Therefore, our results agree with the existing literature, which suggests that old expenditures impact current productivity positively, and their exclusion tells us only a partial story. The coefficients we have obtained in this study indicate that there is room for improvement in extension research and outreach, and that introduction of new research-based knowledge and technology improves productivity. Results also suggest that primary-occupation farmers may be less efficient than those who are able to maintain more than one profession. Efforts could be focused towards improving any existing gaps in efficiency among farmers in different counties.The results of our analysis can guide policymakers during a period of political pressure to reduce public spending for agricultural extension in the state. The county fixed effects results allow a more targeted policy intervention on higher and lower performing regions .By controlling for individual county and fixed-year effects that may be driving productivity in that county, we find that some of the major agricultural counties in California record high positive impacts of UCCE expenditures stock. Out of the 50 county offices in our study, we observe that UCCE expenditures stock has a significant impact on 21 counties for all values of knowledge depreciation. We observe a statistically significant negative impact on a few counties, such as Amador, Calaveras, Humboldt-Del Norte, Modoc, and Siskiyou. For two counties, the impact is not statistically different from 0. In terms of policy, these coefficients can be used as reference points for allocating budgets to different counties. Extension efforts could be targeted to the counties with inconclusive or negative impacts. Monetary impact of cutbacks on county productivity could also be calculated, using the estimates of extension expenditures in this paper. The analysis driven by county performance helps design policies with heterogenous focus, which has been more relevant when public funds have to be allocated among heterogeneous performing recipients of these funds. And finally, as shown in Section 5.3 extension introduces substitutability of traditional inputs with extension knowledge so that higher expenditure on extension in some of the lower-performing counties can substitute for other traditional inputs, dutch buckets for sale which may be scarce in supply.
In particular, our analysis highlighted and measured substitution of extension knowledge for labor and chemicals.Both biodiversity and the human activities that threaten it are unevenly distributed around the globe. Thus, evaluating whether they are spatially congruent and choosing the best areas for conservation actions given the distribution of these conflicts are central problems in conservation bio-geography . The magnitude of the current biodiversity crisis, coupled with the limited resources available for protecting biodiversity, implies that prioritization is unavoidable. Spatial prioritization seeks to identify the areas that are likely to yield the best benefits for biodiversity given a particular conservation investment. It may be applied at a variety of scales, including global , regional , national and sub-national levels. Spatial conservation prioritization analyses can be based solely on the distribution of the biological features to be protected . Alternatively, prioritization analyses can include socioeconomic variables that represent threats to biodiversity or opportunities for conservation, such as human population density, land cost and land use . Agriculture is the human activity that represents the main threat to the environment . It constitutes the largest land use on the planet, using 38% of Earth’s ice-free land surface and 70% of global human freshwater uptake. Food production accounts for 19% of Earth’s net primary productivity and 30-35% of global greenhouse gases, with direct impacts on biodiversity . The burden on the environment may be higher in the future as the human population is expected to increase to more than 10 billion by 2050 . Moreover, a billion people are currently chronically malnourished as a result of lack of access to food . Given the value of biodiversity for human well-being , understanding the potential impacts of future agricultural expansion on biodiversity is a key issue for humanity. The general aim of my PhD thesis was to evaluate the potential impact of agricultural expansion on biodiversity conservation during the 21st century. Specifically, I evaluated four interrelated issues: conservation conflict between agricultural expansion and the global biodiversity conservation priorities and the Brazilian system of protected areas ; the effect of incorporating agricultural expansion data into spatial prioritization models for the conservation of world carnivores ; and the benefits of a globalized conservation strategy for food production and for biodiversity conservation . The impact of future socioeconomic development pathways, including land-use trends, on biodiversity can be accessed by means of quantitative scenarios . For all analyses presented here, I obtained future scenarios of agricultural expansion from land cover maps produced by the Integrated Model to Assess the Global Environment .
IMAGE forecasts, at a resolution of 0.5° × 0.5°, the number of years that each area will be cultivated during the 21st century for six socioeconomic scenarios . For chapter IV, I also included an estimation of potential agricultural productivity in each grid cell, based on climate, relief, soil constraints and irrigation impact . For the first chapter, I overlaid the spatial polygons of the Global Biodiversity Conservation Priorities onto a grid with a spatial resolution of 0.5° × 0.5°. I tested whether areas defined by their higher vulnerability were more affected by agriculture in the year 2000. The opposite was expected for areas with low vulnerability . I also tested whether these priority areas would be more affected by agricultural expansion during the 21st century than expected by chance . To address the aims of chapter II, I overlaid the IMAGE’s land-use model with Brazilian protected areas to calculate the conflict between these two land uses. I obtained Brazilian protected areas’ polygons from the World Database of Protected Area . I also included 10 km buffers around each protected-area polygon to represent the legal buffer zone usually used in Brazil, which is an area where human activity is restricted. I then tested whether these areas were more affected in the present and in the future than expected by chance. Additionally, I tested whether there was difference between the integral protection protected areas and sustainable use protected areas . In both chapters I and II, I evaluated the probability of such conflicts to be found by chance using spatial randomization tests developed in R , considering 1000 iterations . To meet the objectives of chapters III and IV, I performed global spatial conservation prioritization using Zonation . Zonation’s algorithm provides a nested hierarchical ranking of the sites, maximizing the representation of species’ distributions. To define the ranking of importance of sites for conservation, Zonation analyses can also incorporate costs such as potential agricultural production. For all prioritization analyses, I defined the target proportion of areas to be protected as 17%, following the Convention on Biological Diversity , which proposed this percentage as the goal to be met by 2020. I obtained information about mammal species’ distributions from the International Union for Conservation of Nature’s Red List of Endangered Species. I overlaid the spatial polygons onto a grid with a spatial resolution of 0.5° × 0.5°. For chapter III, I focused on 245 terrestrial carnivore species. In chapter IV, I used 5216 terrestrial mammals. These taxonomic groups have been the focus of many conservation programs and they are often considered to represent a potential surrogate for other taxonomic groups . To test whether there is a spatial conflict between the global carnivore conservation solutions obtained in chapter III and the agricultural expansion, I performed spatial correlation analyses using the Spatial Analysis in Macroecology software .
The objectives of chapter IV were achieved by defining global conservation priorities considering three levels of political integration: individual countries, regions , and globalized . I also evaluated the effect of considering, or not, agricultural costs for spatial conservation prioritization. The different conservation solutions were evaluated in terms of the relative amount of food production lost by setting aside sites for conservation and the representation of the geographic distribution of species within those sites. I also evaluated whether the most underdeveloped countries would be subject to higher losses in food production under the global strategy. For this, I correlated the percentage of food production and area lost to sparing land for biodiversity conservation with three development indicators: the Human Development Index , the per-capita gross domestic product , and the percentage of GDP added by agriculture . I found that reactive global biodiversity priorities had about 49% of their area impacted by agriculture in the year 2000 . Conversely, proactive schemes had a low intersection with the agricultural distribution . By the end of the 21st century, there will be an overall increase in world agricultural area from 26.5% of the analyzed area in 2000 to 34.6% in 2100, according to IMAGE, and the difference between the proactive and reactive schemes is predicted to hold true. However, High Biodiversity Wilderness Areas, a proactive scheme,hydroponic net pots is predicted to suffer agricultural impact similar to the reactive schemes, with 73.5% of its area affected, if the worst-case scenarios are realized . In Brazil, a megadiverse country in which agribusiness is the pillar of economy, agricultural expansion is a major conservation concern . According to IMAGE, agricultural land use represented 22% of Brazilian land coverage in 2000 and is predicted to increase up to 40% by 2100, according to a business-as-usual scenario. Moreover, the percentage of protected areas affected is predicted to increase from 11% to 30%, with no difference between IPPAs and SUPAs . I found spatial conflicts between the best areas for terrestrial carnivore conservation and agricultural expansion in the 21st century . These conflicts were alleviated when I incorporated agricultural expansion information into the spatial prioritization process . Nevertheless, accounting for agricultural expansion resulted in a lower representation of species’ geographical ranges: the average proportion of represented ranges was reduced from 58% to 32%. This reduction affected mainly those species with small geographic distributions. In addition, the best solution for global carnivore conservation changed from a spatial distribution closer to that of the reactive global conservation priority schemes to one more like proactive ones. Looking at the impact of globalization for conservation and food production, I found that combining the use of agricultural expansion data and integrating countries in a globalized conservation blueprint to meet the 17% target for terrestrial protected areas, resulted in a 78% reduction in the costs of food production . Furthermore, this globalized conservation approach represented an increase of 30% in the representation of the species in the protected areas network.
The regional-scale conservation solution resulted in similar losses in food production, compared to the globalized solution, and an increase of 17.5% in terms of representation of mammals’ geographical ranges .Conservation actions in the different areas of the world should be planned according to the expected agricultural expansion in the 21st century. Some areas can hold mega-reserves , while other areas should focus on the development of wildlife-friendly agricultural practices. Within Brazil, my findings suggest that the risk of agricultural expansion should be included in the management of protected areas and associated buffer zones. Globally, conservation actions for carnivores should consider agricultural expansion because this may significantly influence the distribution of areas where conservation actions could be more effective in the future . The regional scale may represent an intermediate step towards the global integration. Economic agreements may evolve to common conservation policies, since this has already been done in the European Union by means of the Natura 2000 network . By comparing differences in the distribution of protected areas among countries in the different scenarios, I found that the poorest countries will not be negatively affected by participating in this globalized conservation blueprint. However, the particular cases in which poor countries would be impaired in their development process should be a focus of compensatory policies in order to guarantee the participation of these countries within the global approach. Moreover, such compensatory policies may help to overcome socioeconomic problems such as poverty and inequality, which are known to be detrimental to the success of conservation actions . Feeding an increasing human population, with rising per-capita consumption, while managing the environmental impacts of agriculture, is one of the greatest challenges for global policy. In my thesis, I demonstrated that agricultural expansion will continue to represent an important threat to biodiversity throughout the 21st century. Reducing food waste, increasing agricultural resource efficiency, closing yield gaps, and fostering organic agriculture are tools available for solving this challenge .