Our study uses data from Holocene core samples and modern death assemblages to understand how human-induced environmental change in the northern Adriatic Sea (Italy) may have affected parasite-host dynamics in the economically important bivalve Chamelea gallina. Thirty-one radiocarbon dates confirm temporal distinctness between the periods before and after the onset of significant human influence and confirm that trematode prevalence has decreased by an order of magnitude over the past ~2 k.y. The median number of parasite-induced pits per bivalve host and parasite aggregation has also decreased significantly, signaling a substantial decrease in the effective population size of digenean trematodes. Gaussian finite mixture modeling of pit size does not support the hypothesis of parasite extinction. Combined, these results indicate the (potentially ongoing) collapse of parasite-host interactions in C. gallina in concert with human influence on the Adriatic and its transition to an “urban sea.”Parasites play an important role in ecosystems and can have a significant impact on ecosystem stability (Hudson et al., 2006). Therefore, changes in the prevalence, distribution, or diversity of parasites could signal disruptions in ecological functions due to habitat degradation, pollution, or climate change (Wood et al., 2023). However, ecological monitoring, being restricted in time, suffers from the inability to discriminate short-term fluctuations from long-term regime shifts (i.e., large, persistent changes in ecosystem structure). In this respect, long-term geological data can provide key contexts to current population status and trends. Given their substantial ecological influence, it is useful to have similar long-term data for parasites. Although small, soft-bodied parasites have a low preservation potential, some parasite-host interactions are recorded on the skeleton of their hosts (Huntley et al., 2021), enabling a temporal scale of observation well beyond that of even the longest-term ecological studies (De Baets et al., 2021a). Our goal is to leverage this under-exploited fossil record to reconstruct recent parasite-host dynamics and to place those dynamics in a broader historical context of increasing human influence on ecosystems (e.g., Huntley and Scarponi, 2021).Trematode flatworms (Platyhelminthes) are complex life cycle parasites (i.e., with multiple hosts) that induce a characteristic, preservable growth response (pitting) on their bivalve host’s shell interior (Ruiz and Lindberg, 1989; Huntley et al., 2021). We use the Holocene sedimentary record to assess parasitic infestation in Chamelea gallina—an ecologically and economically important bivalve—before and after the onset of significant human influence in shoreface ecosystems of the Adriatic Sea, one of the most heavily altered marine regions globally (Lotze et al., 2011). Specifically, by integrating radiocarbon dating with analyses of parasite intensity (the number of pits per host) and size distributions of trematode-induced pits, we expand the temporal and geographical scope of neontological studies supporting claims of anthropogenic declines in parasitism in marine environments (Wood et al., 2023). We hypothesize a decline in parasite-host interactions during the past 2 k.y., which was characterized by decreased parasite prevalence and intensity (accounting for differences in depositional environment, time averaging, and host body size) and the extirpation of parasite taxa. In this work, we reveal, from sedimentary successions in Italy, a case for the collapse of parasitic infection in C. gallina that occurred during the transition of the Adriatic into an “urban sea” (coastal estuaries and marginal seas with drainage basins that are extensively developed and populated by humans). While a collapse in parasitism may seem positive, it signals a simplification of the trophic structure and a loss of crucial ecological functions along with the parasites.We analyzed trematode-induced pits from C. gallina valves sourced from five samples (0.375 dm3 each) of two sediment cores (cores 240-S8 and 223-S5, drilled by the Emilia-Romagna Geological Survey) from late-Holocene variably fluvial-influenced shoreface deposits, and 11 modern death assemblages (MDAs; 0.375 dm3) from shoreface settings of the Adriatic Sea (Fig. 1; Table 1). The Holocene stratigraphic architecture of the study area was reported by Amorosi et al. (2016). Sediment samples were dried at 45 °C for 24 h, soaked in 4% H2O2, and wet-sieved with a 1 mm sieve (Huntley and Scarponi, 2012). We selected 31 well-preserved, non-parasitized valves for low-precision radiocarbon dating (Bush et al., 2013).We took scaled photomicrographs of all valves with trematode-induced pits oriented perpendicular to the commissural plane. Using ImageJ 1.53t (https://imagej.net/ij/; Schneider et al., 2012), we measured bivalve anterior-posterior length, dorsal-ventral height, and the primary and secondary axes of the trematode-induced pit (Fig. 1). Trematode prevalence is the proportion of valves within a sample with at least one trematode-induced pit. Trematode pits are equally likely to occur on both valves, and no correction is needed to calculate prevalence, as with predatory drill holes (Huntley, 2007). Pit size is the geometric mean of the primary and secondary axes, and host body size is the geometric mean of valve length and height. We used the mclust package (Scrucca et al., 2016; https://mclust-org.github.io/mclust/) in R (R Core Team, 2021) to produce Gaussian finite mixture models of pit size separately for Holocene and modern samples. This function has two model types for univariate size data: one assuming equal variance, and one assuming varying variance values of the Gaussian distributions. Bayesian information criterion (BIC) values are calculated for one to nine Gaussian distributions under both model types, and the preferred model is the one with the highest BIC value (see Supplemental Material1 for assumptions of this approach). We conducted Spearman correlation analyses comparing the median pit size to the number of pits per valve in the Holocene and modern samples.The ten radiocarbon dates of C. gallina from core 240-S8 are consistent with their stratigraphic position with a median calendar age of 175 BCE (interquartile range of valve calendar ages [IQR] = 267 yr) for the 13.1 m sample and a median of 679 BCE (IQR = 88 yr) for the 13.7 m sample (Table 1). The median age for pooled core samples is 543 BCE (IQR = 444 yr), which is significantly older than the median age (1819 CE, IQR = 297 yr) of the 21 pooled MDA dates (Fig. 2; Table 1; Table S1 in the Supplemental Material). Three outliers from Grado are more than 1 k.y. older than the other specimens, leading us to be concerned about greater time averaging (and possible spurious trends) in the northern MDA samples from such starved coastal settings. Therefore, we conducted analyses on both MDA data and a subset that excludes samples from Grado, Caorle, and Lido di Jesolo (conservative MDA [CMDA]: median age = 1913 CE, IQR = 190 yr; comparable to Holocene sample time averaging).We identified 838 trematode pits in 240 infested valves out of 4567 examined C. gallina valves (Table 1). Trematode prevalence, measured by either pooled valves or median of sample prevalence, is significantly higher in Holocene samples than in the CMDA or MDA samples (Fig. 2; Table S2). Because trematode pits tend to be found on larger valves (Huntley, 2007; Huntley and Scarponi, 2012, 2021), we investigated valve size. CMDA and MDA C. gallina median host valve sizes (21.3 mm and 22.1 mm, respectively) are significantly larger than those from Holocene cores (14.2 mm) and are not distinguishable from one another (Fig. 2; Table S3). Trematode-induced pit size ranged from 0.117 mm to 1.708 mm. The median Holocene pit size (0.497 mm) is significantly larger than that of CMDA and MDA pits (both 0.335 mm, p = 7.99 × 10–9; Fig. 2; Table S4). No significant (rPearson = −0.025, p = 0.48) correlation exists between host size and pit size (Fig. S1).The Gaussian finite mixture models established that the Holocene (n = 715) and MDA (n = 123) samples are best interpreted as two Gaussian distributions with differing variance structures (Fig. 3; Figs. S2−S4; Tables S5 and S6). The Holocene pit sizes are characterized by a component with a mean size of 0.475 mm, which comprises ~84% of the data, and a second component with a mean size of 0.745 mm for the remaining ~16%. MDA pits are characterized by two components with mean sizes of 0.379 mm (78%) and 0.777 mm (22%) (Fig. 3). The CMDA (n = 62) data are best interpreted as three components with mean sizes of 0.320 mm (20%), 0.518 mm (70%), and 0.814 mm (10%). The median number of pits per infested valve is 3 for Holocene samples and 2 for both MDA and CMDA samples (p = 0.02 and p = 0.003, respectively). There is a significant negative correlation (ρ = −0.193, p = 0.01) between median pit size and the number of pits per valve among Holocene samples. These relationships among MDA and CMDA samples are non-significant (ρ = −0.027, p = 0.88; ρ = −0.271, p = 0.13, respectively; Fig. 3; Table S7).Conservation paleobiology uses geohistorical data to address modern conservation challenges and provides baseline data beyond direct ecological monitoring (Dietl et al., 2015). We determined the baseline for trematode-bivalve interactions prior to anthropogenic impact and demonstrate their collapse in the Adriatic Sea over the past ~2 k.y. Specifically, in settings dominated by C. gallina, trematode prevalence decreased by nearly an order of magnitude (Fig. 2). Even though there is substantial geographic variation in parasitic prevalence within our MDA samples, the highest prevalence value in the modern samples (0.145, Montemarciano; Fig. 1) is less than half of the highest prevalence in Holocene samples (0.402, 13.5 m of core 240-S8; Fig. 2, Table 1), suggesting a real temporal difference. Our previous work demonstrated that trematode pits are positively associated with starved depositional conditions, sea-level rise, and large host body size (Huntley, 2007; Huntley and Scarponi, 2012, 2015, 2021; Huntley et al., 2014; Scarponi et al., 2017). All Holocene samples are from similar sediment types and deposited in a progradational setting, further highlighting their high prevalence values. A slight increase in relative sea level since the late nineteenth century (an average rise of 1.2 mm/yr; Carbognin et al., 2011) does not seem to have had a significant effect on modern parasite prevalence values. Despite host sizes being significantly higher in the modern samples, prevalence is significantly lower relative to the Holocene. This size difference is likely due to the 10-cm-diameter Holocene cores. Nevertheless, this strengthens our argument for declining parasite prevalence. Concurrently, the median number of pits per infested valve decreased significantly. Trematode-induced pits are generally highly aggregated among their Holocene C. gallina hosts, meaning that most hosts have only a few pits, and a few hosts have many (Huntley and Scarponi, 2021). This pattern is common among macroparasite-host interactions and effectively increases the parasite population density (Shaw and Dobson, 1995). The strong decrease in aggregation of parasites indicates a collapse in the effective parasite population density in the study area.At first glance, the significant decrease in median pit size through geological time could indicate the extinction of a larger-sized trematode taxon. However, the density estimation modeling results (Fig. 3) for the Holocene and MDA samples indicate that both samples are made up of two Gaussian distributions: one commonplace small-sized taxon and one that is less common, large, and with greater size variation. An analysis of the CMDA samples suggests similar results, with the presence of a third component that is even smaller, comprising ~20% of the observations. We are hesitant to interpret these latter results at face value because the sample size for the CMDA model (n = 62) approaches the minimum needed to model one-dimensional data with high certainty (Psutka and Psutka, 2019). The important point is that even with the most conservative approach, there was no decrease in the number of trematode pit size groups. In other words, there is no evidence for the loss of a trematode taxon over the past ~2 k.y. in the northern Adriatic Sea.These results are consistent with Zuschin et al.’s (2024) work demonstrating a recent collapse in predator-prey interactions in the northern Adriatic Sea due to anthropogenic habitat degradation and an increase in eutrophication-driven hypoxia. Specifically, they found a decrease in predator abundance, a turnover toward less-preferred prey organisms, and a size increase in a dominant hypoxia-related prey species. The depletion of large predators and consumers has caused a strong simplification of the food web in the region. Importantly, both studies documented parasitism and predation beyond the temporal range of ecological monitoring and provided direct insights into the transition of the nearshore settings of the Adriatic into its current urban sea state.The seeming persistence of parasite taxa during this time of decreasing prevalence does not preclude other significant changes in the life history of trematodes. In the Holocene samples, there was a negative correlation between average pit size and the abundance of pits within a valve—a common pattern among parasites—but this relationship broke down in the modern samples. This situation highlights interesting developmental trade-offs, as metacercaria (infective cyst stage) size can predict adult trematode size in the definitive host, which strongly predicts reproductive output (Saldanha et al., 2009). Thus, the higher aggregation of pits indicates higher population density of trematodes in the late Holocene. Conversely, lower pit aggregation in the modern samples results in reduced population density, making it less likely for metacercariae to reach adulthood in their definitive host. This is concerning because parasites are key components of healthy ecosystems. Parasite diversity often mirrors that of the community on which they depend for transmission (Kamiya et al., 2014). Studies of restored ecosystems display a rebound in the prevalence and richness of digenetic parasites (Huspeni and Lafferty, 2004), and parasitism exhibits a strong positive correlation with diversity and negative correlations with origination and extinction rates across the Phanerozoic (De Baets et al., 2021b). However, anthropogenic impacts can decouple the relationship between host and parasite biodiversity (Wood et al., 2018). Our findings are consistent with other studies that show a decrease in the equitability of C. gallina–dominated communities and a reduction of the targeted bivalve over time in the Adriatic Sea (Scarponi et al., 2023; Carlucci et al., 2024), raising concern about the near-future status of C. gallina.Approximately 2.5 k.y. ago, the study area saw a general trend toward a dryer climate (Stefani, 2017) and rapid complex progradational coastal dynamics south of the Po Delta (Amorosi et al., 2017). The region experienced sustained local anthropogenic impacts beginning in Roman times, which became more widespread and intense since the mid-twentieth century (Trincardi et al., 2023). The timing of the collapse of parasitism suggests a link to anthropogenic causes, but testing the specific drivers is not possible with our data. However, pollution and other anthropogenic-related stresses are often associated with a reduction in parasite species richness (Marcogliese, 2004). Parasite transmission may be hindered when intermediate hosts are reduced and free-living life stages (cercariae) are directly affected by these stressors (Mackenzie, 1999). Ocean acidification could be one such factor. The pH of northern Adriatic Sea water decreased by 0.063 units between 1983 and 2008 CE (Luchetta et al., 2010), twice the current global average rate (Hönisch et al., 2012). Experiments have shown that lowered pH reduces cercarial longevity and metacercarial survival (MacLeod and Poulin, 2015). The timing of ocean acidification as a driver of the pattern of parasitism decline is consistent with the ages of our CMDA samples.Our study shows that the dynamics of certain parasite-host relationships reconstructed from recent sedimentary successions can offer a quantitative benchmark against which to evaluate the status of present-day ecosystems. In this case, trematodes became much less common with the onset of human influence, infecting their hosts less frequently and less intensely. We interpret this decline in parasitism to be cause for concern and consistent with numerous other lines of evidence for plummeting ecosystem function and health.We appreciate the helpful feedback provided by K. Flessa, J. Smith, and an anonymous reviewer. L.A. Huntley assisted with fieldwork. This work was financed by the Association for Women Geoscientists–Osage Chapter (Lawrence, Kansas, USA), National Science Foundation grant EAR CAREER 1650745, and the European Union–NextGenerationEU through the Italian Ministry of University and Research under Piano Nazionale di Ripresa e Resilienza (PNRR): Mission 4 Component C2, Investment 1.1 “Conservation of life on Earth: The fossil record as an unparalleled archive of ecological and evolutionary responses to past warming events.”

中文翻译：

---

变化之海：追踪意大利亚得里亚海北部过去千年的寄生动态


我们的研究使用全新世核心样本和现代死亡组合的数据来了解亚得里亚海北部（意大利）人类引起的环境变化如何影响具有重要经济意义的双壳类变色龙的寄生虫-宿主动态。 31 个放射性碳测年结果证实了重大人类影响发生之前和之后时期之间的时间差异，并证实吸虫流行率在过去约 2 公里内下降了一个数量级。寄生虫聚集也显着减少，表明双殖吸虫有效种群规模大幅减少。凹坑尺寸的高斯有限混合模型不支持寄生虫灭绝的假设。综合起来，这些结果表明，与人类对亚得里亚海的影响及其向“城市海洋”的转变相一致，鸡蝇中寄生虫与宿主相互作用的崩溃（可能正在持续）。寄生虫在生态系统中发挥着重要作用，并且可以产生重大影响。对生态系统稳定性的影响（Hudson 等，2006）。因此，寄生虫的流行、分布或多样性的变化可能预示着由于栖息地退化、污染或气候变化而导致的生态功能破坏（Wood et al., 2023）。然而，生态监测由于时间限制，无法区分短期波动和长期格局转变（即生态系统结构的大而持久的变化）。在这方面，长期地质数据可以为当前人口状况和趋势提供关键背景。鉴于其巨大的生态影响，拥有类似的寄生虫长期数据是有用的。 尽管小型软体寄生虫的保存潜力较低，但一些寄生虫与宿主的相互作用记录在宿主的骨骼上（Huntley et al., 2021），使得观察的时间尺度远远超出了最长的观察时间尺度生态研究（De Baets 等人，2021a）。我们的目标是利用这一未充分利用的化石记录来重建最近的寄生虫-宿主动态，并将这些动态置于人类对生态系统影响不断增加的更广泛的历史背景中（例如，Huntley 和 Scarponi，2021）。吸虫扁形虫（扁形动物）是复杂的生命周期寄生虫（即具有多个宿主）会在双壳类宿主的壳内部诱导特征性的、可保留的生长反应（点蚀）（Ruiz 和 Lindberg，1989；Huntley 等人，2021）。我们利用全新世沉积记录来评估在人类对亚得里亚海海岸生态系统产生重大影响之前和之后，Chamelea gallina（一种在生态和经济上都很重要的双壳类动物）的寄生虫感染情况，亚得里亚海是全球变化最严重的海洋区域之一（Lotze 等）等，2011）。具体来说，通过将放射性碳测年与寄生虫强度（每个宿主的坑的数量）和吸虫引起的坑的大小分布的分析相结合，我们扩大了新生儿学研究的时间和地理范围，支持海洋环境中寄生性人为下降的主张（Wood等人，2023）。我们假设过去 2 年内寄生虫与宿主的相互作用有所下降，其特点是寄生虫流行率和强度下降（考虑到沉积环境、时间平均和宿主体型的差异）以及寄生虫类群的灭绝。 在这项工作中，我们从意大利的沉积层序中揭示了 C. gallina 寄生虫感染崩溃的案例，该案例发生在亚得里亚海转变为“城市海”（沿海河口和流域盆地的边缘海）期间。人类广泛开发和居住）。虽然寄生性的崩溃可能看起来是积极的，但它标志着营养结构的简化以及重要生态功能和寄生虫的丧失。我们分析了来自 C. gallina 瓣膜的吸虫引起的凹坑，这些凹坑取自 5 个样本（每个 0.375 dm3）两个沉积物岩心（岩心 240-S8 和 223-S5，由艾米利亚-罗马涅地质调查局钻探），来自全新世晚期受各种河流影响的岸面沉积物，以及来自亚得里亚海沿岸环境的 11 个现代死亡组合（MDA；0.375 dm3）海洋（图 1；表 1）。 Amorosi 等人报告了研究区域的全新世地层结构。 （2016）。沉积物样品在 45 °C 下干燥 24 h，浸泡在 4% H2O2 中，并用 1 mm 筛子进行湿筛（Huntley 和 Scarponi，2012）。我们选择了 31 个保存完好的非寄生瓣膜进行低精度放射性碳测年（Bush 等人，2013）。我们拍摄了所有瓣膜的比例显微照片，这些瓣膜具有垂直于连合平面的吸虫引起的凹坑。使用 ImageJ 1.53t (https://imagej.net/ij/; Schneider et al., 2012)，我们测量了双壳类前后长度、背腹高度以及吸虫引起的凹坑的主轴和次轴 (图。1）。吸虫流行率是指样本中至少有一个吸虫引起的小坑的瓣膜比例。吸虫坑同样可能出现在两个瓣膜上，并且不需要像掠夺性钻孔那样进行校正来计算患病率（Huntley，2007）。 坑尺寸为主轴、副轴的几何平均值，主机体尺寸为阀门长度和高度的几何平均值。我们使用 R（R 核心团队，2021）中的 mclust 包（Scrucca 等人，2016；https://mclust-org.github.io/mclust/）分别为全新世和现代样品。对于单变量大小数据，该函数有两种模型类型：一种假设方差相等，另一种假设高斯分布的方差值不同。在两种模型类型下计算 1 到 9 个高斯分布的贝叶斯信息准则 (BIC) 值，首选模型是具有最高 BIC 值的模型（有关此方法的假设，请参阅补充材料 1）。我们进行了 Spearman 相关分析，比较了全新世和现代样本中的中位坑尺寸与每个瓣膜的坑数量。来自 240-S8 岩心的 C. gallina 的 10 个放射性碳测年与其地层位置一致，中位日历年龄为 175 13.1 m 样本的 BCE（阀门日历年龄的四分位距 [IQR] = 267 年）和 13.7 m 样本的中位数为 679 BCE（IQR = 88 年）（表 1）。汇总核心样本的中位年龄为公元前 543 年（IQR = 444 岁），明显比 21 个汇总 MDA 日期的中位年龄（公元 1819 年，IQR = 297 岁）大（图 2；表 1；表 S1）在补充材料中）。来自格拉多的三个异常值比其他样本早 1 公里以上，这使我们担心来自这种饥饿的沿海环境的北部 MDA 样本的时间平均（以及可能的虚假趋势）更大。 因此，我们对 MDA 数据和不包括来自 Grado、Caorle 和 Lido di Jesolo 的样本的子集进行了分析（保守 MDA [CMDA]：中位年龄 = 1913 CE，IQR = 190 岁；与全新世样本时间平均相当）。我们在检查的 4567 个鸡瓣膜中的 240 个受感染的瓣膜中发现了 838 个吸虫坑（表 1）。全新世样本中的吸虫患病率（通过合并瓣膜或样本患病率中位数来测量）显着高于 CMDA 或 MDA 样本（图 2；表 S2）。由于吸虫坑往往出现在较大的瓣膜上（Huntley，2007；Huntley 和 Scarponi，2012、2021），因此我们研究了瓣膜尺寸。 CMDA 和 MDA C. gallina 中位宿主瓣膜尺寸（分别为 21.3 毫米和 22.1 毫米）明显大于全新世岩心的尺寸（14.2 毫米），并且彼此无法区分（图 2；表 S3）。吸虫引起的凹坑尺寸范围为 0.117 毫米至 1.708 毫米。全新世坑的中值尺寸（0.497 mm）明显大于 CMDA 和 MDA 坑的尺寸（均为 0.335 mm，p = 7.99 × 10–9；图 2；表 S4）。宿主大小和坑大小之间不存在显着的相关性（rPearson = -0.025，p = 0.48）（图S1）。高斯有限混合模型确定全新世（n = 715）和MDA（n = 123）样本是最好的解释为具有不同方差结构的两个高斯分布（图 3；图 S2−S4；表 S5 和 S6）。全新世坑尺寸的特征是平均尺寸为 0.475 毫米的成分，占数据的 84%，第二个成分的平均尺寸为 0.745 毫米，占剩余的约 16%。 MDA 凹坑由平均尺寸为 0.379 毫米（78%）和 0.777 毫米（22%）的两个部分组成（图 3）。 CMDA (n = 62) 数据最好解释为平均尺寸为 0.320 mm (20%)、0.518 mm (70%) 和 0.814 mm (10%) 的三个分量。对于全新世样品，每个受感染瓣膜的凹坑数量中位数为 3 个，对于 MDA 和 CMDA 样品，每个受感染瓣膜的凹坑数量中位数为 2 个（分别为 p = 0.02 和 p = 0.003）。全新世样本中，中位凹坑尺寸与每个瓣膜的凹坑数量之间存在显着的负相关性（ρ = -0.193，p = 0.01）。 MDA 和 CMDA 样本之间的这些关系并不显着（分别为 ρ = -0.027，p = 0.88；ρ = -0.271，p = 0.13；图 3；表 S7）。保护古生物学利用地史数据来应对现代保护挑战并提供超出直接生态监测的基线数据（Dietl 等，2015）。我们在人为影响之前确定了吸虫与双壳类相互作用的基线，并证明了它们在过去约 2 年内在亚得里亚海的崩溃。具体来说，在以 C. gallina 为主的环境中，吸虫流行率下降了近一个数量级（图 2） ）。尽管我们的 MDA 样本中寄生虫患病率存在​​很大的地理差异，但现代样本中的最高患病率（0.145，Montemarciano；图 1）还不到全新世样本中最高患病率（0.402，13.5 m 核心）的一半。 240-S8；图 2，表 1)，表明存在真实的时间差异。我们之前的工作表明，吸虫坑与饥饿的沉积条件、海平面上升和宿主体型较大呈正相关（Huntley，2007；Huntley 和 Scarponi，2012、2015、2021；Huntley 等，2014；Scarponi 等） .，2017）。所有全新世样本都来自相似的沉积物类型，并沉积在渐进环境中，进一步凸显了它们的高流行值。 自 19 世纪末以来，相对海平面略有上升（平均每年上升 1.2 毫米；Carbognin 等人，2011 年）似乎并未对现代寄生虫流行率值产生显着影响。尽管现代样本中的宿主大小明显更高，但相对于全新世，患病率却明显较低。这种尺寸差异可能是由于直径为 10 厘米的全新世核心造成的。然而，这强化了我们关于寄生虫流行率下降的论点。同时，每个受感染瓣膜的凹坑中位数显着减少。吸虫引起的凹坑通常在全新世 C. gallina 宿主中高度聚集，这意味着大多数宿主只有几个凹坑，而少数宿主则有很多凹坑（Huntley 和 Scarponi，2021）。这种模式在大型寄生虫-宿主相互作用中很常见，并且有效地增加了寄生虫种群密度（Shaw 和 Dobson，1995）。寄生虫聚集的急剧减少表明研究区域有效寄生虫种群密度的崩溃。乍一看，随着地质时间的推移，中位坑尺寸的显着减小可能表明较大尺寸的吸虫类群的灭绝。然而，全新世和MDA样本的密度估计建模结果（图3）表明，这两个样本均由两种高斯分布组成：一种是常见的小型分类单元，另一种是不太常见、较大且尺寸变化较大的分类单元。对 CMDA 样本的分析表明了类似的结果，但存在甚至更小的第三个成分，约占观察值的 20%。 我们对于从表面上解释后面这些结果犹豫不决，因为 CMDA 模型的样本量 (n = 62) 接近高确定性建模一维数据所需的最小值（Psutka 和 Psutka，2019）。重要的一点是，即使采用最保守的方法，吸虫坑大小组的数量也没有减少。换句话说，没有证据表明亚得里亚海北部过去约 2 公里内吸虫类群消失。这些结果与 Zuschin 等人 (2024) 的研究结果一致，该研究证明了捕食者-被捕食者最近的崩溃由于人为栖息地退化和富营养化导致的缺氧加剧，亚得里亚海北部发生了相互作用。具体来说，他们发现捕食者数量减少，猎物生物转向较不受欢迎的生物体，以及与缺氧相关的占优势的猎物物种体型增大。大型捕食者和消费者的减少导致该地区的食物网严重简化。重要的是，这两项研究都记录了超出生态监测时间范围的寄生和捕食，并为亚得里亚海近岸环境向当前城市海洋状态的转变提供了直接见解。在流行率下降的这段时间内，寄生虫类群看似持续存在，但这并不意味着排除吸虫生活史的其他显着变化。在全新世样本中，平均凹坑大小与瓣膜内凹坑的丰度之间存在负相关性（寄生虫中的常见模式），但这种关系在现代样本中被打破。 这种情况凸显了有趣的发育权衡，因为囊蚴（感染性囊肿阶段）的大小可以预测最终宿主中成虫吸虫的大小，从而强烈预测繁殖产量（Saldanha 等，2009）。因此，坑的聚集程度较高表明全新世晚期吸虫的种群密度较高。相反，现代样本中较低的凹坑聚集导致种群密度降低，使得囊蚴在其最终宿主中达到成年期的可能性较小。这是令人担忧的，因为寄生虫是健康生态系统的关键组成部分。寄生虫的多样性通常反映了它们传播所依赖的社区的多样性（Kamiya et al., 2014）。对恢复的生态系统的研究表明，双生寄生虫的流行率和丰富度有所回升（Huspeni 和 Lafferty，2004），寄生性与显生宙的多样性呈强烈正相关，与起源和灭绝率呈负相关（De Baets 等， 2021b）。然而，人为影响可能会破坏宿主和寄生虫生物多样性之间的关系（Wood 等，2018）。我们的研究结果与其他研究一致，这些研究表明，随着时间的推移，亚得里亚海以 C. gallina 为主的群落的公平性下降，目标双壳类动物减少（Scarponi 等人，2023 年；Carlucci 等人，2024 年），引起人们对 C. gallina 近期状况的担忧。大约 2.5 年前，研究区域出现了气候干燥的总体趋势（Stefani，2017），波河三角洲以南的沿海地区出现了快速复杂的渐进动态（Amorosi 等，2017）。 ，2017）。 该地区从罗马时代开始就经历了持续的局部人为影响，自二十世纪中叶以来，这种影响变得更加广泛和强烈（Trincardi 等人，2023）。寄生现象崩溃的时间表明与人为原因有关，但用我们的数据不可能测试特定的驱动因素。然而，污染和其他人为相关压力通常与寄生虫物种丰富度的减少有关（Marcogliese，2004）。当中间宿主减少并且自由生活阶段（尾蚴）直接受到这些应激源的影响时，寄生虫传播可能会受到阻碍（Mackenzie，1999）。海洋酸化可能就是这样的因素之一。公元 1983 年至 2008 年期间，亚得里亚海北部海水的 pH 值下降了 0.063 个单位（Luchetta 等人，2010 年），是目前全球平均水平的两倍（Hönisch 等人，2012 年）。实验表明，降低 pH 值会降低尾蚴寿命和囊蚴存活率（MacLeod 和 Poulin，2015）。海洋酸化作为寄生性下降模式驱动因素的时间与我们的 CMDA 样本的年龄一致。我们的研究表明，根据最近的沉积序列重建的某些寄生虫-宿主关系的动态可以提供一个定量基准来评估当今生态系统的状况。在这种情况下，随着人类影响的出现，吸虫变得越来越不常见，感染宿主的频率和强度都降低了。我们认为寄生性的下降值得关注，并且与生态系统功能和健康直线下降的许多其他证据相一致。我们感谢 K. Flessa、J. Smith 和一位匿名审稿人提供的有用反馈。洛杉矶亨特利协助进行实地考察。 这项工作由女性地球科学家协会 - 奥塞奇分会（美国堪萨斯州劳伦斯）、国家科学基金会拨款 EAR CAREER 1650745 以及欧盟 - NextGenerationEU 通过意大利大学和研究部 Piano Nazionale di Ripresa e Resilienza 资助(PNRR)：任务 4 组成部分 C2，投资 1.1“保护地球上的生命：化石记录是对过去变暖事件的生态和进化反应的无与伦比的档案。”