Uncertainty and urban water recharge for managing groundwater availability using decision support.

Quantifying groundwater availability depends upon sound methods and the use of integrated models. To determine availability or sustainable yield, the influence of scientific uncertainty from key sources, such as anthropogenic recharge, must be considered. This study evaluates uncertainty in recharge interpretations on the modeled available water balance for an urban case in Texas, USA. Analyses are completed using the Groundwater Decision Support System, which is a research code-base for an integrated modeling. The case study develops spatially and temporally resolved recharge interpretations based on NEXRAD precipitation and detailed land use data. Results demonstrate the implications of scientific uncertainty as it influences recommendations for policy and urban water management decisions that are based on modeled outputs. Geospatial methods account for spatial and temporal components and can be replicated for other systems. These methods are also useful for resolving uncertainty in relation to the influence of urbanization on recharge through land use change.

Impact of the spatial scale of grass-legume mixtures on sheep grazing behaviour, preference and intake, and subsequent effects on pasture.

The benefits of using white clover in pastures have been widely recognised for many years. However, clover is perceived as being unreliable because of its typically low content, which is spatially and temporally variable, in mixed pastures. One proposed solution to increase the proportion of clover in the diet of grazing animals and composition in the pasture is to spatially separate clover from grass within the same field. In a field experiment ryegrass and white clover were grown in fine mixtures, and in pure alternating strips of ryegrass and clover of 1.5 m, 3 m or 18 m width within a field. Pastures were grazed for two grazing periods of 9 and 12 weeks, and measurements of sward surface height (SSH), herbage mass and composition and clover morphology were taken. Grazing behaviour was also observed. Results showed that spatial separation in the long term, when compared with a fine mixture, increased clover availability (18% to 30% v. 9%, based on standing dry matter) and was not grazed to extinction. Ewes maintained their preference for clover throughout the experiment (selection coefficient 2 to 5), which resulted in a reduction in the SSH of clover in monocultures to <3 cm and significant changes to the morphology of clover (smaller leaves, shorter petioles and thicker stolon), at the expense of maximising their intake. Spatial separation in the short term may therefore allow grazing animals to select their preferred diet; however, in the long term in continuously grazed pasture, their preference for clover depletes its availability.

The distribution of immune cells in the lungs of classical and atypical ovine pulmonary adenocarcinoma.

Ovine pulmonary adenocarcinoma (OPA) is a contagious tumour caused by infection of sheep with Jaagsiekte sheep retrovirus. Two forms of OPA have been identified, classical and atypical, which can be distinguished clinically and pathologically. Most notably classical OPA is progressive until death, while atypical OPA remains subclinical. In the present study the local immune responses in the lungs of cases of atypical OPA were compared with those from classical cases by immunohistochemistry using a panel of mouse anti-sheep mAbs. Distinct differences in the distribution of immune cell subsets in the two forms of OPA were observed. In particular there was an intratumoural influx of T cell subsets and MHC Class II expression on the tumour cells in atypical OPA, neither of which was seen in classical OPA. It is possible that these differences may contribute, at least in part, to determining the progressive course of classical OPA compared with the subclinical nature of atypical OPA.

Eradication of ovine pulmonary adenocarcinoma by motherless rearing of lambs.

The principles of maedi-visna eradication programmes were applied to a field trial for the eradication of ovine pulmonary adenocarcinoma (OPA). In two maternal flocks the prevalence of gross and histological lesions in slaughtered animals was 18.3 per cent and 29.8 per cent, respectively. The lambing period was supervised for three consecutive years from 1999 to 2001, during which the lambs were taken away from their mothers at birth, deprived of maternal colostrum, and hand-reared away from other sheep. Over the three-year period, 322 hand-reared animals, mainly male lambs between 10 and 14 months old, were slaughtered; their lungs were examined grossly, 52.5 per cent of them were examined histologically, and 105 samples of caudal mediastinal lymph nodes were examined by PCR. No OPA tumours were detected in the slaughter specimens from the derived flock, but one lamb had histological lesions in one lung location; intrauterine transmission was ruled out in this case. No clinical OPA has subsequently been observed in the hand-reared flock. Bronchoalveolar lavage samples from the breeding stock were examined by PCR in order to rule out further subclinical cases of OPA. No Jaagsiekte retrovirus was detected in any of the 488 samples.

PCR examination of bronchoalveolar lavage samples is a useful tool in pre-clinical diagnosis of ovine pulmonary adenocarcinoma (Jaagsiekte).

Ovine pulmonary adenocarcinoma (OPA) is a contagious lung tumour of sheep caused by Jaagsiekte sheep retrovirus (JSRV). The disease is a particular problem in flocks in many parts of the world. The aim of the study was to assess screening methods for individual animals as a prelude to future eradication trials. Results of histological examination were used as the standard to evaluate the relative sensitivity and specificity of an established heminested polymerase chain reaction (PCR) test for JSRV proviral DNA from blood and bronchoalveolar lavage (BAL) samples. PCR results from tissue samples are included as control data. PCR testing of blood samples was found to have an estimated sensitivity of only 10% (95% confidence interval (CI) 3-20) while the sensitivity of the PCR test on BAL samples was 89% (CI 79-96) in comparison to the results of histological examination. We conclude that PCR testing of BAL samples is an effective confirmatory test for sheep with suspected clinical OPA. It is also a useful tool for the pre-clinical identification of individual infected sheep within an infected flock and therefore may prove beneficial in future control or eradication programmes.

In-situ demonstration of mitogen-activated protein kinase Erk 1/2 signalling pathway in contagious respiratory tumours of sheep and goats.

Ovine pulmonary adenocarcinoma (OPA) and enzootic nasal adenocarcinoma (ENA) are two contagious neoplastic diseases of secretory epithelial cells in the respiratory system of sheep and goats. Jaagsiekte sheep retrovirus (JSRV) is the aetiological agent of OPA, and enzootic nasal tumour virus (ENTV) is associated with ENA. The genomes of these retroviruses do not contain known oncogenes but products of the env gene are important in the generation of transforming stimuli. However, the cell signalling pathways activated in vivo are not completely understood. This study was based on the use of activation stage antibodies specifically detecting proteins of the extracellular signal regulated kinase Erk 1/2 cell signalling pathway and transcription factors. Tissue sections were collected from four natural cases of OPA, four experimentally induced OPA tumours, four ENA tumours in sheep, four ENA tumours in goats, two normal sheep lungs and two lungs with chronic inflammation. Routine immunohistochemical procedures with phosphorylation stage-specific antibodies were carried out. Representative proteins of the Erk1/2 pathway (Raf-1, Mek1/2 and p44/42MAPK) were activated in natural cases of OPA and ENA in sheep and goats and also in experimentally induced OPA. Transcription factors 90Rsk and Elk-1 were activated in OPA and ENA tumours. However, c-Myc was activated only in OPA tumours. In contagious respiratory neoplasms of sheep and goats the Erk1/2 pathway appears to be important for the in-vivo generation of the transforming stimuli.

The molecular epidemiology of Mycobacterium bovis infections in Tanzania.

A molecular epidemiological study to determine the zoonotic importance of bovine tuberculosis was carried out in Tanzania. Specimens from human cases of tuberculosis as well as from slaughtered cattle were collected from regions with a high proportion of extrapulmonary tuberculosis. In order to determine the similarity of strains from the two sources, molecular typing techniques, namely RFLP and spoligotyping, were used to determine the genetic profile of the strains involved. The results of pTBN12 typing of M. bovis from cattle and man has shown a rather heterogeneous population of this species spread all over Tanzania, assuming that the present sample is representative. There were 13 different pTBN12 RFLP types encountered. The genetic relatedness between the pTBN12 RFLP patterns indicated a high degree of relatedness (86%) between the dominant pTBN12 genotypes existing in Tanzania. There were 13 different spoligotypes found in this study, whose genetic relatedness was also high (79%). DNA profiles were also confirmed by IS986 RFLP, which revealed that strains have 1-13 copies of IS986. Geographically, there was overlap between pTBN12 RFLP and spoligotypes amongst strains isolated from various parts of Tanzania. The diversity of the RFLP and spoligotype patterns observed in Tanzania probably reflects the extensive internal movements of cattle belonging to pastoralists. The evidence of overlap between DNA fingerprints of M. bovis from cattle and man has once more highlighted a need for synergy of veterinary and medical policies in the control of tuberculosis in Tanzania and probably in other developing countries.

An influx of macrophages is the predominant local immune response in ovine pulmonary adenocarcinoma.

Infection with a retrovirus, Jaagsiekte sheep retrovirus (JSRV), causes ovine pulmonary adenocarcinoma (OPA). The excess production of surfactant proteins by alveolar tumour cells results in increased production of pulmonary fluid, which is characteristically expelled through the nostrils of affected sheep. The immune response to JSRV and the tumour is poorly understood: no JSRV-specific circulating antibodies or T cells have been detected to date. The aim of the present study was to obtain phenotypic evidence for a local immune response in OPA lungs. Specific-pathogen free lambs were infected intratracheally with JSRV. When clinical signs of OPA were apparent, the lungs were removed at necropsy and immunohistochemistry (IHC) was performed on lung sections using a panel of mouse anti-sheep mAbs. No influx of dendritic cells, B cells, CD4, CD8 or gammadelta T cells was seen in the neoplastic nodules or in their periphery. MHC Class II-positive cells were found intratumourally, peritumourally and in the surrounding alveolar lumina. In the tumours, many of these cells were shown to be fibroblasts and the remainder were likely to be mature macrophages. In the alveolar lumen, the MHC Class II-positive cells were CD14-positive and expressed high levels of IFN-gamma. They appeared to be immature monocytes or macrophages which then differentiated to become CD14-negative as they reached the periphery of the tumours. A high level of MHC Class I expression was detected on a range of cells in the OPA lungs but the tumour nodules themselves contained no MHC Class I-positive cells. On the basis of these findings, it is proposed that the lack of an effective immune response in OPA could result from a mechanism of peripheral tolerance in which the activity of the invading macrophages is suppressed by the local environment, possibly as a consequence of the inhibitory properties of the surfactant proteins.

Genetic diversity of Mycobacterium avium subspecies paratuberculosis isolates from goats detected by pulsed-field gel electrophoresis.

Paratuberculosis in goats occurs worldwide causing considerable economic losses mainly due to reduced milk production. Nowadays, there is still relatively little knowledge about the epidemiology of this disease in goats, and only a few epidemiological studies have been carried out in goats naturally infected with Mycobacterium avium subspecies paratuberculosis (M. a. paratuberculosis). The objective of this study was to characterize forty four clinical caprine isolates of M. a. paratuberculosis by different molecular techniques (pulsed-field gel electrophoresis [PFGE], restriction fragment length polymorphism analysis coupled with hybridization to IS900, and IS1311 polymerase chain reaction-restriction enzyme analysis) to determine the most useful technique for molecular typing of caprine isolates, as well as to disclose the genetic variation amongst caprine isolates and the relationship with strains isolated from other animal species. PFGE was found to be the most discriminative technique identifying a total of 13 'multiplex' PFGE profiles, ten of which were novel profiles found only in caprine isolates to date. All isolates were genotyped as Type II strains, except two isolates that resembled the intermediate group referred as Type III.

Coexistence of enzootic nasal adenocarcinoma and jaagsiekte retrovirus infection in sheep.

Ten sheep naturally affected with enzootic nasal adenocarcinoma (ENA), a disease associated with ovine enzootic nasal tumour virus (ENTV-1), were found also to be infected with jaagsiekte sheep retrovirus (JSRV), the causal agent of ovine pulmonary adenocarcinoma (OPA). Only one of the sheep showed OPA lung lesions. The animals belonged to 10 flocks located in a geographical area in which OPA is frequently seen. ENTV-1 was found in all the ENA tumours but only occasionally in extra-tumoral sites, confirming the results of a previous study. In contrast, JSRV had a disseminated tissue distribution, similar to that previously reported for animals infected with JSRV. However, the occurrence of JSRV in lymphoid tissues was clearly greater than in sheep infected with JSRV but with no lesions of ENA. The data suggested a synergistic relationship between ENTV-1 and JSRV, resulting in increased proliferation of JSRV.

Analysis of multidrug-resistant Mycobacterium bovis from three clinical samples from Scotland.

OBJECTIVE: To determine whether multidrug-resistant (MDR) strains of Mycobacterium bovis isolated from patients in Scotland were genotypically related. DESIGN: Genotypes of MDR strains were determined using three molecular fingerprinting techniques: pulsed-field gel electrophoresis (PFGE), spoligotyping and restriction fragment length polymorphism (RFLP). PFGE profiles were also obtained for all medical and veterinary isolates occurring in Scotland in 1997-1998. RESULTS: MDR strains showed individual Dra I PFGE profiles. Case III/98 had a profile represented in both veterinary and medical populations, Case I/94 had a profile observed in medical but not veterinary isolates, and Case II/98 had a profile unique to this study. Afl II PFGE discriminated the resistant strains. Spoligotyping grouped Cases I/94 and II/98 (ST-134). Case III/98 had a spoligotype ST-140, which is commonly observed in veterinary isolates. Similarly, DRr-RFLP analysis grouped cases I/94 and II/98, whereas Case III/98 had a common veterinary profile. DRX(PGRS) RFLP gave three unique profiles. CONCLUSION: Three resistant strains were discriminated by PFGE and DRX(PGRS) RFLP, indicating that the three strains are not related in an epidemiologically relevant time scale. However, Cases I/94 and II/98 were more closely linked by spoligotyping and DRr-RFLP data. PFGE and DRr-RFLP linked Case III/98 profiles to the most common veterinary isolate.

Tribal differences in perception of tuberculosis: a possible role in tuberculosis control in Arusha, Tanzania.

SETTING: Arusha, Tanzania. OBJECTIVE: To determine tribal differences in knowledge and practices that might influence tuberculosis control. METHOD: Twenty-seven villages were selected randomly out of 242 villages in four districts. In each village, a general and a livestock keeping group were selected at random. The households were home-visited and 426 family members were interviewed. RESULTS: On average, 40% of respondents practised habits that might expose them to both bovine and human tuberculosis. The Barabaig tribe had a significantly higher number of respondents (50%, chi2(2) = 5.1, P = 0.024) who did not boil milk. Eating uncooked meat or meat products was practised by 17.9% of all respondents. The habit was practised more by Iraqw (21.1%, chi2(2) = 6.9, P = 0.008) and Barabaig (31.6%, chi2(2) = 5.6, P = 0.016) than other tribes. About 75% of the respondents had a poor knowledge of tuberculosis. CONCLUSION: All tribes had habits and beliefs that might expose them to both bovine and human tuberculosis. The Iraqw and Barabaig tribes practised such habits more than other tribes. Knowledge of tuberculosis was limited in all tribes.

The role of livestock keeping in tuberculosis trends in Arusha, Tanzania.

SETTING: Arusha, Tanzania. OBJECTIVE: To assess risk factors that might influence TB control in the general population and in livestock-keepers. METHODS: Of 242 villages in four districts, 27 were selected randomly. In each village, a general and a livestock-keeping group were selected at random. The households were home-visited and 426 family members were interviewed. RESULTS: On average, three-quarters of households practised at least one risk activity for transmission of zoonotic tuberculosis, and respondents had poor knowledge about tuberculosis. In the livestock-keeping group, the risks of having a tuberculosis patient in the family were determined by poor ventilation (OR 2.6, 95% CI 1.1-6.5), confining livestock indoors with people (OR 2.3, 95%CI 1.1-5.0) and multiple determinants including poor ventilation (OR 13.5, 95% CI 2.5-71.7). Risk activities and the risks of having a tuberculosis patient in a family were significantly higher in the livestock-keeping group. CONCLUSIONS: The respondents had limited knowledge about tuberculosis, and the households had practices that posed potential risks for both human and bovine tuberculosis infection. Poor ventilation and confining livestock indoors were associated with tuberculosis spread in the households. These risks were observed more in the livestock-keeping group than in the general population group.

The potential role of wild rabbits Oryctolagus cuniculus in the epidemiology of paratuberculosis in domestic ruminants.

Mycobacterium avium subspecies paratuberculosis, the organism responsible for paratuberculosis in cattle and sheep has been found in wild rabbits (Oryctolagus cuniculus) in the east of Scotland. Few studies have investigated either the level of faecal contamination by rabbits on farms, or the potential infectivity of rabbit excreta. The rate of rabbit faecal contamination deposited and the numbers encountered were estimated for 21 fields on 4 farms with a paratuberculosis problem. 7357 +/- 2571 S.E.M. rabbit faecal pellets were deposited per hectare per day and up to 81,000 pellets/ha ('standing crop') were encountered in October/November 1998. Where access to rabbits was restricted, the standing crop of faeces encountered fell to 22,000 pellets/ha. The prevalence of infection with M. a. paratuberculosis was assessed for 83 rabbits from the four farms. M. a. paratuberculosis was isolated from rabbits on all farms with an overall prevalence of 17%. Out of 17 rabbits from which urine was available, M. a. paratuberculosis was isolated from two--the first reported isolation from urine in wild rabbits. The mean number of colony-forming units per gram of infected rabbit faeces was 7.6 x 10(5) +/- 5.2 x 10(5). A relative estimate of the input of M. a. paratuberculosis onto pasture, at the stocking levels found on the four farms, showed that sheep and cattle potentially contributed 4 and 125 times more organisms/ha per day respectively than rabbits. However, rabbits could still contribute millions of M. a. paratuberculosis organisms per ha per day. Existing rabbit control measures on farms may be inadequate in reducing the risk of transmission to livestock.

Pathology of ovine pulmonary adenocarcinoma.

Clinical, gross pathology, histopathology and electron microscopy of the ovine pulmonary adenocarcinoma (OPA, jaagsiekte) either natural or experimentally induced in sheep, goat and moufflon are described. OPA is caused by an oncogenic betaretrovirus,jaagsiekte sheep retrovirus (JSRV). Most natural cases of OPA appear in animals 1-4 years old. There is no evidence of sex or breed susceptibility. Sheep affected by OPA show an afebrile respiratory illness associated with loss of weight. A very characteristic clinical sign is moist rales caused by the accumulation of fluid in the respiratory airways which is discharged from the nostrils when the head is lowered. Gross lesions are confined to the lungs but occasionally thoracic or extrathoracic structures are also affected. Two pathologic forms of OPA are currently recognized, classical and atypical. In classical forms the neoplastic lesions occurs particularly in the cranioventral parts of all lung lobes. They are diffuse or nodular, light grey or light purple in colour. On the cut surface the tumour is moist, and frothy fluid may pour from the airways on slight pressure. Atypical forms tend to be more nodular in both early and advanced tumours. They are pearly white in colour, very hard in consistency, very well demarcated from the surrounding parenchyma and their surface is dry. Histology of the lung sections reveals the presence of several foci of epithelial cell neoplastic proliferation in both alveolar or bronchiolar regions. The tumours, derived from type II pneumocytes and Clara cells, proliferate into mostly papillary but also acinar or occasionally solid growths. The tumour generally shows a benign histological pattern but intra- and extrathoracic metastases have been detected in some cases. Several considerations suggest that the tumour should be classified as an adenocarcinoma of the lung. The histology of atypical OPA is similar to that of the classical disease, with an increase in the stromal reaction accompanying the epithelial proliferations. Pathological features of OPA induced experimentally in sheep, or of OPA in goats and moufflon are similar to those described in sheep. Detailed electron microscopy of tumour material confirms that type II pneumocytes and Clara bronchiolar epithelial cells are the origin of the neoplasia. Also included in this chapter is a description of the morphology of the viral particles associated with OPA.

Natural history of JSRV in sheep.

Ovine pulmonary adenocarcinoma (OPA) is a contagious lung tumour of sheep and, rarely, goats that arises from two types of secretory epithelial cell that retain their luxury function of surfactant synthesis and secretion. It is classified as a low-grade adenocarcinoma and is viewed as a good model for epithelial neoplasia because of its morphological resemblance to the human lung tumour, bronchioloalveolar adenocarcinoma. OPA is present in most of the sheep rearing areas of the globe and, in affected flocks, tumours are present in a high proportion of sheep. OPA is associated with the ovine retrovirus, jaagsiekte sheep retrovirus (JSRV), and is transmissible only with inocula that contain JSRV. All sheep contain JSRV-related endogenous viruses, but JSRV is an exogenous virus that is associated exclusively with OPA. JSRV is detected consistently in the lung fluid, tumour and lymphoid tissues of sheep affected by both natural and experimental OPA or unaffected in-contact flockmates and never in sheep from unaffected flocks with no history of the tumour. JSRV replicates principally in the epithelial tumour cells, but also establishes a disseminated infection of several lymphoid cell types, including peripheral blood leukocytes (PBLs). Longitudinal studies in flocks with endemic OPA have revealed JSRV in PBLs before the onset of clinical OPA and even in the absence of discernible lung tumour. The prevalence of JSRV infection is 40%-80%, although only 30% of sheep appear to develop OPA lesions. A unique feature of OPA is the absence of a specific humoral immune response to JSRV, despite the highly productive infection in the lungs and the disseminated lymphoid infection. This feature is associated with reduced responsiveness to some mitogens, although the phenotypic profile of the peripheral blood remains unaltered. The reduced response is an early and sustained event during infection and may indicate that the failure of infected sheep to produce specific antibodies to JSRV is a direct consequence of infection.

Enzootic nasal adenocarcinoma of sheep and goats.

Enzootic nasal adenocarcinoma is a contagious tumour of the mucosal nasal glands affecting young adult sheep or goats. The disease occurs naturally in all continents except Australia and New Zealand. Clinical signs include continuous nasal discharge, respiratory distress, exophthalmos and skull deformations. The tumour is classified histologically as a low-grade adenocarcinoma. Nasal glands of both respiratory and olfactory muosal glands seem to be the origin of the neoplasia. It has been experimentally transmitted in sheep and goats using either tumour extracts or concentrated nasal fluids. Two distinct retroviruses are implicated in the aetiology of the neoplasia one in sheep (ONAV) and one in goats (CNAV). We suggest that jaagsiekte sheep retrovirus (JSRV), ONAV, CNAV, and their endogenous counterparts represent a unique family of retroviruses. The similarities between these viruses suggests that any control strategies, including vaccination, may be appropriate to both diseases. The differences, however, represent a unique resource for delineating the function of individual regions of the virus. It is intriguing that whilst ONAV and CNAV appear to be as different to each other as they are to JSRV, that they have very similar disease pathologies, distinct from that of OPA. Additionally, all three exogenous viruses manage to avoid instigating any apparent immune response. Whether this is indeed a result of tolerance induced by the endogenous counterparts or whether the viruses themselves have unique immunosuppressive properties will be an important finding.

Strategies for enhancing monoclonal antibody accumulation in plant cell and organ cultures.

Various strategies aimed at improving IgG(1) antibody accumulation in transgenic tobacco cell and organ cultures were tested. The form of tissue had a significant effect on antibody levels; shooty teratomas were less productive than hairy roots or suspended cells. Although there were several disadvantages associated with hairy roots compared with suspensions, such as slower growth, slower antibody production, and formation of a greater number of antibody fragments, the roots exhibited superior long-term culture stability. Antibody accumulation in hairy root cultures was improved by increasing the dissolved oxygen tension to 150% air saturation, indicating the need for effective oxygen transfer in root reactors used for antibody production. Preventing N-linked glycosylation using tunicamycin or inhibition of subsequent glycan processing by castanospermine reduced antibody accumulation in the biomass and/or medium in cell suspensions. Loss of antibody from the cultures after its secretion and release into the medium was identified as a major problem. This effect was minimized by inhibiting protein transport in the secretory pathway using Brefeldin A, resulting in antibody accumulation levels up to 2.7 times those in untreated cells. Strategies for protecting secreted antibody, such as addition of poly(vinylpyrrolidone) and periodic harvesting from the medium using hydroxyapatite resin, also increased antibody titers. The mechanisms responsible for the disappearance of antibody from plant culture media were not clearly identified; degradation by proteases and conformational modification of the antibody, such as formation of aggregates, provided an explanation for some but not all the phenomena observed. This work demonstrates that the manipulation and control of culture conditions and metabolic processes in plant tissue cultures can be used to improve the production of foreign proteins. However, loss of secreted antibody from plant culture medium is a significant problem that may limit the feasibility of using product recovery from the medium to reduce downstream processing costs relative to agricultural systems.

Variable-density groundwater flow and solute transport in heterogeneous porous media: approaches, resolutions and future challenges.

In certain hydrogeological situations, fluid density variations occur because of changes in the solute or colloidal concentration, temperature, and pressure of the groundwater. These include seawater intrusion, high-level radioactive waste disposal, groundwater contamination, and geothermal energy production. When the density of the invading fluid is greater than that of the ambient one, density-driven free convection can lead to transport of heat and solutes over larger spatial scales and significantly shorter time scales than compared with diffusion alone. Beginning with the work of Lord Rayleigh in 1916, thermal and solute instabilities in homogeneous media have been studied in detail for almost a century. Recently, these theoretical and experimental studies have been applied in the study of groundwater phenomena, where the assumptions of homogeneity and isotropy rarely, if ever, apply. The critical role that heterogeneity plays in the onset as well as the growth and/or decay of convective motion is discussed by way of a review of pertinent literature and numerical simulations performed using a variable-density flow and solute transport numerical code. Different styles of heterogeneity are considered and range from continuously "trending" heterogeneity (sinusoidal and stochastic permeability distributions) to discretely fractured geologic media. Results indicate that both the onset of instabilities and their subsequent growth and decay are intimately related to the structure and variance of the permeability field. While disordered heterogeneity tends to dissipate convection through dispersive mixing, an ordered heterogeneity (e.g., sets of vertical fractures) allows instabilities to propagate at modest combinations of fracture aperture and separation distances. Despite a clearer understanding of the processes that control the onset and propagation of instabilities, resultant plume patterns and their migration rates and pathways do not appear amenable to prediction at present. The classical Rayleigh number used to predict the occurrence of instabilities fails, in most cases, when heterogeneous conditions prevail. The incorporation of key characteristics of the heterogeneous permeability field into relevant stability criteria and numerical models remains a challenge for future research.