Climate change is shifting the growing seasons of plants, affecting species performance and biogeochemical cycles. Yet how the timing of autumn leaf senescence in Northern Hemisphere forests will change remains uncertain. Using satellite, ground, carbon flux, and experimental data, we show that early-season and late-season warming have opposite effects on leaf senescence, with a reversal occurring after the year’s longest day (the summer solstice). Across 84% of the northern forest area, increased temperature and vegetation activity before the solstice led to an earlier senescence onset of, on average, 1.9 ± 0.1 days per °C, whereas warmer post-solstice temperatures extended senescence duration by 2.6 ± 0.1 days per °C. The current trajectories toward an earlier onset and slowed progression of senescence affect Northern Hemisphere–wide trends in growing-season length and forest productivity.

Habitat fragmentation, in which contiguous forests are broken into smaller, isolated patches, threatens biodiversity by disrupting species movement, shrinking populations, and altering ecosystem dynamics. Past assessments suggested declining global fragmentation, but they relied on structure-based metrics that overlook ecological connectivity. We analyzed global forest fragmentation from 2000 to 2020 using complementary metrics that captured patch connectivity, aggregation, and structure. Connectivity-based metrics revealed that 51 to 67% of forests globally—and 58 to 80% of tropical forests—became more fragmented, which is nearly twice the rate suggested by traditional structure-focused methods (30 to 35%). Aggregation-focused metrics confirmed increases in 57 to 83% of forests. Human activities such as agriculture and logging drive this change. Yet protected tropical areas saw up to an 82% reduction in fragmentation, underscoring the potential of targeted conservation.

Deforestation has exacerbated the fragmentation of habitats into smaller, more isolated patches, driving global declines in biodiversity. Yet, a comprehensive global perspective of the trends in forest fragmentation, and its key drivers in relation to forest cover change remains elusive. To provide a comprehensive global overview of recent changes in forest fragmentation, we compare multiple fragmentation metrics, including those that are sensitive to forest cover and those that are not. Our analysis reveals that, according to cover-sensitive metrics that reflect the ecological implications of forest fragmentation, 52% of the world's forests have become more fragmented over the last two decades, a trend that is primarily attributed to increased deforestation in tropical zones. This value is twice as high than estimates from previous research, which estimated that forest fragmentation is declining across 75% of the global forest area. This discrepancy arises from a mathematical artifact, as previous cover-insensitive metrics equate declines in forest cover with decreased fragmentation. By adjusting for this and focusing on metrics that capture the ecologically relevant aspects of forest fragmentation, our study highlights a worrying trend: the ecological integrity of the global forest system has been significantly deteriorating in recent decades. This underscores the importance of using appropriate metrics to accurately assess the ecological impacts of forest fragmentation, especially in the context of global environmental change.

Accurate projections of temperate tree growing seasons under climate change require representing developmental constraints that determine tree resource allocation. Recent work has identified a phenological “switch point” after the summer solstice (21 June), with pre-solstice warming advancing autumn phenology and post-solstice warming delaying it. Here, we propose this switch is flexible and occurs at the compensatory point between the antagonistic effects of early-season development and late-season temperature. We performed trans-solstice climate manipulation experiments on potted European beech ( Fagus sylvatica ) saplings to test (i) how spring leaf-out timing and June-August temperatures influence end-of-season timing (bud set and leaf senescence [50% loss of leaf chlorophyll content]), and (ii) whether daytime and nighttime temperatures before and after the solstice have different effects, given that trees primarily grow at night. Bud set and leaf senescence responses were tightly coupled ( R ² = 0.49), with bud responses being generally stronger. Each day delay in spring leaf-out delayed bud set by 0.24 ± 0.06 days and senescence by 0.22 ± 0.08 days. Post-solstice full-day cooling in July delayed autumn phenology in late-leafing individuals (bud set +4.9 ± 2.6 days; senescence +3.1 ± 2.8 days) but had negligible impact on early-leafing trees (bud set +1.4 ± 2.6 days; leaf senescence +2.2 ± 2.8 days). Conversely, August full-day cooling advanced both stages. Daytime cooling before the solstice had no effect, while after the solstice it advanced autumn phenology. Nighttime cooling always delayed bud set. These findings support the Solstice-as-Phenology-Switch model and highlight the central role of developmental progression in constraining growing seasons. Faster early-season development –especially under nighttime warming– moves trees past the switch earlier, increasing sensitivity to late-season cooling and thereby triggering earlier autumn phenology. To improve growing season length projections, phenology models should account for these developmentally-mediated and diel-specific temperature responses.

Deforestation leads to habitat fragmentation, which adversely affects global biodiversity. Although some studies, using a separation-focused definition, have reported a decrease in fragmentation across 75% of the world's forests over recent decades, a comprehensive and ecologically relevant understanding of global fragmentation patterns remains lacking. In this study, we analyzed global fragmentation trends from 2000 to 2020, employing metrics that emphasize connectivity, aggregation, or separation. Connectivity-focused metrics reveal that 51% of global forests, particularly in tropical regions (58%), have undergone increased fragmentation—a rate nearly double that suggested by previous separation-focused metrics. This increase is corroborated by aggregation-focused metrics, which indicate heightened fragmentation in approximately 58% of forests worldwide and across all biomes. Further analysis attributes this escalation primarily to human activities, such as shifting agriculture and logging. Importantly, tropical protected areas have exhibited reductions in fragmentation by up to 82% compared to non-protected areas, underscoring the success of conservation efforts in these regions.

Climate warming alters the start (SOS) and end (EOS) of growing seasons, impacting biotic interactions and biogeochemical cycles. However, the global constraints between these two stages – how SOS influences EOS (SOS-EOS effect) and vice-versa (EOS-SOS effect) – remain poorly understood, hindering future growing-season projections. Using MODIS satellite-derived phenology data for deciduous vegetation and European ground observations for deciduous tree species, we show that earlier SOS typically advances EOS (on average by 0.19 ± 0.001 days per day [MODIS] and 0.12 ± 0.002 days per day [ground]), while EOS exerts a weaker influence on subsequent SOS (-0.05 days per day). The SOS-EOS effect often outweighed abiotic factors, with SOS being the top predictor of EOS in 34% of pixels (β = 0.27), while preseason temperature was the primary predictor of SOS in 58% (β = -0.33). More importantly, we identified a dampening interaction, where an increase in one carry-over effect reduced the other, with distinct and opposing geographic patterns: the SOS-EOS effect was twice as strong as the EOS-SOS effect in temperate deciduous forests, while the EOS-SOS effect was up to three times stronger in boreal taiga and tundra. Mechanistically, these patterns are likely to reflect developmental (cell and tissue growth) and stress-related constraints (SOS-EOS effect) and chilling requirements during dormancy (EOS-SOS effect at high latitudes). These findings highlight how plant-internal physiological feedbacks constrain phenological responses to climate change, emphasizing the need to integrate carry-over effects into future ecosystem models.