Accurate estimation of the responses of understory plants to natural and anthropogenic disturbance is essential for understanding efficacy and non-target effects of management and restoration activities. However, ability to assess changes in abundance of understory plants that result from disturbance may be hampered by inappropriate sampling methodologies. Conventional methods for sampling understory plants may be robust for common, well-distributed species, but may fail to adequately characterize the abundance of less-common species, which are often the taxa of management concern. I tested conventional and novel approaches to sampling understory plants to determine their efficacy (in terms of number of replicates and time required) for quantifying abundance of plants of varying frequency and spatial heterogeneity on three control and three thinned-and-burned treatment units located within the western Montana block of the Fire and Fire Surrogates Project (FFS) a large-scale investigation of the effects of fuel-hazard reduction treatments on a variety of ecosystem components. In each treatment unit, I used four sampling methods (modified Whittaker plots, Daubenmire transects, point line intercept transects, and strip adaptive cluster sampling) to estimate the cover of 24 understory species that vary in abundance. Compared to Daubenmire and point line intercept transects, modified Whittaker plots estimated cover with the lowest variances and, consequently, for the majority (67%) of species required the smallest sample sizes to accurately measure cover. However, this greater sampling efficiency was offset by increased time required to sample. For species grouped by growth-form and for common species, all three conventional sampling designs (i.e. Daubenmire transects, modified Whittaker plots, and point line intercept transects) were capable of estimating cover with a 50% relative margin of error with reasonable sample sizes (3-36 plots or transects for growth-form groups; 8-14 for common species); however, increasing the precision to 25% relative margin of error required sampling sizes that may be logistically infeasible (11-143 plots or transects for growth-form groups; 28-54 for common species). In addition, all three designs required enormous sample sizes to estimate cover of non-native species as a group (29-60 plots or transects) and of individual less-common species (62-118 plots or transects), even with 50% relative margin of error. Strip adaptive cluster sampling was the only method tested that efficiently sampled less-common species: for Cirsium arvense, an invasive non-native plant, adaptive sampling required five times fewer replicates than needed for modified Whittaker plots and 20 times less than for Daubenmire or point line intercept transects. My findings suggest that conventional designs may not be effective for accurately estimating the abundance of newly establishing, non-native plants as a group or of the majority of forest understory plants, which are characterized by low abundance and spatial aggregation. Novel methods such as strip adaptive cluster sampling should be considered in investigations for which cover of these species is a primary response variable.

Nitrogen (N) is a primary limiting nutrient in all ecosystems. Therefore, a thorough understanding of N cycling processes in forest ecosystems is required to minimize N losses to fire, harvesting, and other forms of land management. The influence of fire, fire exclusion and forest restoration treatments on non-symbiotic N-fixation in the forest ecosystem has been poorly studied. Over the past 100 years, fire has been greatly excluded from low elevation, fire maintained forests as a result of active fire suppression as well as land management activities that create discontinuities in landscape fire patterns. Previous studies have shown this activity to inhibit recolonization of sites by symbiotic N-fixing plant species. The lack of these important N fixing species may make non-symbiotic, free-living N fixing bacteria a more important source for N recovery in these forest ecosystems following disturbance. Recent studies also suggest that free-living N-fixing bacteria colonizing decomposing woody roots have the capacity to fix large amounts of N. The purpose of these studies was to investigate the effect of fire, fire exclusion, and forest restoration on the N contribution of non-symbiotic N-fixing bacteria (colonizing soil, woody roots, and soil crusts) to the forest ecosystem and how their contribution compares to symbiotic N-fixers in Western Montana. Studies were conducted in the laboratory and at numerous field sites throughout western Montana. In order to determine the N-fixation activity of organisms in these systems, we used the acetylene reduction technique. Neither time since fire, nor restoration treatment had any direct influence on free living N-fixation in soil or woody roots. Moisture and N availability were the potent drivers of free living N-fixation in western Montana. Nitrogen-fixation rates were low in decomposing woody roots in these ecosystems and woody roots do not contribute a significant amount of N to low elevation ponderosa pine/Douglas-fir forests in western Montana. Free-living N-fixing bacteria in soils were found to make a significant, yet modest N contribution to the forest ecosystem. Nitrogen demand by trees and shrubs are being maintained by residual soil organic N, symbiotic N fixation, and wet and dry N deposition in the Inland Northwest

This report discusses the taxonomy, distribution and ecology of the herpetofauna of the Transvaal based on a survey conducted over the past 10 years. Seventeen new forms, apart from four new taxa described during the course of this study, are discussed in the section on Systematics. The distribution of many species has been considerably amended. A total of 265 species and subspecies were recorded from the province. The biogeography of the herpetofauna is discussed in the light of past and present climatic events, as determined by palynological record. Factors affecting reptile and amphibian distributions are incorporated in an attempt to elucidate current distribution patterns. The conservation status of each species and subspecies is discussed and a final chapter on threats and conservation recommendations are incorporated. / Thesis (Ph.D.)-University of Natal, Durban, 1989.

The present report, based on a survey conducted over four years and on the accumulation of museum records, provides the most detailed documentation yet of the herpetofauna of Swaziland. One hundred and two new forms are recorded from the country bringing the total number of forms to 154, consisting of 44 amphibians and 110 reptiles. Up-to-date checklists of the amphibians and reptiles are presented and effectively indicate a rich and diverse herpetofauna. The biogeography of the Swaziland herpetofauna is discussed based on distribution records derived from collected specimens as well as reliable sight and audio records. Swaziland does not constitute a distinctive biogeographical unit. The present study indicates that the herpetofauna shows affinities with both the Afrotemperate and Afrotropical biomes. The traditional biogeographical classification in southern Africa, of the presence of a Cape temperate fauna and a tropical East African lowland fauna, is tested by means of a transect and is reinforced. It is also shown that Swaziland, together with Natal and southern Mozambique, forms an integral part of the tropical subtraction zone of south-east Africa. Amphibian diversity and species turnover in southern Africa are investigated by means of a transect from the east coast, through Swaziland, to the interior plateau, and a north to south transect down the eastern lowveld. The Dice-Sorenson Similarity index gives a value of 41% for the entire east-west transect and 89% for the north-south transect. The conservation status of the amphibians and reptiles of Swaziland is discussed. Conservation measures are proposed. / Thesis (M.Sc.)-University of Natal, Durban, 1992.

Molluscs are an important component of South Africa’s biodiversity. The assessment of distribution patterns and factors influencing the biogeographic distribution are an integral part of assessing the conservation status of molluscs and their conservation management needs. The existing terrestrial mollusc data from South Africa were assessed in terms of their value to biodiversity conversation planning and management. Although the data on terrestrial molluscs are incomplete and would be misleading in terms of identifying specific areas for protection, the data do illustrate significant patterns and trends of mollusc endemism and diversity, which can be used to improve biodiversity conservation and management efforts. The distribution of molluscs across the South African landscape illustrated ten broad biogeographical patterns. Two of these patterns reflected ancient distribution patterns of molluscs and consisted of molluscs of the Gondwanaland/southern relict and Laurasian origins. Three biogeographic patterns occurred across the eastern regions. These patterns were defined as the tropical/subtropical east African, subtropical east of southern Africa and east African afromontane patterns. The biogeographic patterns in the west consisted of the characteristic temperate ‘Mediterranean’ Cape centre and the arid regions of northwestern Cape, Namibia and parts of Botswana. An additional biogeographic pattern identified as the nama karoo/central west was recognised. The final two biogeographical patterns described taxa that were widely distributed and taxa that exhibited disjunct distributions. Twenty-six families and forty-three genera were associated with more than one biogeographical pattern. The dominant biogeographic pattern was the tropical/subtropical east African component. Twenty-one families and forty-eight genera were associated with this biogeographical pattern. The east African Afromontane pattern was also a conspicuous biogeographic element in South Africa. Fewer families and genera were distributed in the western and central regions. The distributions of terrestrial molluscs were influenced by a combination of various factors, which included the presence of rivers, the escarpment, altitude, humidity, precipitation, temperature and biomes. Rivers could possibly restrict the distribution of certain mollusc taxa but did not appear to be the dominant factor that influenced the distribution of molluscs across the landscape. In terms of the effect of temperature on the distribution of molluscs, the mean daily and mean annual temperatures appeared to have more of an influence on the distribution patterns than the mean daily maximum and minimum temperatures. Mean annual temperatures influenced the distribution of all families and genera. The mean daily maximum temperature appeared to have little or no effect on the distribution of mollusc taxa. Humidity and biomes also appeared to influence the distribution of taxa. The least inhabited biome was the succulent biome. Many mollusc taxa occurred in the wetter, warmer areas with high humidity levels. Areas of high species richness and high endemic species richness in South Africa were identified using two systems of endemism, one based on distinctive gaps in the frequency distribution of terrestrial molluscs in South Africa and the other based on an existing classification of invertebrate endemism (Hamer & Slotow, 2002). Areas of high mollusc species richness and endemism were also compared to areas of high millipede species richness and endemism. The total number of South African mollusc endemics was 370 (83 % of 447 indigenous species). The dominant mollusc families in South Africa were Achatinidae, Charopidae, Streptaxidae, Subulinidae and Urocyclidae. The first system of endemism identified 56 site endemics (species with only one locality), 50 local endemics (0 < maximum distance < 60 km) and 145 regional endemics (60 km < maximum distance < 330 km). The Hamer & Slotow (2002) classification of endemism classed 67 species as site endemics (maximum distance between localities < 10 km), 47 as local (11 km < maximum distance < 70 km) endemics and 59 as regional endemics (71 km < maximum distance < 150 km). The analysis of mollusc data, with both systems of endemism, showed similar areas of high species and endemic species richness. Quarter-degree grid cells with highest species richness overlapped with grid cells with the highest number of endemic species. However these grid cells coincide with areas that have been intensively sampled and this bias limits the application of the data in conservation planning. The patterns of endemism for molluscs and millipedes within the provinces differed, indicating that the inclusion of a single taxon in conservation planning would inadequately reflect the diversity of invertebrates in South Africa. A preliminary list of specific priority endemic sites for terrestrial mollusc conservation was identified. It is essential that the existing data on invertebrates be evaluated and used to identify key patterns and trends in invertebrate diversity as this will allow for the inclusion of invertebrates in biodiversity conservation planning and management. The analysis of the existing mollusc data identified bio geographical patterns that are important to conservation planning both at the local and national level as well as commonalities and differences between molluscs and millipede distributions. The analysis also highlighted the importance of municipal areas for conservation of hotspots of diversity, particularly in the eastern coastal areas of South Africa. / Thesis (M.Sc.)-University of KwaZulu-Natal, Westville, 2007.

The taxonomic value of adult laryngeal and larval buccopharyngeal characters, not previously used in studies of southern African frogs, was investigated. The amphibians of Natal, recently examined using more traditional characters (Lambiris 1989), were used both for assessing the value of these characters and for testing the earlier taxonomic conclusions. The results indicate that: (a) Laryngeal and buccopharyngeal characters appear to be of value in anuran taxonomy; (b) Most of the taxonomic conclusions of Poynton (1964) and of Lambiris (1989) appear to be justified; (c) Taxonomic problems in certain of the genera and species-groups examined appear to be more easily resolved by the use of these characters than by those more generally used. The status of six taxa has been redefined; these are Bufo pusillus, Bufo poweri, Bufo cruciger, Breviceps maculatus, Rana umbraculata and Cacosternu striatum. Five undescribed species and one new genus have been identified and await description. / Thesis (Ph.D.)-University of Natal, 1991.

Presented in this thesis are the results of a 23-month field study (April 1991 - February 1993) which focused on infant development in a free-ranging baboon troop (Papio cynocephalus ursinus) inhabiting the Cathedral Peak Reserve in the Drakensberg Mountains, Natal Province (South Africa). The troop selected for study was, in demographic composition, representative of the greater Drakensberg baboon population. During the course of the study, five infants were born into the troop. At the end of the data collection period, the infants ranged in age from II-months to 17 -months old. The data presented are extracted primarily from 2678 half-hourly scan samples, representing approximately 1340 hours of observation. Data presented in this study indicate that baboon births in the Drakensberg are seasonally timed such that mothers are afforded some measure of relief from the energetic costs associated with intense infant dependency, particularly during the first few months postpartum. However, the timing of births is not optimal in terms of the development of infant nutritional independence. This has implications for the interbirth interval (38.4 months) which, for the Drakensberg females, is substantially longer than that for comparative studies, while the rate of infant survivorship through to 12-months (95%) is considerably higher than for other baboon populations. The development of independent feeding for the Drakensberg infants is, when compared with infants from other studies, considerably slower. The delayed transition to independent feeding which was observed is explained with reference to seasonally variable ecological conditions in the Drakensberg which necessarily resulted in infants remaining nutritionally dependent on mothers for a longer time than would normally be expected. The relationship between dependent and independent feeding, as alternate infant strategies, is considered within the context of weaning and its role in the iv promotion of independence. A model which purports to delineate the weaning period with reference to the rate of decline in various measures of maternal investment is applied to the data for this study, as well as for comparative studies. No clear consistency is evident in the rate of decline across the various measures for the present study. The trajectories of infant behavioural development toward the adult model of independent activity are described and considered. With the exception of nutritional independence, all other trajectories of infant behaviour followed similar developmental patterns to infants in comparative studies. Thus the appearance of particular behaviours, and the age at which transitions to independence were made, confirmed a general baboon pattern. The implications and consequences of delayed feeding independence for changes in behavioural development are considered within the context of general activity budgets. Data are also presented which examine the contexts within which adult males interact with infants. These data are considered in relation to the possible functional purposes served by male interaction with infants, as well as in terms of the effect male interaction with infants has on inter-male interaction. The data indicate individual male variability in both the extent and contexts of interaction with infants. / Thesis (Ph.D.)-University of Natal, Durban, 1994.

Beach meiofauna were chosen as environmental indicators to investigate the impact of Illovo sugar by-products effluent. The effluent is pumped through a 20 cm diameter pipeline into surf zone at Sezela beach on the coast of KwaZulu-Natal, South Africa. Meiofaunal communities were considered appropriate indicators as they are relatively stable both qualitatively and quantitatively on a seasonal and year to year basis. Most meiofauna also do not have planktonic stages in their life cycles, respond rapidly to pollution due to their fast generation times, and they are often abundant with high species diversity in habitats which are subject to considerable natural physical and chemical fluctuations. In this particular study there was a specific concern about trace amounts of furfural in the effluent. Furfural has been used as the active ingredient in a product designed to kill parasitic nematodes in crop fields. A large proportion of the beach meiofauna consists of nematodes. Eight stations were sampled for meiofauna along the beaches at Sezela on 7 different occasions. Seasonal effects on meiofauna and meiofaunal recovery during the period when the factory was not pumping effluent to sea was assessed. Samples were taken on the following dates: 4 July 2000 (winter); 30 August 2000 (winter); 13 December 2000 (spring); 26 January 2001 (summer); 8 March 2001 (summer); 9 April 2001 (autumn); and 2 January 2002 (summer). PRIMER (Plymouth Routines in Multivariate Ecological Research) was used for statistical analysis and included various univariate indices such as species richness, species diversity and evenness. These indices were then analysed using one-way ANOVA to determine any significant difference between sites over the 7 sampling periods and between the different seasons. Clustering and Ordination multivariate analyses were carried out on the community data and physico/chemical data to determine community patterns and relate them to the effluent and environmental data. The Nematode/Copepod ratio was also calculated. Meiofauna were analysed at major taxa level, as well as to nematode feeding groups and harpacticoid copepod and annelid family level, to determine if analysis to major taxa level is adequate as an indicator of pollution impact. The analyses indicated a possible degree of impact at stations close to the effluent discharge when effluent was being pumped to sea and a recovery was noted at the station closest to the discharge when effluent was not being discharged and analysis was conducted to the major taxonomic rank only. No improved resolution was achieved by analysing some of the meiofaunal major taxa to family level or different feeding groups. The analysis of the Nematode/Copepod ratio was shown to correspond with the multivariate analyses, however, this ratio could not reveal the severity of the impact where both nematodes and harpacticoids i.e. total meiofauna had been reduced by adverse conditions. The physical and chemical variables that showed the greatest correlation with the meiofaunal community patterns were sediment grain size, dissolved oxygen and salinity. There was a very strong positive correlation between Kjeldahl nitrogen in the interstitial waters and total numbers of meiofauna. This and the relationship with salinity may have suggested other possible sources of influence such as enrichment from the three estuaries in the area as well as a storm water drain located 150m north of the effluent discharge. A seasonal effect was observed with increased meiofauna numbers in autumn, but this was possibly influenced by the periods when effluent was not being pumped to sea. / Thesis (M.Sc.)-University of KwaZulu-Natal, Westville, 2007.

Environmental Impact Assessment (El A) is a process that is widely practised as it assists in decision-making and also helps to overcome the environmental problems that could result from development activities. However, the focus is still on EIA as a process and less on EIA follow-up. EIA follow-up is taken to mean the activities, such as monitoring and auditing, that are carried out after the Record of Decision has been made, although the importance of establishing EIA follow-up early in the project cycle is emphasised in this thesis. In most countries, EIA follow-up is not legislated and whilst it is generally recognized as important it is not widely practised. This thesis is aimed at assessing the status of EIA follow-up in Lesotho. Nine development projects were selected and their reports; Environmental Impact Assessment Reports (EIRs), Environmental Management Plans (EMPs) and auditing reports were analysed to determine if there was provision for EIA follow-up. Four criteria were utilised in the analysis. These were: the impacts that were predicted and mitigation measures proposed, the provision made for EIA follow-up before the implementation of the project, the impacts that were experienced and the mitigation measures that were put in place and the EIA follow-up process that was undertaken, and the people responsible for it. All projects had undergone an EIA process, except for one which did not have an EIR prepared, viz. C& Y garment factory at the Thetsane industrial site. Of the remainder, four projects contained provision for EIA follow-up, although in most case studies follow-up focused on the construction phase and little was stated about the implementation of follow-up. Generally, an environmental officer was appointed to monitor the impacts that were experienced and to ensure compliance with the EMP. However, in the Butha-Buthe industrial estate case study, the EIA follow-up process was detailed and specific, even giving the frequency with which EIA follow-up should be undertaken, by whom and how it should be done. This is most likely because it is the most recent industrial estate to be developed and that lessons were learned from previous industrial development sites discussed as case studies in this thesis, where problems were encountered due to lack of EIA follow-up. Apart from the assessment of these reports, questionnaires were also administered to nine environmental consultants practising in Lesotho. Monitoring and auditing were identified as EIA follow-up by the majority of consultants (7 or 78%). Only one person identified it as including public participation, while the other person (11 %) identified it as monitoring, which incorporates EMPs and Environmental Management Systems (EMSs). It was interesting to note that only one person included public participation as part of EIA follow-up, in contrast to the general understanding of EIA follow-up internationally, that the public have a role to play in follow-up activities. One person (1 or 11 %) pointed out that EIA follow-up should start at the planning or design stage, while the majority (89%) stated that it should start after the completion of the EIA process and the Record of Decision, the latter group failing to recognise the importance of collecting baseline data early in the EIA process. Of all the projects, only the Lesotho Highlands Water Project (LHWP) was observed to implement EIA follow-up, such as monitoring and auditing, on a regular basis. An assessment was also undertaken of the environmental legislation in Lesotho and the provision that it makes for EIA follow-up. Sections 31 and 32 of Part V of the Act specifically give provision for EIA follow-up. It is stated that in order to prevent environmental degradation, environmental monitoring and environmental auditing should be undertaken. Moreover, the Lesotho EIA guidelines (1997) do give guidance and procedures on how EIA follow-up should be undertaken. However, it was found that currently, the Environment Act, 2001 is not operational and that EIA follow-up like the EIA process is undertaken on a voluntary basis. It was therefore recommended that at present, the self-regulatory approach to EIA follow-up is the most suitable one for Lesotho. Recommendations were made to strengthen this approach until such time as legislation is in place or an environmentally aware public can participate in EIA follow-up. Several problems were identified that were hampering the practice of EIA follow-up in Lesotho. These included: the un-operational Environment Act, an environmentally unaware public, few environmentalists and lack of sensitive and dedicated government ministries. / Thesis (M.Sc.)-University of Natal, Durban, 2003.

The taxonomic status, distribution and ecology of the herpetofauna of the Orange Free State are discussed, based on the examination of 10 096 specimens housed in various southern African museums. The majority of specimens were collected during the years 1972 - 1978 and 1983 - 1992, and are housed at the National Museum, Bloemfontein. A total of 25 amphibian and 95 reptilian (one translocated species) forms have been determined as occurring in the Orange Free State, 12 (three amphibians, nine reptiles) of which are new (or recently published) records for the province. An additional 10 amphibian and 28 reptilian forms have been determined as occurring nearby but extralimitally to the O.F.S., and several of these forms are expected to occur in the province. New distributional records have resulted in the amendment of the ranges of several species. Detailed taxonomic data on new material, including rare species, have been given. Some problem areas in the taxonomy of O.F.S. taxa have been higlighted. An analysis of habit utilization indicated that 84,0% of amphibians and 61,1% of reptiles are terricolous, whereas up to 21,5% of reptiles are rupicolous. The majority of snakes (72,2%) are terrestrial in habits, but 13,9% are fossorial. More than half of all lizards (52,8%) are terrestrial, although 34,6% are rupicolous. A total of six amphibian and 27 reptile forms utilize inactive termitaria as a microhabitat, including several basically terrestrial forms. Snake forms were particularly well represented in termitaria, 60,6% of all forms known from the O.F.S. having been recorded from inactive termitaria. General features of the ecology of O.F.S. amphibians and reptiles have also been discussed. The biogeographical analysis indicated that O.F.S. amphibian forms can be classified into one of nine range clusters (common patterns of distribution), and reptiles into 13 such cluster groups. By testing these classifications by means of a transect through the northern O.F.S. (from eastern to western borders), it was determined that a fairly distinct east-west subtraction of amphibian and reptilian species and subspecies occurs in the O.F.S. Clustering of range boundaries and high species and subspecies diversity at the western and eastern ends of the transect zone suggest dynamic biogeographical situations occurring in those areas - the western group being associated with the transition from grassland to bushveld, and the eastern group associated with the transition from Highveld Grassland to Drakensberg Mountains. The general eastern and western groupings of taxa appear to be associated with the cooler, wetter and mountainous east vs the warmer, drier and lower-lying west, respectively. Despite a great deal of collecting having been conducted in the O.F.S. from 1972 to 1992, an analysis of the number of taxa collected in each quarter-degree unit in the O.F.S. indicated that additional collecting would be required in order to conduct effectively a mathematically-based biogeographical analysis. / Thesis (M.Sc.)-University of Natal, 1992.