Researchers and advocates in Uganda used data on deforestation, costs and benefits of tobacco farming and other issues to develop public support for effective tobacco control policies and for the Framework Convention on Tobacco Control. In February and March 2004, researchers with the Environmental Action Network conducted a survey among farmers in Uganda focusing on deforestation and economic and health status. Researchers interviewed government officials on the same issues. Findings revealed that farmers in Uganda suffer worsening poverty and poor health associated with tobacco growing. The project is a best practice to retrieve and organize data on the social and environmental costs of tobacco growing. In the U.S., tobacco farmers and tobacco control advocates committed to reducing disease caused by tobacco and ensuring the prosperity and stability of tobacco farmers, their families and communities. Beginning in 1994, national groups such as the National Black Farmers Association and the Campaign for Tobacco-Free Kids, state groups such as the Coalition for Health and Agricultural Development in Kentucky and the North Carolina Council American Cancer Society, and regional groups such the Burley Tobacco Growers Cooperative and the New England Society of Public Health Education participated in meetings with representatives of all groups affected by tobacco, provided expertise to educate participants of similar and opposing positions, vertical farming supplies and encouraged tobacco dialogue to strengthen alliances between farmers and health advocates. Cooperation and commitment to promote tobacco farmer prosperity and public health renders false the dichotomy between policies for tobacco agriculture development and policies directed at the reduction of tobacco use.
The common ground established by farmers and health groups in the U.S. is a best practice that could be used to build partnerships for tobacco farmer welfare and tobacco control in developing countries. Conclusion Tobacco farming contributes to poverty and insufficient economic development in developing countries. Farmers under contractual obligations to tobacco companies or farm landlords are vulnerable to leaf downgrading, suppressed tobacco prices, and inflated prices for inputs. Bonded labor prevents farmers from receiving earnings to cover costs for inputs, food requirements, and health care needs. Child labor undermines children’s education and threatens their health and physical growth, pushing children into a cycle of poverty. Tobacco farming involves wood use for curing and pesticides and fertilizers that destroy forests and pollute soils and water tables. Tobacco farming erodes the lives of present and future generations of farmers, harming human and land capital, key assets for rural development, that could otherwise be devoted to healthy crops and environmentally friendly agriculture.Ecological interactions are increasingly recognized as being highly contingent on their context, shaped by forces that are both historical and contemporary as well as biotic and abiotic . For example, variation between years and sites can have profound influences on the outcomes of field experiments in community ecology . If we want the results of ecological experiments to be general, and not unique to a particular site or time, we need to better explore and understand these and other contingencies. Understanding such contingencies is also crucial for successfully restoring ecosystems. One emerging theme is the phenomenon of priority—how differences in arrival times by different species may have profound effects on the long-term trajectories of communities .
Such priority effects were the centerpiece of initial definitions ofassembly theory, and are currently being explored as potential management techniques in ecological restoration, in particular to assist the establishment of less-competitive species in communities . A number of experimental studies on perennial herbaceous plants have shown that a 1- to 3-week priority can significantly affect initial community structure . In other words, initial community structure is contingent on the relative arrival times of species. This includes research in our study system , where we have extended this concept to show that even small initial priority effects of native perennial grasses over exotic annual grasses can multiply over several years to result in substantially greater cover by the natives . Priority effects may be particularly relevant for testing the mechanisms underlying the competitive advantage of invasive annual plants over native perennials. In many western US ecosystems, these invasives have become community dominants . It has been posited that this competitive advantage is driven by the earlier germination and initially higher growth rates of the annuals . Several short-term priority experiments suggest that this is the case . Most of these studies were carried out at a single site and in a single planting year, and we do not know how the strength and consequences of this priority effect differ though space and time. The structure of communities may also be dependent on conditions in the year in which they were established . Ecologists , and restoration practitioners have noted differences in project outcomes and results from experiments initiated in different years, but these have not been subject to controlled experiments where putative drivers of year differences are manipulated. Community structure may be also contingent on site conditions, and the relative abundances of different species may change over relatively small environmental gradients .
It is likely that these differences are due to a combination of site effects, year effects or differences in restoration practices , but these different factors have rarely been examined together in controlled, replicated experiments. Here we report the results of experimental tests of how seeded native perennial grass cover is influenced by competition with exotic annual grasses, the relative timing of seed arrival , rainfall addition and geographical location . We also tested the interactions among priority, rainfall addition and site effects.Over the previous 6 months , we had collected seeds of local provenance at each of the three sites from four native perennial grasses and four exotic annual grasses . For a few of these 24 provenances for which local reproductive populations could not be located, we purchased seeds from local native seed providers. We made some adjustments at the species level to match local sites: for the annual Avena species, we collected and sowed A. fatua in Davis and the very similar A. barbata at McLaughlin and Hopland; for the annual Vulpia species, we collected and sowed V. myuros at Davis and McLaughlin, and the similar species V. bromoides at Hopland. At each site, we established five blocks, each with two replicates of the following four planting treatments: natives sown alone , natives sown together with exotics , natives sown and exotic sown 2 weeks after the next germinating rain and exotics sown alone 2 weeks after the next germinating rain . Planting treatments were implemented in an additive design . In a splitplot design, blocks were divided in half, with one half designated for rainfall manipulation . Each experimental plot was 1.25 m on a side, and each was separated from adjacent plots by 1 m. Prior to planting, all sites were tilled to control weeds, both before and 1–2 weeks after the first germinating rains in the fall. Within 1 week of the second tilling, we did the first sowing . Each plot was lightly raked, sown and then raked again to increase seed–soil contact. There was a second germinating rain on 24 November. Two weeks later, the plots designated to receive a second sowing where sown. Unusually, there had been little rain in the intervening 2 weeks, and there was no rain in the 5 weeks that followed the second sowing. Therefore, to simulate an early season rain that was more similar to a normal year, weed rack the four treatments designated for rainfall manipulation in each block were watered with the equivalent of 1.25 cm of rain immediately after the second sowing pass .
Over the following weeks, plots were weeded of volunteer forbs. Because grasses are difficult to reliably identify at the seedling stage and because there were volunteer seedlings of sown species at two of the three sites , we only weeded the obvious non-sown grass species. Nonetheless, there were significantly greater exotic grass densities in the plots deliberately sown with exotics than in those without . Surveys were carried out after the main winter rain had ceased in the spring, at the time of peak flowering. For the Davis and Hopland sites, this was 26–31 May 2012. The phenology of the grasses was delayed at the higher elevation McLaughlin site, which was surveyed 8 June 2012. The areal cover of each seeded species was visually estimated for each plot. We also recorded the cover of common non-sown exotic grasses.For each of the following analyses, linear mixed-effects models were specified with the lme function from the R software package ‘nlme’ . Block was included in all of the models as a random effect. Where necessary, variance structures were specified using the VarIdent function to address violations of homogeneity of variance . ANOVA tables were generated by calling the anova command from the ‘stats’ package . Due tothe nested nature of the design we tested the effects of each factor with sequential sums of squares.It is not surprising that the success of sown native grasses was greatly reduced when sown together with exotic annual grasses , and that in general, cover by exotic annual grasses and native perennial grasses were strongly negatively correlated . In grassland restoration projects in the Central Valley of California, the presence of exotic annuals represents perhaps the greatest challenge to successfully establishing native perennial grasses, and aggressive pre-sowing control of exotics is now considered a sine qua non for restoration. Conversely, one of the most effective means of preventing the dominance of exotic annuals is the establishment of cover by native perennial grasses . Together, these processes result in strong negative correlations between exotic and native grasses. The magnitude of the competitive suppression of natives by exotics, however, varied across the three sites. Site effects are a complex array of interacting differences, including different means and patterns of rainfall and temperatures, different intensities and identities of weed challenge, and different herbivore pressures. We can only suggest which are the important drivers, but note that in the coolest site , where native grasses achieved little cover in the first year even when planted alone , they were significantly less affected by the sown exotic annual grasses , which also had reduced cover . Although in practice weed control often seeks to greatly reduce the challenge of exotic annuals for at least the first year of native planting, our results show that even a much briefer respite can have a profound effect. When exotic annual grasses were seeded just 2 weeks after germinating rains for the natives, their ability to suppress these natives was greatly reduced . This provides experimental support for the suggestion that one of the ways the exotic annual species outcompete natives in California grasslands is their demonstrated earlier germination and faster growth . The fact that the tE treatment had nearly as much exotic cover as the NE treatment strongly suggests that the late sowing did not itself greatly reduce eventual exotic cover, but that this occurred only in the presence of natives, i.e. as a priority effect. There are also reasons to believe that these differences in community structure arising from initial differences in our experimental treatments have long-term consequences . Vannette and Fukami made several predictions about the strength of priority effects that apply in this system . In particular, they suggested that priority effects would be greater under higher resource availability . In our system, however, watering reduced the strength of priority effects. This was not because of increased resource availability perse, but rather because the watering treatment effectively reduced the duration of the priority treatment. Greater temporal priority usually results in stronger priority effects . This experiment was initiated in a year when there was a 4-week drought following a few weeks of germinating rains in November . Our watering treatment suggests that one of the reasons that the priority effect was so strong in our experiment was this early wet season drought that allowed sown natives to grow for almost a full month before exotics germinated. When this drought was partially alleviated by watering, the strength of the priority effect was significantly reduced . We would predict that in a year with more consistent fall rain, these priority effects would be milder. Indeed, in a very similar experiment carried out in 2008, this was the case .