If breeders can increase their imported genetic materials by 10 percent, the marginal cost of wheat variety production will fall by 2.2 percent .Such an effect, a type of spillin, has long been known to play an important role in the effectiveness of spending on agricultural research . Our study demonstrates that spillins are also an important source of efficiency gains at the level of the crop breeding institute, and policies and institutions that facilitate the free flow of germplasm will raise the productivity of the agricultural research system. Compared with increasing an institute’s human capital and access to genetic material, the effects of having scientists from other disciplines and the burden of caring for retirees are less clear. Having scientists from other disciplines in a breeding program marginally reduces wheat-breeding costs. It has the opposite effect in maize institutes, although the effect disappears in estimations that correct for measurement error. Hence, at the very least, it seems that the addition of soil scientists, plant pathologists and other scientists does not significantly detract from productivity, even in the types of breeding centers that dominate China’s research system. Our findings also do not provide evidence that would validate the complaints of scientists and research administrators about the adverse effects of bearing the burden of the welfare of retirees. While this result is surprising , it could be that there are two offsetting effects of having breeders remaining formally attached to the institute after they retire. Although retirees probably do take away resources that could otherwise be used for research,round pots their presence could be an asset since they have experience, breeding material inventory, and contacts in the seed system that could help reduce costs.
The scientists responsible for breeding new varieties today will have to meet even greater challenges than those that gave rise to the Green Revolution of the 1970s and 1980s. In an era of waning support and increased demands for output, it will be necessary to become increasingly efficient at producing new varieties. However, there is almost no empirically-based evidence to guide the efforts of reorganizing the current agricultural research system. In this study, we attempt to identify sources of efficiency in China’s crop breeding system. Using a panel data of 46 wheat and maize breeding institutes from 1981 to 2000, we examine the factors that affect the variable costs of wheat and maize varieties. Using a number of approaches and accounting for a number of econometric issues, our analysis produces a set of robust results that can help guide reformers in their efforts to increase the efficiency of China’s crop breeding system. Our most striking finding, and one that is relevant for crop breeding centers around the world, is the existence of strong economies of scale in China’s crop breeding research. The coefficients of economies of scale imply a significant cost saving associated with expanding the scale of crop breeding institutes. According to our findings, the current large number of small crop breeding institutes is the main source of inefficiency. In addition, a number of other sources of inefficiency are identified.
Though not as strong or consistent as the results for economies of scale, we find there are economies of scope in the production of varieties of different crops. Merging a wheat-only institute with a maize-only institute can lead to small, but significant cost savings. We also find that raising the human capital of the breeding staff and facilitating the access of breeders to wider sources of germplasm increases the efficiency of breeding. All of these results fit squarely with our expectations based on knowledge of the crop breeding system in China as well as from a consideration of the counterpart institutions in other countries and in international agricultural research centers. Taken at face value, our findings can support a blueprint for the reform of crop breeding in developing countries, from a system dominated by a multitude of small, fragmented, and isolated breeding stations to one characterized by a smaller number of “super” breeding centers. New centers would be larger, broader in scope, and be staffed by well-trained scientists representing a number of different agricultural science disciplines. Expanding the size of the institutes, either by merging two or more or by expanding a single institute and shutting down others, would take advantage of the strong economies of scale. Our results do not give exact guidance on how big the institutes should be, in part because we are not observing many institutes that have reached or passed the bottom of the average cost curve. However, even casual observation of the descriptive data shows that crop breeding institutes can be expanded by at least several times their current size. Such a move would start to shift the size of breeding programs in China more towards those of developed nations. The new centers could also take advantage of other sources of efficiency gains.
The positive economies of scope mean that the new super centers should have at least two departments, one for wheat and one for maize. It also can be argued that new departments should be created in the centers for the support of work by scientists from other disciplines. Although we did not always find strong efficiency gains from the addition of other scientists, there was even less evidence of any diseconomies associated with institutes that contained non-breeders.13 But, in anticipation of future changes in the technology of crop breeding, that surely will confront any modern agricultural research system , it is likely that there will be substantial gains to having an institutional structure in place that can take advantage of and develop its own high technology products. However, a number of factors potentially could undermine part or even all of these efficiencies, should the government implement an approach based on merging and expanding smaller crop-breeding institutes into a smaller number of super breeding centers. First, there will be non-pecuniary costs associated with mergers or expansions. For instance, researchers who are likely to lose their jobs and directors who are likely to lose their political positions will do whatever they can to prevent any ambitious reorganization from happening. The more ambitious the reform is, the greater will be the opposition. Second, merging or cutting will encounter transaction costs associated with the reform process itself and with reorganizing operations of merged or expanded institutes. Finally, a smaller number of super stations could mean less competition,garden pots plastic leaving less incentive for innovative research. Hence, in deciding how to implement a reorganization of the crop breeding research system, research sector leaders should also take into account these adverse factors and potential transactions costs. Agriculture has increased to meet the demand of a growing and wealthier population that demands more, and more resource intensive, calories . The doubling of agricultural production in the past 40 y has been fueled by technological improvements as well as higher levels of pesticide and fertilizer inputs . Although this increase in food production has contributed to vast improvements in nutrition and reductions in hunger worldwide , the ecological and environmental consequences of these inputs are straining the long-term viability of agricultural systems and the human and natural communities that surround agricultural production . Agricultural intensification at both local and landscape scales has been the workhorse behind production increases. Farms have become specialized on fewer, high-yielding crops grown in shorter rotation cycles on larger fields . In aggregate, agricultural landscapes have become more simplified with less noncrop habitat and fewer crop types in production . Aggregate food production from intensification has undoubtedly increased, reducing the pressure of agricultural land expansion into natural habitats to meet the growing food demand . However, there are numerous unintended consequences of agricultural intensification for biodiversity , water quality, and other ecosystem services . Modern agricultural systems rely on agrochemicals to reduce pest damage, thereby minimizing crop loss . However, many of these chemicals have adverse environmental and ecological effects. Pesticides, broadly, and insecticides, in particular, have been linked to biodiversity declines in numerous taxa in both temperate and tropical regions , as well as declines in water and air quality. Further, off-site pesticide contamination and pesticide resistance are important externalities of pesticide use that have consequences for both chronic and infectious human diseases.
Pesticide use is fundamentally about controlling pest damage. Crops can vary substantially in average insecticide use based on value or susceptibility to pest damage. However, given the set of crops in production, ecologists are seeking means to reduce excess insecticide use by manipulating on-farm and landscape characteristics. Because insect pests and natural enemies often have large dispersal ranges and varied habitat needs, the focus has been on if and when complex landscapes reduce pest abundance or, conversely, if and when simplified landscapes lead to more pest problems . However, ecological field studies seeking to inform more sustainable pest control practices face an enormous challenge. Pest community composition and pest damage may be intricately linked to landscape composition, habitat configuration, and the focal crop type in ways and at spatial scales that are difficult to address in field experiments. As a result, the evidence tying simplified habitats to insecticide use is often specific to one crop and pest combination and is equivocal overall . Data-driven approaches have proven useful in elucidating the larger scale patterns in the relationship between landscape-level agricultural intensification and insecticide use . However, these studies have been limited in spatial resolution of both crop and insecticide data . Thus, the majority of research has focused on one aspect of landscape simplification, namely, cropland extent measured as the proportion of county in cropland. In highly simplified agricultural regions that are dominated by one or a couple of crops, county-level cropland may serve as an appropriate metric for intensification. However, in highly diverse agricultural regions, landscape-level crop diversity, in addition to cropland extent, may be an important driver of pests and enemies . Further, cropland extent may act on both local and landscape scales. Disentangling suchcomplexities requires refined data on crops and insecticides at We conducted all analyses at the field scale. There are ∼13,000 large spatial scales, information that is currently absent for much fields active per year, with crop diversity, cropland extent, field of the world. size, and insecticides varying regionally . Crop diversity Here, we take advantage of unique field-level crop and pesticide was calculated as the Simpson’s Diversity Index of all data for ∼100,000 field-year observations in the agriculturally diverse crops within a 2,500-m radius of the focal field Kern County, California, from 2005 to 2013 to understand if crop and did not include the focal field. SDI was calculated at difdiversity , field size, and cropland extent drive insecticide ferent taxonomic levels or commodity use. Kern County is situated in the southern San Joaquin Valley and levels . Unless otherwise noted, is California’s second ranked county by agricultural production the results discussed below are from models based on SDI at value, with an annual agricultural output of ∼6 billion dollars . the species level. To facilitate comparison, all covariates were Although the leading crops by value are grape varieties and almonds, standardized. over 200 different commodities are produced . We first conduct For models including all crops , we found the analysis pooling all crops to understand general patterns in in- the largest effect of crop diversity on insecticide use in secticide use and landscape simplification using panel data analyses the pooled ordinary least squares model, followed by the that control for regional differences in insecticide use as well as for fixed effects model with region and year controls, and then by the year shocks in pest control. Because crops are not planted haphaz- fixed effects model with crop and year controls . Throughout, we are following the econometric use of the or “fixed effects”) to parse the effect of crop diversity from the term fixed effects to describe panel data models with dummy or differences in crop composition that may be inherent to landscapes with high or low levels of crop diversity. Using these models, we indicator variables for each crop, year, or region.