Abstract Mountain regions are characterized by complex interrelations between human and natural systems. Political and socioeconomic drivers at various scales affect mountain farming systems functioning, resulting in farm structural changes (management, structure and economic performance). Longitudinal studies help fully understand the dynamics of these systems, identify their main drivers of change, and prepare for foreseeable future events. This study aimed to (i) analyse the main changes of cattle farming systems in the Pyrenees from 1990 to 2018, (ii) identify the different trajectories of farm evolution and (iii) to determine the key drivers of those trajectories at global, regional and household levels. We monitored 50 cattle farms in three valleys with different socioeconomic contexts, which were surveyed in 1990, 2004 and 2018. We observed clear changes regarding land and labour production factors. Over the 1990–2004 period, farming systems experienced a land use extensification (one-month increase of grazing season) and capital intensification (55% increase of livestock units (LU) per work unit (WU)) processes, coinciding with a switch from dairy to beef farming with on-farm fattening. Over the 2004–2018 period, land use stabilised but the capital intensification process went on (17% increase of LU/WU) while farms reduced their inputs (43% decrease of feeding costs per LU), in parallel to the decreasing importance of fattening. These changes allowed to globally maintain stable farm economic margins (around 40,000 €/WU). Multivariate statistical analyses enabled to identify four trajectories of evolution, three of them specific to each valley under study and a common across-valleys trajectory. These trajectories resulted from the interaction between global and regional drivers and household particularities. The CAP played a major role at the global level (representing an average of 70% of farm gross margin in 2004 and 2018), while tourism development and household characteristics were the main drivers at the regional level. Several farms responded by maximising their output related to the most limiting production factor (i.e. agriculture land or labour) in each valley. However, the across-valleys trajectory, which comprised 44% of farms, showed limited changes during the studied period. The ability of farms to maintain their adaptation capacity while keeping economic and social viability will determine the future of cattle farming in the region. Our findings highlight the need of reorienting agricultural policies towards promoting new entrants into mountain farming, better integrating CAP instruments with other EU sectorial policies and improving farm monitoring by disaggregating follow-up processes by agroecosystem and management regimes.

中文翻译：

---

山区农业变革的驱动因素：西班牙比利牛斯山脉养牛系统演变轨迹的三十年分析

摘要 山区的特点是人与自然系统之间的相互关系复杂。各种规模的政治和社会经济驱动因素影响山区农业系统的运作，导致农场结构变化（管理、结构和经济绩效）。纵向研究有助于充分了解这些系统的动态，确定其变化的主要驱动因素，并为可预见的未来事件做好准备。本研究旨在 (i) 分析 1990 年至 2018 年比利牛斯山脉养牛系统的主要变化，(ii) 确定农场演变的不同轨迹，以及 (iii) 确定这些轨迹在全球、区域和家庭层面。我们在 1990、2004 和 2018 年对三个不同社会经济背景的山谷中的 50 个养牛场进行了监测。我们观察到土地和劳动力生产要素的明显变化。在 1990 年至 2004 年期间，农业系统经历了土地利用扩大（放牧季节增加一个月）和资本集约化（每个工作单位（WU）的牲畜单位（LU）增加 55%）过程，同时从奶牛养殖与农场育肥。在 2004-2018 年期间，土地利用稳定，但资本集约化过程仍在继续（LU/WU 增加 17%），同时农场减少了投入（每个 LU 的饲养成本降低了 43%），同时育肥的重要性下降. 这些变化允许在全球范围内保持稳定的农场经济利润率（约 40,000 欧元/WU）。多变量统计分析能够识别进化的四个轨迹，其中三个特定于所研究的每个山谷以及共同的跨山谷轨迹。这些轨迹是全球和区域驱动因素与家庭特征之间相互作用的结果。CAP 在全球层面发挥了重要作用（2004 年和 2018 年平均占农场毛利率的 70%），而旅游发展和家庭特征是区域层面的主要驱动力。一些农场的回应是最大限度地提高与每个山谷中最有限的生产要素（即农业用地或劳动力）相关的产量。然而，跨越山谷的轨迹（占农场的 44%）在研究期间显示出有限的变化。农场在保持经济和社会活力的同时保持其适应能力的能力将决定该地区养牛业的未来。