Habitat fragmentation per se ‐ independent of habitat amount ‐ often increases patch occupancy, possibly because patches are closer together in landscapes with higher fragmentation per se, which should increase dispersal success. Here, we ask whether this effect is influenced by the quality and/or heterogeneity of the landscape matrix, i.e. the non‐habitat portion of the landscape. Specifically, we expect the positive effect of fragmentation per seshould be accentuated when matrix quality is high, reducing dispersal mortality. In contrast, when matrix quality is low, high dispersal mortality should lead to fewer colonisations, and accumulation of extinctions across the smaller patches in a more‐fragmented landscape could lead to negative effects of fragmentation per se. Additionally, matrix heterogeneity could obscure fragmentation effects, as the link between habitat spatial distribution and between‐patch dispersal becomes less predictable. We test these ideas using Glanville fritillary butterfly Melitaea cinxia occupancy data for 4291 habitat patches in the Åland Islands, Finland. Habitat patches for the study species are discrete and well‐defined areas where at least one of the two host species occurs. Adult individuals disperse from habitat patches, spending time in the landscape matrix while searching for new habitat patches. Our predictions were mostly supported. Fragmentation effects were more strongly positive when matrix quality was high; however, we did not see the predicted negative effect of fragmentation per se in landscapes with low matrix quality. As predicted, fragmentation effects on patch occupancy were weaker in landscapes with a more heterogeneous matrix. Our findings may explain why fragmentation effects are often weak. They also suggest that the moderating effects of matrix quality and heterogeneity should be explicitly considered when interpreting effects of habitat fragmentation per se on species distributions.

中文翻译：

---

在具有更高质量和更均匀基质的景观中，碎片化本身对斑块占用率的影响更强、更积极


栖息地破碎化本身 - 与栖息地数量无关 - 通常会增加斑块的占用率，这可能是因为在本身破碎化程度较高的景观中，斑块靠得更近，这应该会增加扩散的成功率。在这里，我们询问这种效果是否受到景观基质（即景观的非栖息地部分）的质量和/或异质性的影响。具体来说，我们预计当基质质量高时，碎裂本身的积极影响应该会得到加强，从而降低扩散死亡率。相反，当基质质量低时，高扩散死亡率应导致较少的定植，而在更加破碎的景观中，在较小斑块上灭绝的积累可能会导致破碎本身的负面影响。此外，基质异质性可能会掩盖碎片效应，因为栖息地空间分布和斑块间扩散之间的联系变得难以预测。我们使用芬兰奥兰群岛 4291 个栖息地斑块的 Glanville 贝母蝴蝶 Melitaea cinxia 占用数据来测试这些想法。研究物种的栖息地斑块是离散且定义明确的区域，其中至少会出现两种宿主物种中的一种。成年个体从栖息地斑块中分散开来，在景观矩阵中花费时间，同时寻找新的栖息地斑块。我们的预测在很大程度上得到了支持。当基质质量较高时，碎裂效应更为积极;然而，在基质质量低的景观中，我们没有看到碎片化本身的预测负面影响。正如预测的那样，在具有更多异质基质的景观中，碎片化对斑块占用率的影响较弱。我们的研究结果可以解释为什么碎片效应通常很弱。 他们还建议，在解释栖息地破碎本身对物种分布的影响时，应明确考虑基质质量和异质性的调节作用。