India: Restoration of Banni Grassland in Gujarat

Location of Banni Grassland; Credit: Birds of Banni Grassland
Banni Grassland; Credit: Birds of Banni Grassland
Banni Grassland; Credit: Birds of Banni Grassland


The Gujarat Ecology Commission (GEC), a body created by the Forest and Environment Department of the Government of Gujarat, initiated a restoration programme for Banni grassland in the state’s northern Kachchh district in 1995. Targeting lands degraded by overgrazing, woody species invasion and salinization, the project aims to improve ecosystem productivity, and thereby ensure the socio-economic resilience of local inhabitants in a region entirely dependent upon livestock and the rangelands that sustain them. Implemented by the Gujarat Institute of Desert Ecology (GUIDE) and the Vivekananda Research and Training Institute (VRTI) on a 15-hectare site at Bhirandiyara and a 200-hectare site at Dhordo, the project has employed a combination of methods, including: protection from grazing by a trench and/or barbed wire fencing; removal of weeds (e.g. Prosopis juliflora) to reduce competition; soil amendment via shallow ploughing, slope making etc.; farmyard manure application to improve the soil nutrient composition; and reseeding to improve the seed bank. With the help of a wet monsoon in 1997, the project has yielded promising results to date and has afforded practitioners much insight into the processes and factors that must be addressed in order to ensure an expeditious and sustainable rehabilitation of these grasslands.

Quick Facts

Project Location:
23.7337326, 69.8597406

Geographic Region:

Country or Territory:


Grasslands & Savannas - Temperate

Area being restored:
215 hectares

Project Lead:
The Gujarat Ecology Commission (GEC)

Organization Type:
Governmental Body

Project Partners:
Forest and Environment Department of the Government of Gujarat, the Gujarat Institute of Desert Ecology (GUIDE), and the Vivekananda Research and Training Institute (VRTI)


Project Stage:

Start Date:

End Date:

Primary Causes of Degradation

Agriculture & Livestock, Climate Change, Invasive Species (native or non-native pests, pathogens or plants), Urbanization, Transportation & Industry

Degradation Description

Even though grassland species of Banni have developed a variety of drought adaptations, a severe drought, perhaps, is the most important natural factor responsible for its degradation. It is also important to note that, severe consecutive droughts, which have occurred only twice between the period 1901 and 1980 (1904-1905 and 1968-1969), have increased abruptly during recent years (i.e. between 1981-1996). Three consecutive severe droughts, the first of its kind in this century, occurred from 1985 to 1987, and another two consecutive droughts occurred in 1995 and 1996. This situation has predominantly affected the soil moisture and thereby hastened the process of grassland degradation.

Although the Banni has always exhibited high salinity levels, the rivers that once flowed from the Kachchh mainland to Banni not only deposited the detritus and maintained soil moisture, but they also leached the salinity of the area during good rainfall years. The construction of 6 medium dams–namely, Rudramata, Niruna, Nara, Kaila, Kaswati and Gajansar–has nearly stopped the collection of water from a 1603 km2 catchment area, except during the heavy monsoon years. Thus, natural leaching of the salinity of different parts of Banni has been totally interrupted. In addition to this, seawater from the Kori creek entered the northern part of Banni due to the construction of Punjabi-road during the year 1965. Furthermore, the sparseness or absence of vegetative cover in high saline areas encourages the wind to transport salt particles from the Rann to fertile Banni areas, resulting in further increases in salinity. The cumulative effect of all these factors is the increased salinity in nearly 90 per cent of the Banni grassland.

In addition to these factors, the Gujarat State Forest Department, as a measure to check the advancement of the Rann, has planted initially about 31,550 ha exclusively of Prosopis juliflora. The successive droughts and increasing salinity of Banni provides more suitable conditions for the growth and extension of this species, which has today become the most dominant species and is spreading at the rate of about 25 km2 per year.

Grazing is another major problem in Banni grassland. Large herbivores are known to speed up the nutrient turnover rate in savannas (Ruess 1987, Ruess and McNaughton 1987). However, excessive or over grazing leads to massive degradation. During the normal rainfall years, livestock from neighbouring talukas and districts of Kachchh, and even from other states, totalling over 2 lakhs immigrates into Banni for grazing.

The grazing pressure in these arid rangelands ranges from 1 to 4 ACU (Adult Cattle Unit) per ha per year against the carrying capacity of 0.2 to 0.5 ACU per ha per year (Raheja 1966). In spite of the fact that one ACU in arid areas requires between 2 and 5 ha of pastureland, the available grazing land, which includes all categories–i.e. good, fair, medium and poor condition grasslands–in the Kachchh district, is less than 2 ha per ACU. Therefore, the grazing requirement of the entire livestock in the district cannot be met from the existing grassland resources.

The aforementioned issues have resulted in a severe unsustainability in Banni grassland, which decreased the human population from 14,389 in 1981 to 10,949 in 1991 (24 per cent), and livestock population from 49,240 in 1982 to 26,084 in 1992 (47 per cent). Among the three regions (east, west and central) of Banni, the eastern Banni has already lost its capacity to sustain both human and livestock populations, and the western Banni is slowly loosing its capacity. As a result, a maximum concentration of 55 per cent of the human (4149) and 65 per cent of the livestock population (7333) is located at central Banni region (GUIDE, 1998), exerting excessive pressures in this area which may lead to massive degradation due to over-exploitation of resources.

Reference Ecosystem Description

Banni supports the growth of perennial and palatable grasses of high productivity, which grow in low to moderate saline areas. They are Sporobolus pallidus, Sporobolus helvolus, Dichanthium annulatum, Cenchrus ciliaris, Cenchrus setigerus, Desmostachya bipinata etc. High saline areas are colonised by perennial grasses of low productivity and palatability, such as Aeluropus logopoides, Eurochondra sp. etc. Among the tree species, Acacia nilotica was once distributed all over Banni. The shrub and tree strata were mainly composed of Prosopis cineraria, Acacia nilotica, Acacia leucophloea, Acacia Senegal, Salvadora persica, Salvadora oleodes, Capparis decidua, Tamarix sp. and Prosopis juliflora. Today, palatable grass and tree species like Acacia and Salvadora sp., though present, exhibit a significant decrease in abundance due to the massive invasion and dominance of Prosopis juliflora.

The wild animals of Banni include: Blue bull, Chinkara, Black buck, Blacknaped hare, Wild boar, Jackal, Grey wolf, Caracal, Hyena, Fox, Jungle cat, etc. It also supports a rich diversity of avifauna, herpatofauna, and invertebrates. Further, the water bodies of Banni, during good rainfall years, form important staging grounds for thousands of migratory cranes, and also support over 150 species of other migratory and resident birds.

Project Goals

The primary objective set for the restoration programme is to return areas degraded by heavy grazing, high saline content, and infestation of Prosopis juliflora to former levels of productivity in order to improve ecological integrity and strengthen the local pastoral economy. Through this project, methods and techniques for the restoration of Banni grassland will be developed and evaluated, as will guidelines for the long-term sustainable management of this ecosystem.


The project does not have a monitoring plan.

Description of Project Activities:
As a pilot measure, the restoration of Banni grassland was initially intended to cover a total area of 500 ha at 5 different sites. However, due to administrative and local issues in acquiring the land, the restoration programme could be successfully launched at only two sites--one located 11 km before Bhirandiyara and the other at Dhordo village, covering an area of 15 and 200 ha respectively. The Banni grassland could be improved only in the good rainfall years. Since good years are generally followed by droughts in a cycle of every 3 years, or sometimes alternative years, the best method to improve grasslands is through protection from frequent grazing to facilitate the replacement of undesirable annual grasses with productive perennial species. Protection from grazing allows the perennials to produce large quantities of seeds. Thus depending on the condition of a site, a continuous protection from grazing develops the ground layer principally made up of perennial grasses (Kanodia and Patil, 1983). The same improvement can be achieved, albeit at a slower rate, with either a deferred or rotational grazing system where the grasses are protected up to seed dispersal. Therefore, the first step in any restoration strategy, of course, is to protect the disturbed habitats and communities from being further wasted, and from losing the extant genes (Singh and Jha, 1993). As a measure to protect the Bhirandiyara restoration site from livestock grazing, a trench fence consisting of 1.2 m width and 1 m deep was laid along the periphery of the site. There are views that trench fencing is not a foolproof method to protect the grass plot from cattle. However, its effectiveness as a barrier depends upon the width and depth of the trench. Therefore, a barbed wire fence around the site was also raised to ensure double protection from any kind of grazing. It has been reported that the protection of grasslands with the help of barbed wire fencing increases forage production of better grasses (Bhimaya et. al. 1967, Ahuja 1977, Kanodia et. al. 1978). At Dhordo restoration site, due to the coverage of a vast area of 200 ha, it was decided to have only trench fencing. However, to ensure full protection from cattle, a 2-m wide and 1.5-m deep trench was dug along the periphery of the plot, which covers a length of 6000 m. The mud removed from the trench was also used to raise a 1.5 m high mud wall on the inner fringe of the trench, thus ensuring complete protection from cattle. This method of fencing, though less expensive as compared to barbed wire fencing, requires maintenance every year. Apart from its cattle-proof function, the trench also helps in harvesting rainwater due to its vast length, depth and width. Prosopis juliflora infestation is believed to be the major cause in reducing the productivity of pastureland and hindering the growth of grasses. Unwanted thorny bushes need to be removed mechanically or manually to eliminate the competition. Even useful fodder bushes, such as Zizyphus nummularia, should not be allowed to grow more than 4 per cent in a grazing area, so as to maintain the forage production from the pasture (Ganguli et. al. 1964). Hence, Prosopis juliflora and other unwanted bushes growing inside the restoration site were totally uprooted with the help of the Gujarat Forest Development Corporation. Succession of vegetation in a natural sequence is a time consuming process and can be expected to succeed in the desired direction only if the nucleus of the plant material is available in a productive stage on a given site. Plantation of seed slightly below the surface after scratching it or in strips with appropriate species of grasses is the quickest and most ensured method of improvement of degraded grassland (Kanodia and Patil, 1983). For reseeding, first preference was given to indigenous grass varieties, which normally grow under arid climatic and soil conditions of Banni. Therefore, indigenous seeds were obtained from Banni Development Agency, Bhuj, as well as from the Indian Grassland and Fodder Research Institute, Jhansi. Indigenous species such as Cenchrus setigerus, Sporobolus sp., Eurochondra sp., Dichanthium annulatum, Chloris barbata and Cenchrus ciliaris, were selected for sowing in the restoration site. Species such as Chrysopogan fulvus and Pennisetum pedicellatum were also sown in small quantities, although the later species was not reported from Banni area. The performance of this species in the restoration site (i.e. growth and productivity) will aid in planning future large-scale introductions of such species over the entire Banni area. The reseeding activity, it is hoped, will increase the seed bank of the area, and thereby enrich the grass density and diversity of the restoration site. In arid areas, under rainfed conditions, the sowing of grass seeds is normally done in the months of June/July with the onset of the monsoon. To enhance the germination and uniform distribution of seeds, the seeds were mixed with soil and farmyard manure. Then they were thoroughly mixed by sprinkling water. This reduces the transport of seeds through wind, and the moisture in the seed mixture hastens the process of germination. Care was taken to ensure that the seed sown was covered by a 1- to 2-cm soil layer. Because the grass seeds are very small, the stored germplasm is insufficient to support the germination and growth of seedlings from deeper soil layers. The stocking rate of seeds varied with the species. The normal stocking rate of seed reported is: 2.5 kg/ha for Dichanthium annulatum; 5 kg/ha for Cenchrus ciliaris and Cenchrus setigerus; 5 kg/ha for Sporobolus sp. etc. to 75 kg/ha for Lasiurus species. However, keeping in mind the near absence of a seed bank in the restoration sites, a stocking rate nearly 3-4 times higher than the normal was used. The Bhirandiyara restoration site was divided into four sub-compartments, each covering an area of approximately 3.4 ha. The first compartment was kept under natural conditions (Compartment 1), where the fencing (trench and barbed wire) ensured complete protection from livestock grazing. Both pure and mixed combinations of grass seeds (e.g. Cenchrus sp. with Sporobolus sp.) were sown within the compartment. However, a strip (10m width and 250 m length) covering an area of 3750 m2 was kept completely under natural conditions, where no sowing of grass seed was undertaken. The basic view was that the area devoid of seed sowing resembled very much that outside the fenced area of the site; the only difference between these two was that the former received protection from grazing, while the later was subjected to grazing pressures. Hence, a comparison of the productivity of grass species of the said area with that outside the fenced site would provide data about the effect of grazing pressure that existed in the area. The mixed combinations of seeds were used to ascertain with species associations were more favorable in terms of productivity. At Dhordo restoration site, the whole area was divided into 30-m wide strips. Ploughing, broadcasting of grass seeds and addition of farmyard manure were carried out at alternative strips. In natural pastures of arid regions, the scope of fertilisation is limited mainly due to the uncertainty of rainfall. In Banni, the sale of cow dung has further limited the natural supply of manure, affecting soil nutrients and plant growth in the area. Under favorable circumstances, the two-way system whereby nutrients are consumed through grass feed and then returned to the soil through dung ensures that little or no loss of nutrients from the grassland area takes place. In order to rectify the nutrient deficit and improve soil composition at the restoration site, farmyard manure was added at the rate of 2 tonnes per ha. Apart from improving the nutrient status of the degraded land, this practice will improve such physical properties of the soil as permeability, soil structure, etc. Moreover, the chemical binding of dung with that of soil is known to reduce the soil salinity to some extent. Grass-legume mixtures are always desirable because of their complementary functions in providing nutritive and palatable forage for livestock. Legumes usually maintain their quality better than non-legumes, even at maturity, and being rich in protein, they enhance the forage value and also add substantially the needed nitrogen to the soil. Thus, grasses directly or indirectly benefit from the use of nitrogen components manufactured by the legumes. On death and decay of the legume plant, its roots and root nodules break down in the soil and release their stored nitrogen, which then becomes available to the associated crops. In order to take advantage of these qualities, legume species like Afylosia sp., Sesbania sp. and Stylo hamata were sown at the restoration sites. Sesbania, in particular, produces high quality fodder that can also be used as cattle feed. The restoration site selected at Bhirandiyara was located in a highly saline track of Banni. In this site, experimentation on salinity reduction and control was undertaken to improve the grass diversity and productivity. This site was divided into four equal compartments, of which, compartment 1 is left under natural condition where no soil modifications were carried out. Compartment 2 was ploughed to a depth of 50 cm to improve the soil permeability, water infiltration and to break up-surface crust. In this compartment, organic manure was also added at the rate of 2 tons/ ha. To leach the salinity of the research site, 5-m wide gentle slope with a height of 30 cm, and with drainage channels 50 cm deep and 50 cm wide, was dug in Compartment 3. Similar activities, with changes in the width of the slopes to 3 m and height of the slopes to 50 cm, were conducted in Compartment 4 to study the effects of increasing the slope height on the grass species diversity and productivity. Compartment 4 was further sub-divided such that half the slopes and the trench faced northwest and the other half faced east-west. The basic view of this amendment was to study the effects of wind-borne salt deposition on the diversity and productivity of grass species growing on the slopes facing towards and against wind direction. It is hoped this will help generate additional information on the deposition of salt particles through wind action and its impacts on the productivity of different grass species growing in the area. The Dhordo restoration plot encompasses an area of 200 ha (width 1km x length 2km), located near village Dhordo. The whole area of the site was divided into 30-m wide strips. Ploughing, adding farmyard manure and sowing grass seeds were carried out in alternating strips. Prosopis juliflora growing inside the fenced site was removed, with the exception of a few individuals, in which the side branches were removed in order to change the structure of the plant. There is a common view that the spreading cover of Prosopis juliflora reduces ground cover, and that a modified structure would assist in supporting understory vegetation. Furthermore, it is thought that the tree may also act as a windbreaker, thus providing a valuable service under certain circumstances.

Ecological Outcomes Achieved

Eliminate existing threats to the ecosystem:
GRASS COVER At Dhordo, there was a marked variation in grass cover at three study sites, i.e., restoration site (site A), outside unprotected area (site B) and Prosopis juliflora infested areas (site C). The restoration site was broadly classified into two categories: unploughed area where no soil working was carried out and ploughed area where soil working such as ploughing, addition of farmyard manure and reseeding of grass species was undertaken. The grass cover estimated in the ploughed and unploughed area of site A gradually increased from 56.7 and 54.7 per cent in September to 67.4 and 72.1 per cent respectively at the end of November, with a slight fall (57 and 65 per cent) in the first fortnight of November. Further, the unploughed and ploughed area did not record any significant variation in grass cover between the initiation and the end of the study period. Nevertheless, the grass cover was slightly higher in unploughed areas since October when compared to ploughed area. Contrary to this, unprotected and Prosopis juliflora infested areas showed very low availability of cover, which, during the study period, decreased from 15.4 to 10.7 per cent and 30.7 to 8.3 per cent respectively. However, site C showed a higher cover than site B during the initiation of the study. A comparison between the unploughed protected area (restoration site) and unprotected area (site B), which have similar climo-edaphic conditions, showed significant variation in grass cover. The grass cover of the former was 3.6 to 6.7 times higher than the latter during the entire period of study. This clearly demonstrated that grazing pressure in the unprotected area was significant. The cover increased by 10.7 and 17.4 per cent respectively in the ploughed and unploughed areasof site A, while it has decreased to 6.1 per cent at site B and 22.4 per cent at site C. At Bhirandiyara, the monthly mean cover in the restoration site was 49 per cent at the initiation of the study in October, and thereafter it showed a slight increase in November and remained almost stable up to the first fortnight of December, except that it showed a slight fall in the second fortnight of December. In the unprotected (outside) area, maximum cover was recorded in the first fortnight of November (69.4 per cent) and gradually decreased during the post-monsoon period, attaining a minimum of 19.4 per cent at the end of December. Thus, initially (i.e. first fortnight of November) the unprotected area showed a higher cover than the restoration site, but then decreased significantly by the end of the study, indicating clearly the effects of grazing pressure on the unprotected area. The grass cover recorded under various amendments of the restoration site showed similarities in certain areas and variation in others. The cover estimated was very high (76.9 per cent) in the unploughed area and drastically low (17.5 per cent) in the narrow horizontal slopes. Interestingly, looking at the various amendments of the restoration site, the cover showed a decrease during the study period in the unploughed, ploughed and broad slopes, whereas it showed an increase in the narrow vertical and horizontal slopes. The unploughed area in the restoration site had higher cover availability during the beginning and end of the study period, while the ploughed area exhibited a slightly higher cover during the middle period--i.e. first fortnight of November to first fortnight of December. Among the three types of slopes in the restoration site, almost similar cover availability was recorded at broad (50 per cent) and narrow vertical slopes (46.3 per cent), while on the narrow horizontal slopes it was low (17.5 per cent) during October. However, at the end of December, the narrow vertical (50.6 per cent) and horizontal slopes (45.6 per cent) had the maximum cover, which in turn decreased to 38.8 per cent on the broad slopes. Overall, narrow vertical slopes showed a higher cover than other two slopes except in October and first fortnight of December. SPECIES DIVERSITY In Dhordo, a total of 25 species belonging to 20 genera were recorded in the restoration site, which in turn decreased to 18 species (72 per cent) in the unprotected area (site B) and 17 species (68 per cent) in the Prosopis juliflora infested area (site C). The occurrence of species such as Cenchrus ciliaris, Cyperus haspan, Chrysopogan fulvus, Scirpus sp., Setaria pallide, Themeda triandra, Tragus sp. and Sporobolus pallidus was restricted to the restoration site; whereas Sporobolus helvolus, Aeluropus logopoides, Dichanthium annulatum, Tetrapogan tenellus, Tragus sp., Chloris barbata, Cyperus rotundas, Echinocloa sp., Cenchrus setigerus, Eragrostis sp., Dinebra rotroflexa, Gandhiro, Digera rnuiicata, Aristida funiculata and Cressa erotica are some of the species found in all the three sites (A, B and C) at Dhordo. At Bhirandiyara, a total of 12 plant species belonging to 11 genera were recorded, which in turn decreased to 6 species (50 per cent) in the unrestored area. Further, within the restoration site, the number of individuals of the species fluctuated under different soil amendments. The species occurrence was high in the unploughed area (91.7 per cent) and slightly lower on vertical slopes (75 per cent). In other areas, about 66.7 per cent of the species were recorded. Cyperus haspan, Cyperus rotundus, Scirpus sp., Sporobolus helvolus, Aeluropus logopoides and Cressa erotica are the common species, which were recorded in all the amendment areas of the restoration site, as well as the unprotected site. Dichanthium annulatum was seen only on vertical slopes, while the occurrence of Chrysopogan fulvus and Setaria sp. were restricted to the unploughed area of the restoration site. Sporobolus sp. forms the dominant species in both ploughed and unploughed areas of the Dhordo restoration site while at site B and C, Cyperus rotundus was dominant. The dominance index of seven species; Cyperus rotundus, Chloris barbata, Aeluropus logopoides, Gandhiro, Echinocloa sp., Cenchrus sp. and Tetrapogan tenellus was comparatively higher in the ploughed area than the unploughed area. At site B, Aeluropus logopoides forms the second dominant and was followed by Chloris barbata and Sporobolus sp. At site C, the second dominant species was Chloris barbata while Sporobolus sp. and Cenchrus sp. forms the less and least dominant' species respectively. In Bhirandiyara, Cyperus haspan forms the dominant species in all the amendments and unprotected area except in narrow vertical slopes where Aeluropus logopoides exerted domination. Sporobolus sp. was found to be second dominant in ploughed and broad slopes whereas it was third dominant in narrow vertical and horizontal slopes. Cenchrus sp. was least dominant in unploughed and narrow horizontal slopes but, totally absent in ploughed and unprotected area. BIOMASS - Dhordo At Dhordo, data on biomass was collected from three sites: site A, site B and site C. Among the five samplings carried out between September and November 1997, maximum biomass obtained at site A, B and C was 3546.9 kg/ha in the second fortnight of October (mid study period); 618.3 kg/ha in September (beginning of study); and 410 kg/ha during end of November (end of study) respectively. It was also noted that unploughed and ploughed areas showed variation in the productivity of grass species. Interestingly, the biomass recorded was higher in the unploughed areas than in ploughed areas, although the monthly increase in productivity recorded was higher in the ploughed area. Between September (1511.9 kg/ha) and November (2736.9 kg/ha), the biomass increased by 81 per cent in the ploughed area, while it increased by only 67 per cent (2177.8 to 3639.8 kg/ha) in the unploughed area. This increase was precipitated by the improved productivity of such species as Cenchrus sp. (221 to 692 per cent), Chloris barbata (96 to 149 per cent), Echinocloa sp. (26.5 to 442 per cent), and Dichanthium annulatum (23 to 60 per cent). Contrary to these observations, increase in Sporobolus productivity was higher in unploughed area (216 per cent) than in ploughed area (104 per cent). There was a considerable variation in the productivity of different grass species in the three sites of Dhordo. A total of 12 important and commonly available grass species (Cyperus rotundus, Chlohs barbata, Sporobolus sp., Setaria sp., Aeluropus logopoides, Eragrostis sp., Gandhiro, Echinocloa sp., Cenchrus sp., Dactyloctenium aegyptium, Tetrapogan tenellus and Dichanthium annulatum) were selected for analysis. The total biomass estimated for these 12 species at three sites varied from 2505 kg/ha at site A, 230.3 kg/ha at site B and 201.9 kg/ha at site C. The biomass recorded at site C was 12.3 per cent less than site B. The 12 species encompass 87.2 per cent of the biomass estimated at site A, 67.3 per cent at site B and 64.1 per cent at site C. The biomass of a few highly palatable species such as Sporobolus sp., Cenchrus sp. and Dichanthium annulatum encompasses 31.7, 11.1 and 10.3 per cent respectively at site A, whereas it decreased to 14.5, 0.0 and 1.3 per cent respectively at site B and 3.6, 0.004 and 0.3 per cent respectively at site C. Contrary to this, the biomass contribution of species like Cyperus rotundus, Aeluropus logopoides and Chloris barbata, which are comparatively less palatable at matured stage, was least at site A--7.6, 3.8 and 11.2 per cent respectively--while it was 27.9, 23.3 and 16.7 at site B and 21.6, 12.2 and 19.2 at site C. Between September and November, total biomass of these selected 12 species showed an increase of 75.6 per cent at site A, whereas it decreased to 44.7 per cent at site B and 40.1 per cent at site C. The comparison of biomass of grass species between September and November at site A shows that, except three species (Cyperus rotundus, Setaria sp. and Dactyloctenium aegyptium), all other species showed an increase. Among the species at site A, maximum increase was recorded for Sporobolus sp., which alone encompassed 22 per cent of the biomass in September to 36 per cent in November, followed by Cenchrus sp. (4.8 per cent in September to 12.9 per cent in November) and Chloris barbata (5.9 per cent in September to 9.8 per cent in November). These three species alone represented nearly 32.7 per cent biomass in September, which increased to 58.8 per cent in November. BIOMASS - Bhirandiyara The biomass in the restoration site gradually increased from October to December; 1115.9 kg/ha in October to 2228.7 kg/ha in December and thus, showed an increase of 99.7 per cent, whereas it decreased to 31.5 per cent in the unprotected area during the same period. Among the five sub-compartments (unploughed, ploughed, broad slope 0.3m height x 5m wide, narrow vertical slope 0.5m height x 3m wide, and narrow horizontal slope 0.5m height x 3m wide) of the restoration site, narrow slopes showed a maximum increase of biomass between October and December--279.6 per cent in vertical slopes and 171.9 per cent in horizontal slopes. This was followed by the unploughed area (112.8 per cent), broad slopes (42.9 per cent) and ploughed area (25.9 per cent). This indicates clearly that the recovery of saline area in narrow steep slopes was faster. The biomass estimated (i.e. all grass species) within the restoration site was higher in the unploughed area than in the ploughed area for all months (excepting the first fortnight of December). Among the slopes, the broad slopes with a width of 5m had more biomass production (except during the end of December) than narrow vertical slopes of 3m width. The biomass of the narrow horizontal slopes was found slightly higher than the narrow vertical slopes during the first fortnight of November and December, while in other months, it showed lower biomass than the vertical slopes. Cyperus haspan made a major contribution to the biomass of the area, representing 41.6 per cent of the productivity of restoration site and 82.6 per cent of the unprotected area. Grass species such as Sporobolus sp., Cenchrus sp., Aeluropus logopoides, Dichanthium annulatum etc., sown in the area were capable of growing even under low moisture availability, whereas C. haspan - a water-loving species - requires high moisture for its survival and growth. Therefore, it was presumed that there will be a change in the contribution of C. haspan in the various amendment areas following the monsoon--i.e. from October to December. To confirm the above, and to understand the recovery process of different sub-compartments of the restoration site, biomass was analysed with and without the contribution of C. haspan. During the month of October, C. haspan constituted more than 67.8 per cent of the biomass of the restoration site, which in the month of December decreased to 34.9 per cent. Further, this species was found almost equal in ploughed (49.8 per cent) and unploughed areas (46.7 per cent) of the restoration site. However, it decreased to 43.7 per cent in the broader slopes and 38.4 per cent in narrow horizontal slopes, and a very low contribution of 17.5 per cent was recorded in narrow vertical slopes. The contribution of C. haspan decreased from 75 per cent in October to 45.6 per cent in December in the unploughed area, while in the unprotected area, it decreased from 83.4 to 74.9 per cent. A comparison of data during the same period in other sub-compartments showed that the percentage contribution of C. haspan changed from 61.8 to 50.1 at ploughed area; 75.2 to 19.4 at broad slopes; 54.9 to 12.5 at narrow vertical slopes; and 56.6 to 34.7 at narrow horizontal slopes. An increase in other grass species (excluding C. haspan) was recorded maximum at narrow vertical slopes (635.9 per cent), followed by broad slopes (363.2 per cent), narrow horizontal slopes (309.3 per cent), unploughed (164.8 per cent) and ploughed areas (64.4). In contrast to this, the outside area showed a decrease of 23.5 per cent. Apart from C. haspan, the biomass of five other common grass species of the site (Sporobolus sp., Aeluropus logopoides, Echinocloa sp., Cenchrus sp., and Scirpus so.) was analysed. The biomass contribution of these species varied from 67.5 per cent of the mean biomass production on vertical slopes to 35.2 per cent on broad slopes. Among the five grass species, biomass of Aeluropus logopoides, a saline tolerant species, was found maximum in all sub-compartments except ploughed and broad-slope areas. Ploughing and the addition of farmyard manure favoured the growth of Sporobolus sp., where it contributed 41 per cent of the biomass amongst the five species. In the broad slopes, both Sporobolus sp. and Aeluropus logopoides were found in almost equal abundance. In the unprotected (outside) area, (excluding C. haspan), Aeluropus logopoides was found to be the dominant species (88.6 per cent of the biomass). Amongst the five sub-compartments, contribution of Sporobolus sp. was maximum in the narrow vertical slopes (29.1 per cent), followed by unploughed (22.7 per cent) and ploughed (24.4 per cent) areas. Species like Aeluropus logopoides and Echinocloa sp. were dominant in unploughed areas, while Cenchrus sp. dominated (32.9 per cent) the narrow vertical slopes. Surprisingly, this species was totally absent in the ploughed area. The species Scirpus sp. was more prevalent in the ploughed area (38.3 per cent) when compared to unploughed area (23.2 per cent). The herbaceous stratum of the Bhirandiyara restoration site, and the surrounding area, was dominated by Cressa cretica. This species contributed about 21.2 per cent of the biomass. Among the sub-compartments, the unploughed area had the lowest biomass contribution of Cressa cretica (4.4 per cent) and was followed by vertical slope (20.4). In the other sub-compartments, it contributed more than 25 per cent to the total biomass, with a maximum contribution of 34.7 per cent in horizontal slopes. In the unprotected area, it represented about 15.2 per cent of the biomass.

Factors limiting recovery of the ecosystem:
In the restoration sites, the productivity was found to be higher in the unploughed area than in the ploughed area. In the ploughed area, the natural seed bank, along with tillers and rhizomes of grass plants, were disturbed due to ploughing activities, while in the unploughed area, these were left undamaged. Therefore, in the unploughed area, fewer showers were required to activate the vegetative growth of grasses from tillers or rhizomes. In the ploughed area, grass has to develop from the germination of seeds of different species sown in that area. The grass seeds have certain adaptations to survive in the harsh conditions, and not all seeds of a grass plant will grow after a single shower. Even in a single grass plant, seeds produced in the same spike have different levels of dormancy (an adaptive mechanism) and might germinate after a single shower, or they might require multiple showers. For example, seeds of grass species such as Cenchrus sp. require a certain amount of moisture to wash the inhibitors present in the seed coat before germination. This could be one reason for the late recruitment/germination of certain grass species like Cenchrus sp. in the ploughed area. Further, after the germination, the plant would take at least one full season to establish, and this may be an important reason for obtaining more biomass from unploughed areas than ploughed areas at both the restoration sites. At Bhirandiyara site, waterlogging is a common feature during heavy rainfall years. This creates a favorable environment for the growth of Cyperus haspan and Cyperus rotundus (unwanted annuals). Such a favourable situation prevailed during September 1997. Within 36 hours, the area received about 458 mm rainfall, which resulted in waterlogging for more than a fortnight. This enabled the rhizome of C. haspan to grow enormously and vigorously in and outside the restoration site, which is less palatable due to its high fibrous nature. Further, the implementation of soil amendments carried out on about 75 per cent of the restoration site disturbed the rhizome of C. haspan. These brought out higher cover and biomass availability during the initial period in the unprotected area, which was slightly higher than the restoration site of Bhirandiyara, as well as unprotected and Prosopis juliflora area of Dhordo site.

Socio-Economic & Community Outcomes Achieved

Economic vitality and local livelihoods:
Livestock is the mainstay of the inhabitants of Banni, and constitutes the bulk of their assets. Out of 10,949 people in Banni (1991 census), 2443 people are entirely engaged in livestock rearing and other allied activities. This pastoral economy depends primarily on the extent and condition of available grasslands to support grazing, especially given the region's aridity. Besides just pasturelands, though, Banni grassland is also a source of other products of local economic importance, such as: leaves, seeds and tubers for human consumption, as well as medicinal plants, building materials, thatch, fencing, gums and other products important in the rural economy (Sale 1981 and NRC 1983). During December 1997, over 1,75000 kg of grass was harvested from the Dhordo restoration site, a quantity of fodder sufficient to meet the livestock needs of 6 surrounding villages for more than 1 year. Similarly, at the Bhirandiyara site, biomass increased from 1115.9 kg/ha in October 2007 to 2228.7 kg/ha by December, and thus showed an increase of 99.7 per cent. This improvement has favorable implications for local herders.

Key Lessons Learned

Although there are reports of failure in the artificial restoration of degraded grasslands (Bronner 1990 and Westoby et al. 1989), the GEC’s activities have been relatively successful, and recovery of the degraded and saline land could be clearly visualized from first-year results. These results reflect higher species diversity, grass cover and biomass in the two restoration sites than the unprotected and Prosopis juliflora infested areas. The net above-ground production estimated for nine USA grasslands by Sims and Singh (1971) ranged from 107 g/m2 (1070 kg/ha) (ungrazed short prairie) to 512 g/m2 (5120 kg/ha) (grazed tall grass prairie). The obtained biomass during the month of November/December in a high saline tract of Bhirandiyara was 2 times higher than the minimum biomass recorded by Sims and Singh, and Dhordo was about 3 times higher. The dense grass cover that developed in the restoration sites reduced the evaporation rate, and this, in turn, reduced the surface soil salinity–from 13 EC to less than 4 EC in the Bhirandiyara restoration site. These findings highlight the fact that the productivity of any degraded saline lands can be improved through proper management actions.

Long-Term Management

The activities currently being undertaken by the Gujarat Institute of Desert Ecology (GUIDE) and the Vivekananda Research and Training Institute (VRTI) promote the rehabilitation and natural recovery of the two Banni restoration sites, Bhirandiyara and Dhordo. However, because successional processes take considerable time, the measures employed thus far have enabled only moderate progress toward restoration, most of that being achieved after the 1997 monsoon. Under these circumstances, it will take at least two more years to develop well-founded management guidelines for this area.

Sources and Amounts of Funding

Funding for this programme is provided through the Forest and Environment Department of the Government of Gujarat.

Other Resources

Gujarat Institute of Desert Ecology

Gujarat Ecology Commission

Primary Contact

Organizational Contact