以中型生態系統探討不同水深對農塘甲烷循環微生物群落與甲烷通量之影響

Abstract

池塘為重要的內陸水體甲烷（CH₄）排放源，但水深如何透過沉積物氧化還原狀態、孔隙水地質化學與甲烷循環微生物群落共同調控甲烷排放，仍有待釐清。本研究以臺灣北部關渡地區之淡水農塘為對象，結合溫室中型生態系統實驗與現地農塘觀測，探討不同水深對農塘沉積物甲烷循環微生物群落與甲烷通量之影響。中型生態系統設置 50、100 及 150 cm 三種水深處理，每組三重複，並整合沉積物固相性質、表面水與孔隙水化學、甲烷通量、16S rRNA 細菌群落，以及產甲烷功能基因 mcrA 與甲烷氧化功能基因 pmoA 之定量與群落分析。 結果顯示，水深處理可改變沉積物表層氧化還原環境，深水處理整體呈現較低之氧化還原電位，顯示其底床環境較具還原性。然而，三種水深處理之甲烷通量並未呈現顯著差異，50、100 及 150 cm 處理之通量中位數分別為 2.33、0.98 及 1.32 µmol·m⁻²·h⁻¹，且未隨水深呈單調變化。相較之下，產甲烷功能基因相對豐度於 150 cm 處理較高，孔隙水溶解態甲烷亦大致呈深水略高之趨勢，顯示深水處理可能具有較高的底床產甲烷潛勢。此一「底床產生潛勢較高，但表面通量未同步升高」的現象，指出農塘甲烷排放並非由沉積物產甲烷潛勢單獨決定，而可能受到水柱傳輸距離、滯留時間與甲烷氧化作用共同調節。 孔隙水與微生物資料進一步顯示，硫酸鹽還原作用是本研究中最一致的地質化學調控主線。孔隙水 SO₄²⁻ 與溶解態 CH₄ 呈顯著負相關，SO₄²⁻ 亦與 mcrA/16S rRNA 及 pmoA/16S rRNA （基因拷貝數比值）呈負相關；同時，硫酸鹽還原菌相對豐度約為產甲烷菌之數倍，顯示硫酸鹽還原菌在沉積物厭氧代謝中具有明顯數量優勢。上述結果共同指向硫酸鹽還原作用可能透過基質競爭與維持較高氧化還原條件，抑制產甲烷作用與甲烷累積。微生物群落分析亦顯示，mesocosm 三水深處理間之 mcrA 與 pmoA 群落組成未形成明顯分群，表示水深在本研究中主要改變甲烷循環之背景條件與功能基因量級，而非造成明確的功能菌群組成分化。 此外，連續監測結果顯示，氣泡逸散雖僅於少數量測中發生，但一旦出現，即主導該時段甲烷釋放，反映農塘甲烷排放具有高度事件性與時空異質性。現地農塘結果則顯示，真實系統中水深常與岸際位置、有機質輸入、底泥條件、擾動與氣泡逸散頻率等因子共變，因此中型生態系統所得之水深純效應不宜直接外推為自然農塘之水深效應。整體而言，本研究指出，水深本身並非農塘甲烷排放的單一主控因子；農塘甲烷排放應視為產甲烷作用、硫酸鹽還原作用、甲烷氧化、氣體傳輸與氣泡逸散共同作用下的結果。未來農塘甲烷減排管理不宜僅依賴水位調控，而應同時考量有機質負荷、底泥累積、氧化還原條件與高頻通量監測。

Ponds are important sources of methane (CH₄) emissions from inland waters, yet how water depth regulates CH₄ emissions through sediment redox conditions, porewater geochemistry, and methane-cycling microbial communities remains unclear. This study investigated the effects of water depth on methane-cycling microbial communities and CH₄ fluxes in agricultural pond sediments by combining a greenhouse mesocosm experiment with field observations in a freshwater agricultural pond in Guandu, northern Taiwan. Three water-depth treatments, 50, 100, and 150 cm, were established in triplicate mesocosms. Sediment solid-phase properties, surface water and porewater chemistry, CH₄ fluxes, 16S rRNA bacterial communities, and the abundance and community composition of the methanogenesis marker gene mcrA and methane oxidation marker gene pmoA were analyzed. The results showed that water depth altered the redox environment at the sediment surface. Deeper-water treatments generally exhibited lower oxidation-reduction potential, indicating more reducing conditions in the sediment bed. However, net CH₄ fluxes did not differ significantly among the three water-depth treatments. The median CH₄ fluxes under the 50, 100, and 150 cm treatments were 2.33, 0.98, and 1.32 µmol·m⁻²·h⁻¹, respectively, and did not show a monotonic trend with increasing water depth. In contrast, the relative abundance of the methanogenesis marker gene was higher in the 150 cm treatment, and porewater dissolved CH₄ generally showed a slightly higher tendency in deeper-water treatments. These results suggest that deeper water may enhance sediment methanogenic potential. The observed decoupling between higher sediment methanogenic potential and unchanged surface CH₄ flux indicates that CH₄ emissions from agricultural ponds are not controlled solely by sediment methanogenic potential, but may also be regulated by water-column transport distance, residence time, and methane oxidation. Porewater and microbial data further indicated that sulfate reduction was the most consistent geochemical control in this study. Porewater SO₄²⁻ was significantly negatively correlated with dissolved CH₄ and was also negatively correlated with the mcrA/16S rRNA and pmoA/16S rRNA gene copy number ratios. In addition, the relative abundance of sulfate-reducing bacteria was several times higher than that of methanogenic archaea, suggesting that sulfate-reducing bacteria had a clear numerical advantage in sediment anaerobic metabolism. Together, these results suggest that sulfate reduction may suppress methanogenesis and CH₄ accumulation through substrate competition and the maintenance of relatively higher redox conditions. Microbial community analysis also showed that mcrA and pmoA community compositions did not form distinct clusters among the three mesocosm water-depth treatments, indicating that water depth mainly altered the background conditions and functional gene abundance of methane cycling rather than causing clear differentiation in functional microbial community composition. Continuous monitoring further showed that ebullition occurred only in a small number of measurements, but once it occurred, it could dominate CH₄ release during that period. This finding highlights the strong event-driven nature and spatiotemporal heterogeneity of CH₄ emissions from agricultural ponds. Field observations further indicated that, in real pond systems, water depth often covaries with distance from the shoreline, organic matter inputs, sediment conditions, disturbance, and ebullition frequency. Therefore, the pure water-depth effect observed in mesocosms should not be directly extrapolated as the water-depth effect in natural agricultural ponds. Overall, this study demonstrates that water depth itself is not the sole controlling factor of CH₄ emissions from agricultural ponds. Instead, CH₄ emissions should be understood as the combined outcome of methanogenesis, sulfate reduction, methane oxidation, gas transport, and ebullition. Future CH₄ mitigation strategies for agricultural ponds should not rely solely on water-level regulation, but should also consider organic matter loading, sediment accumulation, redox conditions, and high-frequency flux monitoring.