Effects of herbivores and fire on riparian and upland savanna ecosystems
Field Operations Manual for Herbivore & Fire Exclosures on the Sabie and Letaba Rivers in the Kruger National Park
Compiled by
Thomas O’Keefe and Glynn Alard
December 2002 1
Contents
Project Summary 3
Introduction 5
Project Background and History of Development 5
Guiding Research Questions 7 Vegetation Community and Disturbance Processes 7 Soil Ecology and Biogeochemistry 9 Additional areas of research interest 11
Exclosure Layout 12
Vegetation Sampling Protocol 13 Transects 13 Plots 14 Methods 16 Woody species 16 Herbaceous species 17 Alien plants 19
Instrumentation Protocol 19 Atmospheric Monitoring 19 Soil Monitoring 20 Ground water Monitoring 21
Fixed Point Photos 21 Fine scale 21 Coarse scale 21 2
Detailed Site Map Methodology 22 Aerial Photography 22
Management 23 Site Management 23 Patrolling & Maintenance 23 Data Management 24 Access 24 Rules and Safety 25 Contacts 27 References 27
Appendix
Kruger Park Map Kruger Park 2000 Flood photos Sabie and Letaba River photos Schematic of Exclosure Layout Fencing Specifications Letaba Site, species list Sabie Site, species list Aerial view of Sabie sites showing basic vegetation associations Field Equipment List Elephant and Fire Impact assessment Aerial photo of Sabie pre and post flood Aerial photo of Letaba pre and post flood Channel detail of Sabie pre and post flood Sabie River site map Letaba River site map Data Sheets 3
Project Summary
Recent (February 2000) floods on the eastern seaboard of Southern Africa inundated large areas of the Mozambique coastal plain. Rivers draining Kruger National Park (KNP) in South Africa experienced flows (c. 8,000 m3 s-1 for the Sabie River) with return intervals of approximately 1:50 years, causing widespread removal of riverine vegetation and extensive alterations to channel habitat. The floods caused widespread human suffering but also provided a unique research opportunity. For the first time in nearly a century we will be able to characterize the long-term development and recovery of riverine forests from a major disturbance.
Savanna landscapes in South Africa are important reservoirs of biodiversity that are affected by expanding human populations and shifting patterns of land use. Within these landscapes, the riverine forests are important boundaries between aquatic and terrestrial systems that define the biodiversity and heterogeneity of these ecosystems.
South African National Parks rates research along with ecosystem conservation as two of its core activities. Together with an international team of scientists we have constructed large exclosures to study the role of herbivory and fire in the alteration and determination of spatial and temporal heterogeneity patterns of vegetation. This research will directly benefit our ecosystem conservation planning.
In their basic form, exclosures are fenced areas designed to keep out animals. At each locality, a pair of exclosures will limit the accessibility of herbivores; one being a “total” (electrified) exclosure, i.e. to exclude all herbivores (plant-eaters) from a hare upwards in size, and a “semi-exclosure”, consisting only of two cables and electric wiring at 1.8 m and 2.2 m above ground. The purpose of the semi-exclosure is to exclude elephant, and, by virtue of their size, giraffe, but will allow access to all other herbivores. Each exclosure will be divided in half, with one half allowed to burn should fire pass through, and the other kept unburned. These exclosures will extend from within the river channel to the crest, in order to enclose the full sequence of terrain morphological 4
features and their associated soils, vegetation, and smaller fauna. This will allow for the study of the relationships of habitats along the topographic gradient, from the crest to within the river channel. Funding for construction of the exclosures was provided by a grant from the U.S. National Science Foundation and funding from the Kruger Park Marathon Club. Kruger Park staff will maintain the exclosures while continuing to promote involvement with international scientists to conduct research at the sites.
The placement of large herbivore exclosures in the Kruger National Park (KNP) is a fundamental research activity that provides solid evidence on which our whole management policy rests. A critical component of the current research initiative in the KNP concerns the effects of herbivores (especially elephant) and fire on the vegetation and biodiversity of the Park, with the ultimate objective of devising (or revising existing) scientifically based management policies. The rationale behind exclosures is that by excluding these two important modifiers of the ecosystem, their effects on the ecosystem can be determined over time. The proposed exclosures are thus designed to address these issues, as well as post-flood recovery of riparian vegetation which is thought to be heavily influenced by herbivory. As the nature of some ecological changes or processes are of a long-term nature, the exclosures will serve as long-term monitoring and research sites for at least 25 years. At that time, the objectives for the future use of the exclosures should be reviewed. These exclosures will thus provide extremely valuable long-term information on key ecological processes; information that would be nearly impossible to obtain.
The exclosures along the Sabie are located across the river from Nkuhlu picnic site, in the area between the Sabie River and the Salitjie road. The Letaba exclosures are located approximately midway between the low-water bridge over the Letaba, and Mingerhout dam, and thus far from public view in a wilderness area. Special permission was obtained to locate the exclosure here and it may be subject to special research conditions.
5
Introduction
Savanna landscapes in South Africa are important reservoirs of biodiversity, and are affected by expanding human populations and shifting patterns of land use. In order to understand and manage these landscapes, several concepts must be employed. Foremost is the application of a modern systems viewpoint (e.g., Likens 1992), which hypothesizes that spatial and temporal heterogeneity of savannas are key to their function (Rogers 1997). In these landscapes, large herbivores and fire are recognized to play an important role in determining these heterogeneous spatial and temporal patterns through their effects on vegetation and the physical landscape. Despite this general knowledge that herbivory and fire are important, careful long-term studies that rigorously document the changes in isolation and as interactions have yet to be conducted.
Through the construction of large herbivore exclosures overlapping with fire treatments our primary research objective is to determine how herbivory and fire alter spatial and temporal heterogeneity patterns. These exclosures will extend from within the river channel to the crest, and encompass the riparian corridor representing a full sequence of geomorphological features and their associated soils, vegetation, and smaller fauna. Riparian corridors are perhaps the most obvious expression of boundaries in savanna regions, and these boundaries defining the heterogeneity of these ecosystems are of particular ecological interest.
Project Background and History of Development
This cooperative program builds on the extensive knowledge base generated by the many decades of research conducted by the Kruger National Park Scientific Services, and the decade-long Kruger National Park River Research Program and more recent River Savanna Boundaries Program in collaboration with several South African and American universities, government departments, and research agencies. The Kruger Park has a long history of interest in the possibility of erection of formal research exclosures (rather than the “incidental exclosures” which have become available because of for instance, enclosures being built to breed up rare antelope). This interest relates 6
particularly relating to fire and elephant effects. Fortunately this historical interest, and the more recent interest of other groups in riparian and riparian-upland issues in the Kruger Park, has provided an opportunity for unified structures dealing with all these interests, to be put in place.
In February 2000 the Kruger National Park was presented with a unique opportunity to develop a long-term experiment designed to examine the effects of herbivory. The largest flood since 1925 for the Sabie (7,000 to 8,000 m3 s-1) and Letaba rivers removed most of the vegetation along these rivers, primary research sites in the park, resetting the system to bedrock and sand. The vegetation before the flood was a mature riparian forest including stately fig and other majestic riparian trees. The Sabie River catchment is 7,086 km2 with a river length of 230 km and a mean discharge of 633 million m3/a. The Letaba catchment is 13400 km2 with a length of 490 km and a mean discharge of 631 million m3/a. Construction of large exclosures subsequent to this major flood event will allow us to follow the successional development and pattern formation of vegetation along riparian zones. Riparian zones are recognized as “hotspots” of activity because they integrate terrestrial and aquatic systems (Naiman and Décamps 1997). We believe the research we are conducting (viewed across aquatic, riparian and upland zones) will generate a novel understanding of savannas as integrated yet heterogeneous ecological systems. The systems approach to ecosystem study requires ecologists to expose the connections and fluxes between the elements of heterogeneity, and the feedbacks between heterogeneity and ecosystem function (Risser 1995). Therefore, boundaries in the landscape that define this heterogeneity are of particular ecological interest, and riparian corridors are perhaps the most obvious expression of boundaries in savanna regions. The emerging view of system heterogeneity emphasizes that biological richness has many facets (Noss and Cooperrider 1994). The first facet highlights the kinds of ecological systems that are present in a region. The second facet indicates the relative abundance of each kind of entity present in the area. The third facet indicates that the 7
various ecological components are dynamic in time, and that they are functionally connected with one another. Finally, the three facets of kind, relative abundance, and function are expressed in all ecological realms. Therefore, genes, species and populations, communities and ecosystems, and landscapes all can be considered as components of system heterogeneity (Kolasa and Pickett 1991, Collins and Benning 1996). Riparian systems, and their connection to in-stream processes and to interchanges with the upland components of savanna landscapes, provide a powerful test of the functional significance of spatial and temporal ecosystem heterogeneity. The development of this understanding is the overarching mission of our scientific programme.
Guiding Research Objectives
Vegetation Community and Disturbance Processes Under our overall objective of examining the role that herbivory and fire play in the alteration and determination of spatial and temporal heterogeneity patterns, we have identified the following primary research questions:
How does herbivory change the vegetation pattern? When herbivory is imposed on the basic geomorphic template, greater spatial heterogeneity in vegetation may occur resulting in a more structurally and biologically complex habitat. Released from pressure by herbivores, we expect that large trees will display higher growth rates although it will likely take up to 25 years to see this effect.
How does fire affect the vegetation pattern? Although previous long-term studies have been conducted in the park to examine the role of fire in structuring vegetation patterns, we will add focus on the riparian zone and the interaction effects with herbivory.
8
How do herbivory and fire affect the regeneration of vegetation following a major flood event? Floods such as those experienced during February 2000 are a rare and major disturbance. We predict that recovery of vegetation will be more rapid in the absence of herbivores, but with less diversity and habitat complexity.
How do herbivory and fires affect seed dispersal, seed germination, and then seedling survival? Because herbivores are important transport vectors for seeds, we expect that in the absence of herbivores, seed dispersal and germination decline. In the absence of fire, we expect seed germination declines as disturbance by fire is recognized to be important for successful germination. Our plots will allow us to examine the interaction between this two factors affecting germination. Finally, we expect that once they are established, fire and herbivory will both decrease individual seedling survival rates.
What effect do animals have on the physical and biogeochemical features of the landscape? One important example of this are pans which are a ubiquitous feature on the landscape. We expect that animal activity is necessary for their continued maintenance. Without animals to keep them open, we predict vegetation will invade and they will close up. In the interim small semi-aquatic fauna may benefit from conditions were there is no major large animal disturbance. Animals likely have an effect on soil structure and erosion rates through physical disturbance of soils and we will be able to examine this.
Our basic methodology to address these research questions will be to characterize riparian and upland vegetation in the different herbivory and fire treatments by measuring basal area, stem density, and canopy dimensions for all woody and herbaceous species at our sites, and individual growth rates of key species of interest that include Sclerocarya birrea, Acacia nigrecens, Diospyros mespiliformis, Kigelia africana, Breonadia salicina, and Trichilia emetica. We will also use fixed-point photography from the ground and aerial photography to document overall changes through time. In addition we will conduct biogeochemistry analyses to characterize the sites themselves and to better understand the features of the physical template that affect the vegetation community. 9
Soil ecology and Biogeochemistry In conjunction with studies on the vegetation community we will also conduct studies on soil ecology and biogeochemistry. These will include long-term experiments of water and nutrient availability on the development of vegetation and soil. The observation of soil water and nutrient status and fluxes will be used to address a number of hypotheses.
Within treatments “The different vegetation types along each transect from crest to river are influenced by the retention capacity and flux of water and nutrients within the soil profiles prevalent at each vegetation type.”
This involves comparison of soil water and nutrient uptake within each transect. Hence soil water monitoring and sampling sites on each transect are located in the riparian zone (2-3 sites), in a sodic zone where these exist and in the upland combretum zone. Here, surface and subsurface water and nutrient fluxes will be studied.
Across treatments “The rate of recovery of vegetation with and without herbivory will depend on the availability of soil water and nutrients.”
Observation of soil water and nutrient fluxes in profiles of similar vegetation types under different treatments will enable factors affecting growth and survival of vegetation to be distinguished between herbivore effects and effects due to water and nutrient availability. Observations will include surface and subsurface water and nutrient balance determination at each site in each treatment.
Transect dynamics “Surface and subsurface fluxes and accumulation both vertically and laterally influence the availability of water and nutrients in specific topographic locations in the transect.”
10
Surface runoff and subsurface water and nutrients observations will be used to define the hydrological processes, which result in accumulations of water and nutrients at specific locations in the profile and thereby influence the spatial distribution of vegetation recovery.
Instrumentation that would otherwise be disturbed by large animals will include freestanding equipment to measure atmospheric parameters to provide rainfall input and estimates of potential evapotranspiration boundary conditions. Parameters to be measured at the meteorological stations will include break-point rainfall, air temperature, air humidity, wind speed and direction and solar radiation. Soil water monitoring will include volumetric water content by Time Domain Reflectometry (TDR) and soil water tension status by tensiometry. Techniques have been developed to house this instrumentation within and outside of exclosures so that comparisons of soil water status can be made. The TDR method used will allow for simultaneous measurement of volumetric water content as well as bulk soil electrical conductivity, giving a continuous indication of nutrient and salinity status. The nutrients and salinity will also be sampled periodically to determine the make up of soil water dissolved solids and N-P ratios. The samples will be extracted, either by suction cup lysimeter or by a modified wick method. The exclosures will, for the first time, allow experimental studies of nutrient limitation on plant diversity and productivity. Surface runoff quantity and quality will be observed in runoff plots located at the soil monitoring stations.
Fundamental questions remain as to which nutrients are limiting to riparian vegetation (N or P) and in what elemental ratios, how water availability affects growth and survivorship of key riparian species, and what is the course of ‘terrestrialisation’ once water is removed. Instrumentation that would otherwise be disturbed by large animals will include equipment to measure and sample the following parameters: atmospheric monitoring (air temperature and rainfall), soil moisture, soil temperature, lysimeter and surface water collectors to collect soil and runoff solution for nutrient analysis, and groundwater monitoring wells. The exclosures will, for the first time, allow experimental studies of nutrient limitation on plant diversity and productivity. 11
Additional areas of research interest In addition to these primary research questions, we have identified other areas of research interest particularly in the area of biogeochemistry and will encourage collaborative relationships to develop other projects of interest that are compatible with the primary research objectives set out for the exclosures. The objective is that these exclosures will be used primarily for projects of a long-term nature.
The KNP has a well-documented objectives hierarchy (available at http://parks- sa.co.za, go to Kruger park, Scientific services) which includes explicit research interests, and many of these are obviously catered for by these exclosures. However, more detailed fleshing out, or unpacking, of particular objectives still needs to be carried out to further focus such questions.
Opportunities exist to examine the role of exotics in savanna landscapes. While the most pervasive exotics will be removed in accordance with park management activities (see management section of this manual) we predict that many exotic annuals will be displaced over time as the native riparian forest community develops. With changes in habitat structure we also anticipate changes in fauna including birds, rodents, and insects. This can in turn have feedbacks on the vegetation community that could be studied through the use of long-term mini-exclosures for these groups located within the large exclosure.
Additionally, in the next few years KNP will be embarking on a large-scale experiment to actively move elephants into high and low-density areas within the Park. The Sabie River exclosure will be constructed in an area destined to have low elephant densities, whereas the Letaba site is destined to have high elephant densities. Therefore, secondarily the exclosures will be used in cooperation with the large-animal researchers to quantify the long-term effects of differential elephant browsing and fire on ecosystem processes.
12
Exclosure Layout
Cost and practical considerations prevented us from replicating exclosures in a manner consistent with traditional statistical techniques. Despite this, large-scale ecosystem experiments have made important contributions to ecology, and this approach was selected for the design of these exclosures. While the scale selected makes replication prohibitive, designing smaller exclosures would have introduced edge effects and prevented us from effectively addressing many of the ecosystem questions that are of primary interest.
The design selected addresses the shortcoming of non-replication by incorporating basic principles that others have successfully used to address ecosystem questions (Carpenter and Kitchell 1993). The manipulations are strong and sustained with exclosures being used to apply a strong treatment effect with regard to herbivory. Statistical techniques designed for large-scale ecosystem experiments have been developed and successfully implemented (Matson and Carpenter 1990) which will be used in conjunction with ecological interpretation of changes that occur.
Exclosures are located at two sites within the Kruger National Park; one along the Sabie River and one along the Letaba River. At each locality, a pair of exclosures were erected; one being a “total” (electrified) exclosure, i.e. to exclude all herbivores (plant- eaters) from a hare upwards in size, and a “semi-exclosure”, consisting only of two cables and electric wiring at 1.8 m and 2.2 m above ground. The purpose of the semi-exclosure is to exclude elephant, and, by virtue of their size, giraffe, but will allow access to all other herbivores.
The exclosures are approximately 50 ha (1.0 x 0.5 km) in size and separated by an unfenced distance of approximately 400 m. They extend along the catena, from within the river channel to the crest, in order to enclose the full sequence of terrain morphological features and their associated soils, vegetation, and smaller fauna. This will 13
allow for the study of the relationships of habitats along the topographic gradient, from the crest to within the river channel. Part of the latter will be included in "sacrificial" exclosures, so-called because they may sustain damage and are designed to be reconstructed when the river floods (bankfull events with a 5-10 year recurrence interval) Nevertheless, they will serve a very important function as it is the lower riparian zone that took the brunt of the impact by the recent floods and underwent the greatest change
In order to introduce fire as an experimental treatment in the areas, each exclosure will be divided approximately in half with a firebreak. All natural fires will be allowed to enter the exclosures from outside, but hopefully, will always be excluded from the no- burn treatments. In addition, the open area between the two exclosures will act as a no burn reference area, and the unfenced reference area at the outside ends of the exclosures will serve as a natural fire regime treatment A total of six treatments will thus be possible. See the diagram in the appendix for details.
The exclosures along the Sabie are located across the river from Nkuhlu picnic site, in the area between the Sabie River and the Salitjie road. The Letaba exclosures are located approximately midway between the low-water bridge over the Letaba, and Mingerhout dam, and thus far from public view. Permanent interpretive signs at Nkuhlu picnic site and along the Salitjie road provide an explanation and general background on the exclosures, for the benefit of park visitors.
Exclosure Vegetation Sampling Protocol
Transects We have established two permanent transects in each of the six treatments. Transects extend from the river to the crest of the catena, perpendicular to the river channel. Transects are as evenly spaced as possible within each area, and at least 100 m from fences and at least 30 m from firebreaks and areas where construction activity was 14
concentrated. The location and bearing of the transects was determined using aerial photographs and 1:50 000 maps.
Permanent markers (25 x 25 x 50 cm concrete benchmarks) are located at the ends of each transect. The transects have been laid out in two steps. In the riparian zone a digital theodolite with a compass was used to establish a straight line from the benchmark along the bearing of the transect. In the uplands, where tree density increases and a straight, line of sight for the theodolite would be obscured by tree canopies, the line was extended along the same using a differential GPS geo-referenced to a nearby base-station to ensure accuracy.
Along the transect, stakes (12 mm steel reinforcing bar) are placed every 10m in the riparian zone and every 20 m in the upland. These points correspond to the location of permanent vegetation plots located on the downstream side of the transect but will also be used to relocate the transect for other studies. In the riparian zone, these distances were measured using the theodolite and in the uplands where the transect was continued with the GPS, the distances were measured using a laser rangefinder.
Plots Permanent vegetation sampling plots have been established 5 m downstream from the transects (to reduce disturbance on the plots from other uses of the transect lines.) Plots are 10m x 20m, with the long side parallel to the river channel.
Within the riparian zone, the permanent vegetation plots are contiguous to the top of the macrochannel bank. As the upland limit of the riparian zone may be unclear following the flood, a minimum of 6 plots are sampled at the beginning of each transect. This allows for sampling from the edge of the active channel up to the top of the macrochannel bank (where the sodic areas start). Most of the riparian zone will be between the permanent and temporary fences (temporary fences are the ones that we expect may be damaged in flood events, but they will be repaired as soon as possible after such events). The plots in the riparian zone will be contiguous as this area has a high 15
biodiversity and it is a narrow zone. Also, it is in this zone that most vegetation succession will occur due to post-flood community development. Contiguous plots ensure that data of this vital zone is maximized and thus future understanding of the spatial and temporal heterogeneity of the riparian zone is improved.
Plots in the sodic sites (where present) and uplands will be the of same dimensions (10m x 20m) and distance from the transect (5m). These plots will be spaced 20m apart along the transect. Plot placement may be subject to minor adjustment so as to avoid areas of construction activity including roads used by the building crew. The sodic sites and uplands are more extensive and vegetation patterns are more uniform; these areas were not altered to the great extent that the riparian zone was during the floods. Therefore, the scale of temporal heterogeneity will not be as great as in the riparian zone. For these reasons, sampling need not be as intensive.
The river-upstream (UU) corner of each plot will be 5m perpendicular from the transect, and correspond to the 20m stakes along the transect. All 4 corners will be exactly located and permanently marked by hammering 12mm steel re-bar stakes into the ground. Each stake is 0.5 m long, and approximately 10cm will be left exposed. Each stake will be marked with grooves that will indicate their position so that if only corner can be found, it will be possible to re-establish the plot correctly. • 1= Upland-upstream (UU) • 2= River-upstream (RU) • 4= River-downstream (RD) • 3= Upland-downstream (UD) Further, the GPS co-ordinates of the UU corner of each plot will be found using a differential GPS. Finally, a reference photograph will be taken diagonally across the plot from the UU to the RD corner using a 28 mm lens and a monopod at a height of 2 meters.
In order to measure the degree of alluviation and illuviation (riparian zones only), the length of the RU stake extending from the ground will be measured to the nearest mm. 16
Methods
Woody vegetation sampling A brightly coloured 60 m rope will be extended around the stakes to clearly mark the edges of the plot and clarify which individuals fall within and outside the plot. All plants rooted within the plot are recorded, i.e. even if the entire canopy of a tree extends into a plot but it is rooted outside, it will not be recorded, and vice versa.
When it is not clear what constitutes an individual within a clump of multiple stems or when stems may result from vegetative reproduction, clumps which are 0.5m or further apart will be classified as different individuals.
Two canopy heights will be taken for each individual:
1. The height to the top of the tallest branch will be measured using a clinometer if the individual is too tall to measure with a stadia rod. Tall trees that cannot be measured from within the site will be measured from as far back as necessary to see the top of the tree. 2. The height of the lowest foliage (which could be a single twig with leaves) will be measured.
Two canopy diameter measurements perpendicular to each other will be taken for each individual across the longest and shortest perpendicular axes of the canopy.
The diameter of every stem will be measured just above the basal swelling, using callipers for stems <5 cm in diameter and a diameter tape for stems >5cm in diameter. Where an individual consists of more than 10 stems, the number of stems will be counted and recorded; 10 stems will then randomly be measured for diameter. Dead stems will be recorded as such, but will be measured. These records of dead stems are to be written and then circled. Obviously, no foliage exists on a dead tree, therefore only the total height will be recorded.
17
Further, elephant damage to the individual will be noted. Qualitative data will be recorded (type of damage e.g. broken side branch, tree pushed over, scar in trunk) and quantitative data will also be recorded (the extent of damage will be described e.g. what percentage of the tree is damaged) following the methods of Anderson & Walker (1974). If the cause of the damage is not known, it will not be speculated upon, as this is often very difficult to determine.
Fire damage will be assessed in order to capture the extent of past fires (e.g. main, primary or secondary stems killed or alive).
Any impact of a known or unknown cause other than elephant and fire will be recorded listing the cause of impact, if known. The following mammals have visible impacts on woody vegetation and will be indicated by the use of the following codes: • Porcupine = P • Giraffe = G • Kudu = K • Buffalo = B • Unknown = X • Other = O
If the impact is not the cause of any of the above, and the cause is known, it will be noted as such in the notes column.
Finally, there will be notes field in the data sheet to record additional discretionary information that may be important.
Herbaceous vegetation sampling The herbaceous biomass of each plot will be estimated by dropping a disc pasture meter every 2 paces across the E to W diagonal of the plot. The disc is not to be dropped onto any woody or other non-herbaceous material, as this would give a false reading. A minimum of 7 readings is required for calculation. The disc pasture meter is calibrated 18 for the conditions and vegetation of the lowveld, and is used to calculate the standing crop or fuel load.
Within each plot, two 1 m2 circular sub-plots will be sampled for herbaceous vegetation. A metal hoop 1.12 m in diameter (1m2 in area) will be placed in the N and then in the S corner. A circle is used as it has the least circumference relative to area, and thus decreases the disputes of “borderline” plants. As with woody plants, only individuals rooted in the circle will be recorded. Two thin metal poles will be laid at right angles at ground level to sub-divide the hoop. This makes sampling much easier as mat forming herbs are difficult to keep track of. Every life form will be counted and recorded, including algal mats (% cover will be estimated), mosses, and ferns. All plants will be identified to the furthest level possible, i.e. species level.
Within each 1m2 circular sub-plots: 1. Every forb will be identified and recorded to determine species abundance and density. When a species occurs in very large numbers, rendering it impractical and inaccurate to count individuals, sub-sampling will be used. A circle of known diameter will be used (e.g. cellotape role). Also, when the total diameter of any forb is less than 1 cm, and many occur in the plot, subsampling will be done by tossing the sub-sample into the ring randomly four times and counting all such minute individuals.
For stoloniferous forbs, each rooted point will be counted as an individual.
2. Every grass species will be identified and recorded.
For grasses that form a tuft base, basal diameter will be measured. For stoloniferous grasses, the diameter of all rooted points will be measured and treated as individuals. This can be used to measure the amount of basal cover.
3. % cover of algal mats will be estimated. 19
Herbarium specimens of all plant species will be collected, preferably not from within the plots, so that these can be compared in future surveys if/when confusion arises regarding species identifications (especially if species have been taxonomically split). Although taking specimens from within the plot destroys seed producers and individuals which may need to be surveyed in the future, this trade-off has been decided upon as the future benefits of a reference collection outweigh the minimal damage that will be done to the site. However, unnatural disturbance (of which specimen collection is an example) must at all times be kept to an absolute minimum within the plots where possible.
Surveying of Alien plants Woody alien plants present at the time of survey will be recorded but not in the same detail as the indigenous species. Total canopy height will be measured and a subjective visual estimate of percentage cover will be made. It would be valuable to monitor the presence of alien populations historically over time subjected to the various treatments. However, alien plant eradication/removal will continue as part of the SANParks/Working for Water management programme.
Instrumentation protocol
Instrumentation is designed to facilitate study of spatial and temporal variation in environmental parameters and nutrient dynamics at critical points along a transitional toposequence from the river edge to the upland boundary. This will provide us with greater insight on the physical template at our sites. We anticipate deploying one meteorological station and twenty-six soil water-monitoring stations at each site (Sabie and Letaba). These will comprise either four or five stations along each of six transects located in each treatment. Sampling stations will be selected to capture local topographic and vegetation variation.
Atmospheric monitoring A simple weather station at each site will measure point rainfall, air temperature, air humidity, wind speed and direction and solar radiation. These data will be logged on a 20
10-minute interval and resolved to yield daily summaries.
Soil monitoring Volumetric soil water contents will be continuously monitored at specific sampling stations using TDR probes installed at various depths in the profile and connected to a data logger. These will provide data that are critical to understanding water and solute movement, both locally and across the toposequence and will be useful in assessing the importance of soil moisture to nutrient transformations. The TDR measurements are sophisticated and expensive, yielding both volumetric soil water contents and soil water electrical conductivity (EC) measurements simultaneously. Two sets of recording instrumentation will be available. These can be moved from station to station to record intervals of soil water and EC measurements where probes have been installed. In addition to the TDR, each station will be equipped with three or four automatic recording tensiometers, positioned at different depths within the profile. The tensiometers will record the soil moisture tension and provide a record of the prevailing direction and gradient of the soil water potential. These data will be used to determine both the water and nutrient balance at each location as well as the mechanisms and rates of vertical and lateral accumulation or uptake.
Soil solution will be collected either in ceramic cup lysimeters or modified wick samplers. Two samplers will be installed at each station: a shallow lysimeter (15 cm depth) for sampling nutrient concentrations in surface soil horizons and a deep lysimeter to sample for changes in solute chemistry as rainwater infiltrates into deeper soil layers.
Gaseous losses are of major interest in determining the fate of nitrogen and carbon. A permanent collar for attachment of a gas collection enclosure will be installed in each sampling station. Permanent installation of collars is necessary to minimize the effects of soil disturbance on gas evasion rates.
Surface runoff is an important vector for loss of nutrients during storms. A sampler for collecting surface runoff during storms will be installed at each sampling 21
station. Runoff collectors will be simple standpipes installed vertically in the soil, with the upper lip slightly above the level of the soil to intercept surface-draining water.
Soil temperature sensors will be installed in six of the sampling stations at different distances along the toposequence
Ground water monitoring Changes in ground water height are important indicators of surface water ground water interactions. Wells for monitoring ground water height and for sampling ground water chemistry will be installed in selected upland sites with deep water tables. At sites where the water table is three meters or less deep and either perched or permanent water tables are anticipated, wells will be installed to a depth of 3m. Water level sensors will be installed in six wells to continuously monitor groundwater height.
Fixed-Point Photography
Fine-scale A final decision has yet to be made on the exact locality of fixed-point sites. The suggestions were made to take one (1) photograph per transect plot along a diagonal line (will still have to decide on the direction) and at a fixed height which necessitates the usage of a tri- or monopod. The problem with this way of taking fixed-point photos is that because only short distances are covered, trees and taller shrubs will not fit entirely onto photographs. An alternative way would be to cover longer distances along the transects, i.e. to take a photograph every 50 m or so while standing on the centre line (middle of transects), thereby including the whole spectrum of height classes of woody plants. Frequency and timing must be consistent, in other words photographs will have to be taken annually during the same time of day of same month.
Coarse-scale In addition to the above-mentioned points, other sites can be selected in order to gain an 22
overall bigger overview of the vegetation across the exclosures. Potential sites include corner points in the fences and or junctions of internal firebreaks with fences. Such sites would be easy to find and also show out differences between burned and unburned blocks. Frequency and timing must be the same as above.
Detailed Site Map Methodology
A once-off detailed (fine-scale) set of aerial photographs will have to be taken during 2003 in order to produce a digital elevation model (DEM) at a spatial resolution of 0.5 m, reflecting three-dimensional differences between vegetation types along the catenas as well as indicating contour lines to illustrate changes in topography. The DEM can be used to assist in compiling detailed maps of the entire exclosures as well as immediate surrounding areas in terms of plant communities, soil types, geology and drainage.
A company will be hired to take the aerial photographs and produce the DEM. The classification of the vegetation, soils and geology can be handed out as research projects to students under good supervision of an academic institution.
Aerial photography
Large and/or small-format aerial photography will have to be conducted on an annual basis approximately the same time of the year. Fine-scale photography at high spatial resolution will be required if even small changes in vegetation cover and drainage lines are to be detected. Changes in the number of large trees, size of sodic patches, bush encroachment and NDVI are amongst the parameters that can be extracted from such photographs. The possibility of Scientific Services conducting the large-format photography will be investigated since the camera belongs to CSIR but is out on loan to Scientific Services. Should it not be possible to pursue this option, the alternative would be to use small- format photography as is currently applied in the rest of the KNP.
23
Management
Site Management An exclosure committee has been formed in order to manage the site in terms of maintenance decisions and scientific studies/surveys. The various members of the committee as well as brief explanations of their respective roles in the exclosure project are: