazing indicators could be trimmed or
augmented to ensure a practicable and
precise rapid grassland appraisal meth-
od. Increased knowledge of forb dy-
namics would guide the management
of grassland for multiple objectives and
help sustain that fabulous flower show.
References
1. Chamane S, Kirkman KP, Morris C, O’Connor T. 2017. Does high-density stocking affect perennial forbs in mesic grass-
land? African Journal of Range & Forage Science 34: 133-142.
2. Hurt CR, Hardy MB. 1989. A weighted key species method for monitoring changes in species composition of Highland
Sourveld. Journal of the Grassland Society of Southern Africa 6: 109-113.
3. Morris CD, Scott-Shaw R. 2019. Potential grazing indicator forbs for two mesic grasslands in South Africa. Ecological
Indicators https://doi.org/10.1016/j.ecolind.2019.105611
4. Scott-Shaw R, Morris CD. 2015. Grazing depletes forb species diversity in the mesic grasslands of KwaZulu-Natal, South
Africa. African Journal of Range & Forage Science 32: 21-31.
5. Siebert F, Dreber N. 2019. Forb ecology research in dry African savannas: knowledge, gaps, and future perspectives.
Ecology and Evolution https://doi.org/10.1002/ece3.5307
Grassroots
Vol 19
No 4
November 2019
08