Agriculture represents humankind’s largest engineered ecosystem, and agricultural ecosystems are managed chiefly to meet food, fiber and fuel needs [1] [2] . Although agriculture can contribute to ecosystem services, it can also be a source of disservices, including loss of biodiversity and emissions of greenhouse gases (GHG) and pollutants [3] . However, agricultural management can be designed to host wild biodiversity of many types [4] - [6] , and to reduce GHG emissions [7] - [9] .

Stork-friendly farming (hereinafter referred to as SF farming), which has been practiced in Toyooka City in the prefecture of Hyogo in Japan, is a biodiversity-conscious farming method. It is characterized by the reduced usage of chemical fertilizers and agrichemicals, and is an endeavor to assist with the reintroduction of storks into the wild [10] . Rice produced by the SF farming method has been marketed under the “stork-friendly rice” brand, and has its higher selling price than conventionally grown rice from the same geographical area. In addition to protection of storks and preservation of biodiversity, the stork-friendly farming method will be appreciated even more as a superior environment-friendly farming method, if it also helps mitigate climate change.

In the present study, we attempt to evaluate the SF farming method, which has established a good reputation as an endeavor to help biodiversity conservation, in terms of the impact on global warming. Specifically, the rice production by SF farming method is compared with that by conventional farming method on the basis of carbon dioxide (CO₂) emissions estimated using the life cycle assessment (LCA). LCA can assist to identify more sustainable options [11] .

Although few studies have applied LCA to organic or environment-friendly rice farming in Japan [9] [12] [13] , the impacts on global warming of these farming systems compared to the conventional farming systems are debatable. Hokazono et al. [12] revealed that organic and sustainable rice production systems had the potential to mitigate global warming and eutrophication. Hokazono and Hayashi [13] applied LCA to organic rice farming taking the year-to-year variations of agricultural production into account by using time series data. They concluded that although environmental impacts associated with organic farming fluctuated widely, it was higher than those of conventional rice farming on average. Masuda and Tomioka [9] showed that GHG emissions from the environmentally friendly rice farming were much more than from conventional rice farming by using the LCA method.

Figure 3 shows the estimated CO₂ emissions for each of the farm types. In Figure 3, the emission from energy input denotes the CO₂ emission from the fossil fuel usage in the farm for rice production, and the emissions from material input denotes the CO₂ emission occurring in manufacturing of materials used such as seeds and fertilizers. The findings from the analysis are described below.

First, the type with the greatest CO₂ emissions per unit area was small scale SF farming, low-agrichemical

type (Type C) with the estimated emission of 237 kg CO₂/10a. The conventional farming type was estimated least in the emission (145 kg CO₂/10a). Thus, the CO₂ emission per unit area from SF farming was greater compared with the emission from conventional farming, with the difference of up to 1.6-fold (=237/145).

Second, within the farms of the SF farming type, large scale farms (type B and type D) were found to emit a lower level of CO₂ per unit area than small scale farms (type A and type C) in both agrichemical-free and low-agrichemical types.

Third, within the farms of the SF farming type, agrichemical-free farms (type A and type B) were found to emit a lower level of CO₂ per unit area than low-agrichemical farms (type C and type D) in both large scale and small scale types.

The results show that the efforts for preservation of storks and CO₂ emissions reduction may be in a trade-off relationship, since the CO₂ emissions from all type of SF rice farming are greater compared with the emission from conventional farming. The results also imply that within the farms of the SF farming type, large scale farms and agrichemical-free type farms are considered favorable in terms of CO₂ reduction, comparing with small scale farms and low-agrichemical farms respectively.

Masuda and Tomioka [9] and Hokazono and Hayashi [13] showed that GHG emissions from the environmentally friendly rice farming is higher than from conventional rice farming by using the life cycle assessment method. Masuda and Tomioka [9] pointed out that the organic fertilizer used in environmentally friendly rice farming cased a large increase in methane emissions from paddy field and it increased GHG emissions from environmentally friendly rice farming. Hokazono and Hayashi [13] pointed out that although agrichemical and chemical fertilizers are eliminated in organic rice farming, large quantities of organic fertilizer are required.

In addition to CO₂ emission from energy and material input, CH₄ emission from paddy field needs to be studied in the future. Application of organic materials, which are believed to effectively absorb greenhouse gases by acting as a carbon reservoir, should also be included in the future analysis.

We evaluate the biodiversity-conscious farming method in terms of the impact on global warming by using the life cycle assessment (LCA) taking stork-friendly farming in Japan as a case of farming method. The results show that efforts for preservation of storks and countermeasures against global warming may be in a trade-off relationship. The results suggest that expansion of the farming scale and switch from low-agrichemical to agrichemical-free farming may be two possible paths towards a lower CO₂ emission from stork-friendly farming than the current level.

SF farming is also considered to have several positive effects such as the biodiversity local brand development for biodiversity friendly consumer market, eco-tourism and environmental education [10] [19] . Evaluating these effects needs to be focused in future research.

This research was supported in part by JSPS KAKENHI Grant Number 25660175. The research benefited from expert discussions at the New Zealand Life Cycle Assessment Conference 2014, on 2-3 September 2014, at the Comfort/Quality Inn, Wellington, New Zealand. The views expressed in this study are not shared by the agencies that the authors are affiliated with or from which they received funding. The points expressed reflect the views of the authors.