Skip to main content
NCBI home page
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2016 Jun 13;113(25):6934–6938. doi: 10.1073/pnas.1604909113

Natural and anthropogenic drivers of calcium depletion in a northern forest during the last millennium

Bérangère A Leys a,1, Gene E Likens b,c, Chris E Johnson d, Joseph M Craine e, Brice Lacroix f, Kendra K McLauchlan a
PMCID: PMC4922151  PMID: 27298361

Significance

This research breaks new ground by showing that, contrary to generally accepted theories of ecosystem development, calcium depletion has been occurring for millennia as a natural consequence of long-term ecosystem development. This natural process predisposed forest ecosystems in the region to detrimental responses to acid rain in the 20th century. We also show that nitrogen availability was increasing concurrently with the depletion of calcium. This is the first study, to our knowledge, to reconstruct continuous changes in nutrient availability for a northern forest ecosystem on the millennial time scale. The results alter our assessments of the speed and trajectory of nutrient limitation in forests and suggest that reformulation of global models of forest productivity may be necessary.

Keywords: calcium, natural depletion, acidification, nutrients, hardwood forest

Abstract

The pace and degree of nutrient limitation are among the most critical uncertainties in predicting terrestrial ecosystem responses to global change. In the northeastern United States, forest growth has recently declined along with decreased soil calcium (Ca) availability, suggesting that acid rain has depleted soil Ca to the point where it may be a limiting nutrient. However, it is unknown whether the past 60 y of changes in Ca availability are strictly anthropogenic or partly a natural consequence of long-term ecosystem development. Here, we report a high-resolution millennial-scale record of Ca and 16 other elements from the sediments of Mirror Lake, a 15-ha lake in the White Mountains of New Hampshire surrounded by northern hardwood forest. We found that sedimentary Ca concentrations had been declining steadily for 900 y before regional Euro-American settlement. This Ca decline was not a result of serial episodic disturbances but instead the gradual weathering of soils and soil Ca availability. As Ca availability was declining, nitrogen availability concurrently was increasing. These data indicate that nutrient availability on base-poor, parent materials is sensitive to acidifying processes on millennial timescales. Forest harvesting and acid rain in the postsettlement period mobilized significant amounts of Ca from watershed soils, but these effects were exacerbated by the long-term pattern. Shifting nutrient limitation can potentially occur within 10,000 y of ecosystem development, which alters our assessments of the speed and trajectory of nutrient limitation in forests, and could require reformulation of global models of forest productivity.

Elemental limitation of terrestrial primary productivity is a master variable that can determine key ecosystem processes such as carbon exchange, ecosystem biodiversity, and water quality (1). Terrestrial primary productivity is most frequently limited by the availability of nitrogen (N), but phosphorus (P) is generally recognized as the limiting nutrient in terrestrial ecosystems with older soils (2). Consequently, a paradigm of progression from N to P limitation has developed (3, 4), but this is far from a predictable phenomenon (5, 6). For example, Ca can also be a limiting element in base-poor ecosystems (7, 8), and experimental Ca addition can increase forest productivity (8, 9).

As a base cation, soil Ca2+ is highly sensitive to pH, and several mechanisms that acidify soil could lead to the development of Ca limitation (10). Understanding nutrient limitation requires consideration of multiple elements, including macronutrients and rock-derived elements such as Al, Si, Ca, Na, Mg, and K. Acid rain—atmospheric deposition of all wet and dry substances that cause acidification—commenced in Europe and eastern North America in the middle of the 20th century and caused nutrient leaching from soils (11). A decline in forest productivity has been observed in affected ecosystems since about 1980 CE (8, 12), demonstrating clear evidence of the effects of acid rain on forest ecosystems (11, 13). Nonanthropogenic Ca depletion in soils also occurs in terrestrial ecosystems due to natural weathering of parent material in concert with plant uptake, long-term N enrichment, and inputs of organic acids produced by plants (14). Although some long-term records seem to indicate patterns of ecosystem acidification over millennial timescales (15), whether these processes were widespread or sufficient enough to cause Ca limitation is largely unknown.

To assess the dynamics of Ca and other key nutrients linked to acidification processes during long-term ecosystem development, we examined the concentrations of 17 elements in a 1,200-y-long sedimentary record from Mirror Lake, New Hampshire, a 15-ha oligotrophic lake in a temperate forest watershed in the White Mountains of the northeastern United States (16) (also see Materials and Methods). These forests were established ∼10,000 y ago following deglaciation and are situated on base-poor gneissic bedrock with relatively slow mineral weathering rates. Past research has suggested Ca limitation of productivity for these forests (8) as well as increasing N availability over the past millennium (17). The site has a typical history for the region: a period of Euro-American deforestation and agriculture beginning ∼1770 CE, followed by abandonment and development of second-growth mixed hardwood forest (16) and chronic acid rain from the mid-20th century to present (12, 18), which caused the export of large quantities of Ca from forests in this region via surficial streamflow (12).

Results and Discussion

Long-Term Calcium Depletion Reflects a Decrease in Weathered Ca.

To determine sedimentary elemental concentrations, we used X-ray fluorescence (XRF) analysis of dried and homogenized sediments (Table S1 and Figs. S1 and S2). Before 1800 CE, four elements showed significant linear declines in concentrations over time: Ca, Mn, Rb, and K (Fig. 1A and Fig. S1). Of these, Ca showed the strongest linear declines (r2 = 0.89; Fig. 1A). Several lines of evidence indicate that the decline in Ca toward the present likely reflects lower inputs of weathered Ca from soils in the watershed. The decline is not solely due to a reduction only in plant-available forms of Ca. The amount of exchangeable Ca remained similar over this time period (P > 0.3; Table S2), indicating that the decrease of Ca concentration in the sediment is not due to labile Ca leached to the bottom of the core. There is also no evidence of a change in the Ca-bearing minerals being supplied to the lake. The estimation of mineral composition of the lake sediment by combining XRF and X-ray diffraction (XRD) shows that the relative abundances of quartz (78 ± 1.3%), albite (11 ± 0.5%), chlorite (8 ± 0.5%), muscovite (2 ± 0.4%), and apatite (0.5 ± 0.05%) did not change significantly through time (Tables S3 and S4). To assess whether the decline in Ca could be due to dilution from an increase in the concentration of other sedimentary components, we compared Ca concentrations with titanium (Ti; Fig. 1B), which is an erosion proxy (19), and organic carbon percentage (%C; Fig. 1C), used as an organic matter proxy (20). Both [Ti] and %C are independent of [Ca] over this period (P > 0.15 for both), indicating that the decline in [Ca] before 1800 CE was not due to dilution by other elements or change in the organic matter proportion of sediment samples. Together with low local atmospheric Ca inputs (12), these data indicate that the declines in sediment Ca concentrations did not reflect increases in erosion, increases in organic matter delivery, changes in sedimentation processes, or changes in the source of Ca. The decrease of Ca concentration thus most likely reflects a decrease of weathered Ca inputs to Mirror Lake.

Table S1.

Elemental concentration from XRF on Mirror Lake sediment samples, expressed in g·kg−1

Sample_ID Up_Depth Bot_Depth Up_Age Bot_Age Na Mg Al Si P S K Ca Ti Mn Fe Co Ni Cu Zn Pb Mo
Mirror Lake 0 0 1 2005 2002 5.814 5.715 31.092 190.275 0.505 6.990 7.607 5.814 2.867 0.465 42.644 0.008 0.003 0.000 0.153 0.019 0.076
Mirror Lake 1 1 2 2002 1998 5.717 5.345 33.341 204.986 0.576 7.033 8.019 6.098 2.979 0.521 31.048 0.010 0.014 0.010 0.172 0.027 0.057
Mirror Lake 2 2 3 1998 1994 5.701 6.332 31.637 191.280 0.453 6.347 7.764 6.025 2.746 0.519 24.844 0.007 0.015 0.015 0.197 0.025 0.063
Mirror Lake 3 3 4 1994 1988 5.566 7.073 31.086 190.522 0.365 6.391 7.739 6.163 2.924 0.492 25.715 0.008 0.016 0.016 0.206 0.028 0.055
Mirror Lake 4 4 5 1988 1981 5.466 6.848 32.881 196.543 0.456 6.741 8.080 6.047 2.929 0.490 24.681 0.010 0.015 0.018 0.217 0.028 0.055
Mirror Lake 5 5 6 1981 1972 5.417 7.321 32.023 191.768 0.385 6.336 7.914 5.974 2.964 0.464 24.153 0.007 0.012 0.020 0.210 0.028 0.056
Mirror Lake 6 6 7 1972 1962 5.290 8.628 29.822 182.522 0.316 6.578 7.699 5.637 2.860 0.444 23.867 0.005 0.016 0.034 0.226 0.027 0.058
Mirror Lake 7 7 8 1962 1949 5.451 7.496 30.275 187.916 0.372 7.348 7.725 5.778 2.987 0.432 24.433 0.008 0.014 0.022 0.245 0.028 0.054
Mirror Lake 8 8 9 1949 1934 5.359 7.552 30.273 191.183 0.425 8.103 7.522 6.060 2.939 0.430 25.009 0.009 0.015 0.022 0.253 0.027 0.055
Mirror Lake 9 9 10 1934 1920 5.265 8.551 29.614 184.588 0.374 6.946 7.359 6.023 2.886 0.433 23.979 0.008 0.011 0.019 0.243 0.025 0.053
Mirror Lake 10 10 11 1920 1907 5.254 8.006 31.225 193.798 0.528 6.002 7.795 6.507 3.004 0.454 21.957 0.008 0.011 0.008 0.184 0.021 0.051
Mirror Lake 11 11 12 1907 1894 5.153 8.961 29.437 183.759 0.521 5.583 7.749 6.485 2.864 0.444 22.312 0.007 0.011 0.005 0.156 0.019 0.054
Mirror Lake 12 12 13 1894 1882 5.203 8.884 29.144 181.571 0.584 5.005 7.958 6.696 2.756 0.456 22.169 0.006 0.010 0.000 0.126 0.016 0.052
Mirror Lake 13 13 14 1882 1872 5.056 8.455 25.921 197.283 0.591 4.595 6.423 6.400 2.424 0.449 20.017 0.005 0.008 0.000 0.112 0.013 0.053
Mirror Lake 14 14 15 1872 1862 4.859 8.013 21.487 215.188 0.599 4.737 4.452 6.031 1.921 0.455 17.398 0.004 0.009 0.000 0.098 0.012 0.058
Mirror Lake 15 15 16 1862 1853 4.190 13.710 13.799 171.621 0.198 3.683 2.356 4.926 1.249 0.393 13.430 0.005 0.011 0.000 0.085 0.011 0.060
Mirror Lake 16 16 17 1853 1841 4.362 9.176 15.639 219.156 0.440 4.417 2.392 5.548 1.307 0.446 14.215 0.004 0.007 0.000 0.080 0.010 0.062
Mirror Lake 17 17 18 1841 1829 4.588 8.022 15.638 228.547 0.516 4.795 2.315 5.696 1.266 0.459 13.912 0.004 0.005 0.000 0.068 0.010 0.060
Mirror Lake 18 18 19 1829 1818 4.378 8.639 14.096 225.430 0.447 4.535 1.816 5.589 1.161 0.451 13.541 0.004 0.005 0.000 0.067 0.010 0.062
Mirror Lake 19 19 20 1818 1806 4.475 7.696 14.274 237.905 0.458 4.897 1.652 5.741 1.046 0.469 10.885 0.005 0.005 0.000 0.061 0.010 0.060
Mirror Lake 20 20 21 1806 1795 4.705 4.578 15.003 268.420 0.687 5.692 1.735 5.965 0.985 0.485 12.560 0.004 0.005 0.000 0.061 0.009 0.060
Mirror Lake 21 21 22 1795 1784 4.715 5.078 14.235 260.389 0.629 5.633 1.558 5.985 0.951 0.482 12.730 0.003 0.003 0.000 0.067 0.009 0.060
Mirror Lake 22 22 23 1784 1772 4.626 5.418 13.926 258.648 0.610 5.573 1.416 6.017 0.908 0.480 10.507 0.003 0.004 0.000 0.062 0.009 0.061
Mirror Lake 23 23 24 1772 1761 4.621 5.803 14.197 254.983 0.602 5.667 1.631 6.042 0.965 0.493 10.863 0.003 0.003 0.000 0.065 0.009 0.060
Mirror Lake 24 24 25 1761 1750 4.701 5.687 13.681 253.596 0.619 5.788 1.478 6.060 0.898 0.485 10.629 0.003 0.005 0.000 0.057 0.009 0.062
Mirror Lake 25 25 26 1750 1738 4.654 5.945 13.875 253.030 0.597 5.604 1.443 6.121 0.898 0.488 10.872 0.003 0.005 0.000 0.064 0.009 0.060
Mirror Lake 26 26 27 1738 1727 4.564 5.629 13.704 259.768 0.664 5.735 1.462 6.171 0.920 0.496 10.763 0.004 0.003 0.000 0.061 0.009 0.061
Mirror Lake 27 27 28 1727 1716 4.579 5.210 13.997 263.607 0.667 5.795 1.508 6.224 0.904 0.492 10.562 0.004 0.002 0.000 0.068 0.009 0.062
Mirror Lake 28 28 29 1716 1704 4.574 5.724 14.395 257.415 0.639 5.910 1.635 6.300 0.966 0.491 11.436 0.005 0.006 0.000 0.066 0.009 0.063
Mirror Lake 29 29 30 1704 1693 4.630 5.076 14.396 264.886 0.657 6.119 1.494 6.345 0.925 0.504 13.624 0.003 0.005 0.000 0.063 0.010 0.060
Mirror Lake 30 30 31 1693 1682 4.592 5.634 14.329 258.706 0.633 6.567 1.526 6.367 0.908 0.500 11.539 0.003 0.005 0.000 0.062 0.010 0.060
Mirror Lake 31 31 32 1682 1670 4.516 5.972 14.354 258.905 0.584 6.693 1.537 6.415 0.905 0.503 11.175 0.004 0.009 0.000 0.063 0.009 0.060
Mirror Lake 32 32 33 1670 1659 4.622 5.549 14.306 257.508 0.539 6.563 1.519 6.379 0.922 0.500 10.991 0.004 0.007 0.000 0.066 0.009 0.059
Mirror Lake 33 33 34 1659 1648 4.532 5.868 13.459 254.446 0.504 6.248 1.348 6.288 0.849 0.492 13.229 0.003 0.005 0.000 0.063 0.009 0.060
Mirror Lake 34 34 35 1648 1636 4.537 5.742 14.242 258.767 0.530 6.440 1.596 6.410 0.930 0.508 11.099 0.004 0.005 0.000 0.064 0.010 0.059
Mirror Lake 35 35 36 1636 1625 4.520 5.411 13.893 261.865 0.525 6.239 1.431 6.504 0.889 0.514 10.884 0.004 0.003 0.000 0.058 0.009 0.060
Mirror Lake 36 36 37 1625 1614 4.504 6.003 13.939 255.589 0.507 6.112 1.376 6.465 0.854 0.517 10.863 0.004 0.005 0.000 0.057 0.010 0.060
Mirror Lake 37 37 38 1614 1602 4.529 5.944 13.685 255.332 0.490 6.396 1.389 6.573 0.858 0.508 10.992 0.004 0.005 0.000 0.062 0.009 0.059
Mirror Lake 38 38 39 1602 1591 4.454 6.542 13.576 251.568 0.511 6.590 1.383 6.423 0.872 0.505 10.940 0.004 0.004 0.000 0.060 0.009 0.060
Mirror Lake 39 39 40 1591 1580 4.543 5.346 13.875 261.261 0.556 6.597 1.377 6.455 0.864 0.510 10.914 0.004 0.005 0.000 0.059 0.009 0.062
Mirror Lake 40 40 41 1580 1568 4.623 5.492 13.968 257.125 0.538 6.322 1.477 6.483 0.889 0.523 10.932 0.005 0.001 0.000 0.055 0.010 0.060
Mirror Lake 41 41 42 1568 1557 4.624 5.090 13.266 262.031 0.556 6.612 1.174 6.430 0.850 0.508 10.250 0.004 0.005 0.000 0.050 0.009 0.060
Mirror Lake 42 42 43 1557 1546 4.558 4.477 13.790 273.489 0.647 6.323 1.272 6.490 0.863 0.517 12.601 0.005 0.006 0.000 0.056 0.009 0.062
Mirror Lake 43 43 44 1546 1534 4.616 3.520 14.081 283.535 0.697 6.007 1.276 6.583 0.877 0.522 12.191 0.004 0.002 0.000 0.051 0.008 0.060
Mirror Lake 44 44 45 1534 1523 4.483 4.893 13.328 271.649 0.581 5.713 1.165 6.402 0.823 0.515 11.926 0.004 0.007 0.000 0.054 0.009 0.061
Mirror Lake 45 45 46 1523 1511 4.640 3.376 14.261 286.968 0.732 6.023 1.179 6.616 0.861 0.532 11.977 0.003 0.004 0.000 0.044 0.009 0.061
Mirror Lake 46 46 47 1511 1500 4.632 3.276 14.783 286.268 0.715 6.005 1.325 6.584 0.849 0.532 12.020 0.003 0.004 0.000 0.049 0.009 0.060
Mirror Lake 47 47 48 1500 1489 4.653 3.244 13.931 286.986 0.692 5.934 1.196 6.617 0.819 0.524 12.316 0.005 0.008 0.000 0.050 0.009 0.061
Mirror Lake 48 48 49 1489 1477 4.485 6.400 12.811 255.434 0.403 5.141 0.964 6.349 0.782 0.505 12.263 0.004 0.006 0.000 0.050 0.009 0.062
Mirror Lake 49 49 50 1477 1466 4.696 3.033 14.108 287.779 0.709 6.132 1.077 6.705 0.859 0.522 12.537 0.004 0.011 0.000 0.059 0.009 0.061
Mirror Lake 50 50 51 1466 1455 4.593 3.297 13.743 287.058 0.759 5.847 1.106 6.756 0.795 0.533 12.222 0.003 0.005 0.000 0.049 0.009 0.063
Mirror Lake 51 51 52 1455 1443 4.721 2.832 13.832 290.141 0.741 6.145 1.166 6.696 0.804 0.541 12.164 0.004 0.005 0.000 0.053 0.010 0.060
Mirror Lake 52 52 53 1443 1432 4.626 4.897 12.852 264.387 0.493 5.346 1.009 6.485 0.794 0.508 12.063 0.003 0.008 0.000 0.054 0.008 0.063
Mirror Lake 53 53 54 1432 1421 4.687 3.183 13.515 285.296 0.678 5.856 1.155 6.686 0.814 0.539 11.911 0.004 0.007 0.000 0.047 0.009 0.062
Mirror Lake 54 54 55 1421 1409 4.508 5.649 12.759 260.217 0.462 5.374 0.979 6.426 0.792 0.515 11.964 0.004 0.006 0.000 0.049 0.008 0.062
Mirror Lake 55 55 56 1409 1398 4.563 5.903 12.640 252.871 0.425 4.965 0.916 6.292 0.776 0.516 12.035 0.004 0.000 0.000 0.045 0.009 0.060
Mirror Lake 56 56 57 1398 1387 4.666 3.194 13.779 287.253 0.722 5.927 1.142 6.674 0.826 0.525 11.947 0.004 0.002 0.000 0.055 0.010 0.061
Mirror Lake 57 57 58 1387 1375 4.737 2.992 13.706 286.244 0.792 6.008 1.184 6.690 0.800 0.530 12.095 0.004 0.004 0.000 0.051 0.009 0.062
Mirror Lake 58 58 59 1375 1364 4.582 3.332 13.874 286.655 0.786 5.868 1.016 6.680 0.796 0.533 11.975 0.003 0.003 0.000 0.049 0.009 0.061
Mirror Lake 59 59 60 1364 1353 4.555 5.018 12.887 267.209 0.561 5.521 1.084 6.563 0.749 0.529 12.274 0.004 0.002 0.000 0.055 0.009 0.063
Mirror Lake 60 60 61 1353 1341 4.564 3.354 13.628 287.516 0.841 5.968 1.379 6.696 0.774 0.529 12.335 0.004 0.009 0.000 0.062 0.009 0.064
Mirror Lake 61 61 62 1341 1330 4.673 2.939 14.045 289.580 0.888 6.262 1.374 6.752 0.841 0.542 12.105 0.003 0.009 0.000 0.049 0.009 0.061
Mirror Lake 62 62 63 1330 1319 4.668 3.267 13.574 284.522 0.813 6.011 1.300 6.700 0.851 0.538 12.549 0.006 0.008 0.000 0.054 0.009 0.061
Mirror Lake 63 63 64 1319 1307 4.555 3.221 14.365 289.248 0.836 6.267 1.397 6.758 0.843 0.548 12.428 0.004 0.003 0.000 0.048 0.009 0.063
Mirror Lake 64 64 65 1307 1296 4.605 3.337 14.026 284.968 0.809 6.137 1.523 6.915 0.810 0.545 12.502 0.005 0.005 0.000 0.053 0.008 0.060
Mirror Lake 65 65 66 1296 1285 4.512 5.773 13.205 257.465 0.570 5.372 1.326 6.592 0.784 0.533 12.558 0.004 0.004 0.000 0.052 0.009 0.060
Mirror Lake 66 66 67 1285 1273 4.588 3.354 13.936 287.301 0.812 6.247 1.527 6.849 0.927 0.544 12.263 0.004 0.004 0.000 0.052 0.008 0.062
Mirror Lake 67 67 68 1273 1262 4.598 3.512 14.111 284.410 0.762 6.348 1.449 6.911 0.804 0.540 12.693 0.006 0.009 0.000 0.057 0.009 0.060
Mirror Lake 68 68 69 1262 1251 4.547 5.645 13.245 259.620 0.505 5.451 1.402 6.688 0.724 0.530 12.519 0.004 0.008 0.000 0.053 0.009 0.060
Mirror Lake 69 69 70 1251 1239 4.656 2.948 13.961 289.316 0.762 6.128 1.510 6.915 0.849 0.546 12.286 0.005 0.004 0.000 0.050 0.009 0.060
Mirror Lake 70 70 71 1239 1228 4.633 3.239 13.942 286.991 0.709 6.024 1.389 6.906 0.915 0.535 12.547 0.003 0.001 0.000 0.051 0.008 0.061
Mirror Lake 71 71 72 1228 1217 4.564 3.206 14.307 288.591 0.724 6.040 1.610 6.805 0.834 0.546 12.397 0.003 0.004 0.000 0.057 0.008 0.061
Mirror Lake 72 72 73 1217 1205 4.556 3.627 14.299 285.930 0.719 6.026 1.556 6.936 0.913 0.558 12.648 0.004 0.010 0.000 0.055 0.010 0.060
Mirror Lake 73 73 74 1205 1194 4.667 3.629 13.698 282.158 0.725 5.988 1.492 6.930 0.843 0.543 12.577 0.003 0.011 0.000 0.049 0.008 0.060
Mirror Lake 74 74 75 1194 1183 4.543 3.758 13.829 284.458 0.727 6.059 1.638 7.010 0.936 0.546 12.229 0.003 0.001 0.000 0.053 0.009 0.058
Mirror Lake 75 75 76 1183 1171 4.619 3.397 14.384 284.608 0.693 6.255 1.749 6.875 0.850 0.537 12.585 0.003 0.004 0.000 0.053 0.008 0.061
Mirror Lake 76 76 77 1171 1160 4.639 3.711 14.239 283.502 0.699 6.122 1.723 7.000 0.911 0.543 12.725 0.004 0.000 0.000 0.056 0.009 0.060
Mirror Lake 77 77 78 1160 1149 4.560 4.734 13.503 270.799 0.657 5.956 1.613 7.044 0.875 0.546 10.996 0.003 0.005 0.000 0.055 0.009 0.060
Mirror Lake 78 78 79 1149 1137 4.586 4.827 13.678 266.945 0.605 5.847 1.705 6.883 0.897 0.535 12.891 0.004 0.006 0.000 0.060 0.008 0.059
Mirror Lake 79 79 80 1137 1126 4.610 5.200 13.559 262.912 0.560 5.859 1.808 7.091 0.922 0.536 12.762 0.005 0.003 0.000 0.055 0.008 0.059
Mirror Lake 80 80 81 1126 1115 4.496 6.263 12.974 252.488 0.484 5.538 1.562 6.852 0.890 0.535 13.127 0.004 0.005 0.000 0.053 0.009 0.060
Mirror Lake 81 81 82 1115 1103 4.598 5.016 14.089 264.651 0.592 5.936 1.833 7.219 0.953 0.548 11.454 0.004 0.005 0.000 0.056 0.009 0.060
Mirror Lake 82 82 83 1103 1092 4.564 5.835 13.013 258.081 0.529 5.686 1.665 6.976 0.858 0.539 11.472 0.004 0.008 0.000 0.058 0.009 0.058
Mirror Lake 83 83 84 1092 1081 4.474 5.972 13.522 257.147 0.450 5.777 1.767 6.934 0.926 0.534 11.363 0.004 0.008 0.000 0.060 0.008 0.058
Mirror Lake 84 84 85 1081 1069 4.547 5.552 13.572 258.918 0.459 5.820 1.851 6.996 0.920 0.533 13.690 0.005 0.008 0.000 0.053 0.009 0.059
Mirror Lake 85 85 86 1069 1058 4.579 4.430 13.759 271.086 0.560 6.172 1.792 7.210 0.885 0.537 11.469 0.004 0.003 0.000 0.057 0.009 0.058
Mirror Lake 86 86 87 1058 1047 4.533 5.252 13.478 262.483 0.478 5.766 1.764 7.057 0.883 0.540 13.025 0.003 0.006 0.000 0.061 0.009 0.059
Mirror Lake 87 87 88 1047 1035 4.502 5.782 13.525 259.639 0.480 5.788 1.718 7.051 0.893 0.539 13.146 0.003 0.005 0.000 0.062 0.008 0.060
Mirror Lake 88 88 89 1035 1024 4.556 5.406 13.638 262.010 0.447 5.963 1.788 7.141 0.932 0.533 13.458 0.004 0.007 0.000 0.059 0.009 0.060
Mirror Lake 89 89 90 1024 1013 4.554 5.981 13.104 254.118 0.438 5.778 1.681 7.125 0.868 0.532 11.154 0.004 0.008 0.000 0.060 0.009 0.059
Mirror Lake 90 90 91 1013 1001 4.670 6.103 13.429 247.463 0.431 5.771 1.626 7.202 0.862 0.541 13.599 0.004 0.007 0.000 0.062 0.009 0.060
Mirror Lake 91 91 92 1001 990 4.575 8.073 12.584 229.890 0.314 5.498 1.709 6.947 0.894 0.517 13.047 0.004 0.009 0.000 0.065 0.008 0.061
Mirror Lake 92 92 93 990 979 4.594 5.637 13.765 255.654 0.462 6.196 1.803 7.336 0.930 0.548 11.463 0.005 0.006 0.000 0.065 0.009 0.063
Mirror Lake 93 93 94 979 967 4.658 6.202 13.176 247.121 0.444 5.993 1.706 7.241 0.877 0.545 11.643 0.003 0.005 0.000 0.059 0.008 0.061
Mirror Lake 94 94 95 967 956 4.627 6.983 13.014 238.906 0.415 5.710 1.702 7.114 0.918 0.539 11.812 0.004 0.010 0.000 0.056 0.009 0.062
Mirror Lake 95 95 96 956 944 4.723 5.227 14.054 258.007 0.517 6.277 1.826 7.337 0.940 0.552 12.001 0.004 0.007 0.000 0.065 0.009 0.060
Mirror Lake 96 96 97 944 933 4.645 5.855 13.723 251.314 0.478 6.324 1.907 7.301 0.931 0.550 13.676 0.004 0.006 0.000 0.062 0.010 0.061
Mirror Lake 97 97 98 933 922 4.717 5.351 13.582 254.992 0.542 6.668 1.856 7.398 0.940 0.557 14.072 0.004 0.008 0.000 0.062 0.009 0.061
Mirror Lake 98 98 99 922 910 4.573 7.198 13.270 239.032 0.439 6.150 1.745 7.228 0.867 0.548 12.717 0.004 0.006 0.000 0.060 0.009 0.059
Mirror Lake 99 99 100 910 899 4.662 5.334 14.076 257.167 0.570 6.746 1.741 7.563 0.905 0.565 12.674 0.005 0.005 0.000 0.064 0.009 0.059
Mirror Lake 100 100 101 899 888 4.713 7.145 13.366 232.470 0.416 6.396 1.814 7.497 0.925 0.551 11.984 0.004 0.009 0.000 0.048 0.008 0.070
Mirror Lake 101 101 102 888 876 4.583 6.427 12.888 246.375 0.475 6.242 1.629 7.302 0.871 0.541 15.225 0.004 0.010 0.000 0.062 0.008 0.062
Mirror Lake 102 102 103 876 873 4.639 7.239 12.243 233.785 0.439 5.655 1.572 7.027 0.879 0.527 16.904 0.006 0.008 0.000 0.069 0.009 0.060
Open in a new tab

Fig. S1.

Fig. S1.

Open in a new tab

Concentrations of 12 elements in Mirror Lake, New Hampshire sediments plotted over the presettlement period (from 876 to 1800 CE). Black curves are lowess smooth splines (span, 0.9). Seven events significantly different from baseline (on PC1 from 876 to 1800 CE) are colored red. Coefficients of determination (r2) are reported only for significant linear regressions (P < 0.01).

Fig. S2.

Fig. S2.

Open in a new tab

Concentrations of 17 elements derived from XRF and the percentage of carbon in Mirror Lake, New Hampshire sediments plotted over time from 1600 to 2000 AD. Black curves are lowess smooth splines (span, 0.5).

Fig. 1.

Fig. 1.

Open in a new tab

Elemental composition of sediment deposited from 876 to 1800 CE (A) Ca concentration, expressed in g·kg−1, P value, and r2 are derived from a linear model. (B) Ti concentration, expressed in g·kg−1, a proxy of erosion of the catchment. (C) Percent organic C, a proxy of organic matter concentration. (D) BSI calculated as the quotient of the sum of concentrations of Ca, Mg, K, and Na and the sum of concentrations of Ca, Mg, K, Na, and Al. Black curves are lowess smooth splines. Seven events significantly different from baseline conditions signified by a principal component axis from 876 to 1800 CE are colored red.

Table S2.

Concentrations of exchangeable and total potassium (K), calcium (Ca), magnesium (Mg), and sodium (Na) in the sediments, expressed in ppm

Depth, cm Age, CE Exch. K, ppm Total K, ppm Exch. portion of K, % Exch. Ca, ppm Total Ca, ppm Exch. portion of Ca, % Exch. Mg, ppm Total Mg, ppm Exch. portion of Mg, % Exch. Na, ppm Total Na, ppm Exch. portion of Na, %
10 1920 84.0 66,166.3 0.1 2,143.1 84,317.7 2.5 181.5 3,583.6 5.1 80.0 3,792.6 2.1
43 1545 97.6 27,733.1 0.4 2,906.4 83,161.7 3.5 173.2 3,620.4 4.8 76.5 3,790.0 2.0
50 1466 100.0 26,859.6 0.4 2,968.7 85,571.0 3.5 180.0 3,644.6 4.9 67.4 3,812.5 1.8
61 1341 87.1 28,368.4 0.3 2,827.6 85,377.0 3.3 168.5 3,724.9 4.5 57.3 3,867.2 1.5
70 1239 104.2 28,571.4 0.4 2,992.2 87,560.3 3.4 175.5 3,671.0 4.8 60.9 3,863.2 1.6
78 1148 94.5 30,390.9 0.3 2,962.0 87,217.6 3.4 162.3 3,523.6 4.6 60.1 3,688.6 1.6
Open in a new tab

The exchangeable portion of each element has been calculated as exchangeable concentration/total concentration * 100. and is expressed in %. Exch., exchangeable.

Table S3.

Calcium concentrations, expressed in g·kg−1, from XRF techniques on 10 Mirror Lake dried powdered samples and on the same samples after an LOI at 1,000 °C for 2 h

Depth, cm Age, CE Ca concentration in the sediment samples Percentage of matter remained after an LOI at 1,000 °C for 2 h Ca concentration in the samples after an LOI at 1,000 °C for 2 h
14 1872 6.03 9.66 7.68
20 1806 5.97 6.44 7.32
34 1647 6.41 5.90 7.92
48 1488 6.35 6.26 7.84
59 1364 6.56 6.11 7.81
63 1318 6.76 6.04 8.03
65 1296 6.59 6.35 8.01
68 1262 6.69 5.84 8.16
73 1205 6.93 5.80 8.21
96 944 7.30 6.58 9.26
Open in a new tab

Percentage of matter remaining after the LOI at 1,000 °C for 2 h is also indicated.

Table S4.

Main mineral composition of Mirror Lake sediments, derived from XRD techniques, expressed in normalized percentages

Minerals Quartz Albite Chlorite Mica Apatite
Chemical formula SiO2 (Na, Ca)[AlSi3O8] (Mg,Fe2+)5Al2Si3O10(OH)8 KAl3Si3O10(OH)2 Ca5(PO4)3(OH)
Mirror Lake 20 76.69 11.56 8.89 2.26 0.61
Mirror Lake 34 78.62 10.61 8.24 2.04 0.49
Mirror Lake 59 79.43 10.55 7.99 1.51 0.52
Mirror Lake 68 79.68 10.30 7.57 1.88 0.56
Mirror Lake 96 77.47 10.75 8.81 2.49 0.49
Mean 78.38 10.75 8.30 2.04 0.53
SD 1.28 0.48 0.56 0.37 0.05
Open in a new tab

Percentages are derived from XRF element concentration, transformed in oxides, and matched with chemical formula of main minerals.

Ca Depletion as a Natural Consequence of Long-Term Ecosystem Development.

To determine whether declining presettlement Ca availability was due to serial large-scale disturbances that depleted Ca stocks or gradual weathering, we tested whether Ca concentrations increased during any disturbance events identifiable in the sediment record before 1800 CE Principal component analysis (PCA) of the elemental concentration data before 1800 CE and calculation of a Base Saturation Index (BSI) identified seven disturbance events between 1250 and 1550 CE (Fig. 1D, Fig. 2A, and Supporting Information). BSI indicated that sediments deposited during the seven events exhibit slightly more basic characteristics during an overall acidic period. This period corresponds to the glacierization phase of the Little Ice Age in Europe and North America, a period characterized by cool summers and wet winters (21). These conditions would be favorable to the formation of ice storms, which are documented as common disturbances in this region, and have been shown to damage trees and increase the flux of nutrient elements from forests (22). During the seven disturbance events, the concentrations of five elements are significantly lower (P = 0.001) than the intervening periods between 1250 and 1550 CE, including Al and Si, two elements that are weak acids in their dissolved forms (Fig. 2 B and C and Fig. S1). Higher Ca concentrations in sediments during these events would have suggested an increase in loss rates of Ca from the bulk soil in the watershed during disturbance, but Ca concentrations are actually 4% lower during these events than the surrounding time (P = 0.05). Therefore, the decline in Ca does not appear to be driven by these events but instead seems to be a consequence of sustained weathering stimulated by plant uptake of Ca during ecosystem development (e.g., refs. 21–24).

Fig. 2.

Fig. 2.

Open in a new tab

Patterns of elemental composition of sediment deposited from 876 to 1800 CE (A) Included are the first principal component axis of elemental concentrations (λ = 17%). (B and C) Element concentrations of Al and Si. (D) δ15N in units of ‰. Seven events significantly different from baseline (on PC1 from 876 to 1800 CE) are colored red.

To understand the interactions between forest Ca availability and N dynamics better, we examined patterns of sedimentary δ15N, an index of N availability. In general, in Mirror Lake, high sedimentary δ15N appears to indicate high N availability in the terrestrial ecosystem as enriched plant material, soil organic matter, and inorganic N enter the lake (17). From the beginning of the sedimentary record in ∼876–1800 CE, the decline in Ca concentrations in sediment was directly associated with increasing N availability. As Ca declined over this time period, sedimentary δ15N increased (r = –0.68; Fig. 2D and Fig. S3), suggesting increasing Ca limitation and declining N limitation. Alternatively, the different dominant sources of these two elements—the atmosphere for N and bedrock weathering for Ca—can explain the different trajectories recorded in Mirror Lake sediment since 876 CE The BSI also displays a long-term decline since 876 CE, reflecting increasingly acidic conditions (Fig. 1D). This long-term decline in BSI suggests that watershed soils were becoming increasingly base-poor through this time period. Thus, these data seem to indicate changes in nutrient availability for both N and Ca after only ∼10,000 y postglacial retreat and before the onset of significant human activities in the region.

Fig. S3.

Fig. S3.

Open in a new tab

Correlation between calcium concentration in sediments (Ca, expressed in g·kg−1) and sediment δ15N (black dots) tested with a model 2 regression (r = 0.45, red line) on the presettlement period (876–1800 CE). Gray lines are 95% confidence intervals.

Large Changes in Nutrient Concentration After Euro-American Settlement.

Sedimentary elemental concentrations changed markedly after Euro-American settlement in the region began ∼1770 CE As the Mirror Lake watershed experienced deforestation and agricultural activities (16), all 17 elements exhibited large changes. PCA reveals that the concentrations of 11 elements increased logistically after 1800 CE, reaching 95% of their upper asymptote by 1930 CE (Fig. 3A and Fig. S2). The identity of these elements indicates a signature of increased atmospheric pollution (e.g., increases in [Pb], [Zn]), an increase in erosion ([Ti]), and an increase in acidification (BSI; Fig. 3B), which may have increased leaching of Al from the forest floor and upper soil horizons. The increases in concentrations of heavy metals (e.g., Pb, Zn) are a general phenomenon in northern temperate lakes and have been documented in Mirror Lake previously, coupled with the increase in atmospheric deposition and the absence of change in lake productivity (16). In concert with these increases, six elements had lower concentrations in sediments after 1800 than before. Most notably, Si concentrations declined by 24% (Fig. 3C), likely reflecting the decline in biogenic silica concentration (16) rather than a diminution of rock weathering, as estimated mineral proportions indicate no increase in silicate-rich rock inputs (Table S4). In addition to the increase in heavy metal concentrations, the increases in erosion processes and acidification indicate that other pre-Industrial human activities such as forest clearance were strongly influencing the biogeochemistry of the lake and its watershed.

Fig. 3.

Fig. 3.

Open in a new tab

Elemental composition of sediment deposited from 1600 to 2000 CE, including post–Euro-American settlement and regional industrial activity. (A) First principal component axis (λ = 62%). (B) BSI equal to the ratio of sum of base cations (Na, Ca, Mg, K) divided by the sum of base cations and Al, no unit. Low values indicate acidic conditions. Also shown are Si concentrations (C) and Ca concentrations (D).

Despite these observed changes in elemental concentrations after 1800 CE, there is no record of a significant increase in Ca flux as recorded in the sediments relative to other elements (Fig. 3D). Although there are no records of direct human manipulation of Ca inputs in the Mirror Lake watershed, Euro-American land use changes affected Ca concentrations in the sediment. After 1800, there is a decline in sedimentary [Ca] associated with a decline in %C and organic matter inputs (Fig. S2) during extensive forest harvest, followed by a reestablishment of organic Ca inputs in the late 1800s during a period of reforestation. The sedimentary Ca concentration was further marked by increases in the early to mid-20th century, consistent with elevated Ca levels in streamwater flux as an effect of acid rain, and decreases after 1970, possibly reflecting the reduced deposition of acid rain and the depletion of Ca in soils (12). To contextualize the sedimentary elemental record with the long-term stream chemistry data from nearby Hubbard Brook watersheds, the onset of acid rain ∼1955 CE led to significant leaching of base cations—up to 50% of exchangeable Ca—from upland forests (12). This hydrologic loss pathway of dissolved Ca would not necessarily be recorded in lake sediments that are composed of organic and clastic material, and thus more closely reflect Ca concentration of organic material entering to the lake. However, the two records are complementary. On average, sedimentary Ca concentrations were 12% less after 1800 than during the period before 1800. Calcium concentrations were declining before Euro-American settlement, and it is difficult to assess the role of anthropogenic activities in the 20th century from these sedimentary data alone. Nevertheless, this decrease in sedimentary Ca from 876 CE to the beginning of the presettlement period reflects a large depletion of the total Ca pool from upper soil horizons, and this pool remains a particular concern, as it is larger than the entire exchangeable Ca pool in the soil and the forest floor pool (10, 25).

Conclusions

Our data indicate that nutrient availability on base-poor parent materials is sensitive to acidifying processes on millennial timescales. Long-term acidification of the system from natural processes has led to changes in both Ca and N dynamics, which may have affected nutrient limitation. Long-term acidification coincided with a steady decline of [Ca] and an increase of [Al] in lake sediments. Further, anthropogenic activities caused rapid (decadal) acidification in the 20th century, affecting many aspects of ecosystem biogeochemistry. Although Ca addition experiments at Hubbard Brook have demonstrated the reversibility of some of the modern Ca loss (26), there was also a decrease in N availability and no effect on Al concentration for 10 y after Ca application (8). Thus, at Hubbard Brook, additions of Ca alone are not sufficient to stop acidification on a long time scale. Regulations to minimize acidic precipitation are important to maintain, because mineral weathering rates are too low to neutralize additional acidic input to these forest ecosystems (12, 27), and currently there are minimal aerosol and precipitation inputs of Ca to sustain productivity (12). Thus, trajectories of increased acidification are likely to continue in this base-poor system.

Two major components of ongoing global change—increased atmospheric carbon dioxide concentrations and production of reactive N—both lead to ecosystem acidification. Global N additions may have already shifted soils from base cation buffering to Al3+ buffering (28). Nonetheless, Ca limitation is considered rare in temperate forest systems because of the relatively young age of most temperate-zone soils. However, our results demonstrate that shifting nutrient limitation can potentially occur within 10,000 y of ecosystem development. Lacustrine sedimentary records can be used to assess terrestrial nutrient dynamics, and due to the widespread occurrence of lakes, continuous depositional processes, and good chronological control, these records could help us build more generalized understandings of slow processes like the development of nutrient limitation in terrestrial ecosystems.

Materials and Methods

Study Site.

Mirror Lake is a 15-ha oligotrophic lake in the White Mountains of New Hampshire (16). The surrounding vegetation is northern hardwood forest containing deciduous and coniferous tree species. Common tree species in the watershed include Pinus strobus, Tsuga canadensis, Fagus grandifolia, Fraxinus americana, Acer saccharum, and Betula alleghaniensis. There are three stream inlets that drain the watershed and one outlet from the lake. Elevation in the watershed ranges from 213 m at lake level to 469 m at ridge tops.

Parent material in the watershed is primarily sandstone and mudrock deformed and metamorphosed from the Littleton Formation. The most abundant primary minerals are quartz, plagioclase, biotite, and sillimanite in a coarse-grained matrix. In the southwest corner of the watershed, there is an inclusion of the Kinsman Formation, a metamorphosed granitic material. The minerals oligoclase, andesite, potassium feldspar, biotite, and muscovite define this medium- to coarse-grained formation. Soils in most of the watershed are Spodosols, generally acidic (pH is less than 4.5), and infertile (16). At pH 5.5 and under, the concentration of dissolved aluminum increases and limits plant root growth (16). At pH 5.5 and less, soil water concentrations of aluminum increase, limiting plant root growth. More generally acid rain in this region has caused nutrient leaching in the soil, and a decline in forest productivity has been observed since 1980, demonstrating clear evidence of the effects of acid rain on forest ecosystems (11, 29).

Historical Context of the Study Site.

The first Euro-American settlers in the watershed were recorded around 1770 CE, and soil tillage for row-crop agriculture, grazing by sheep in pastures, and selective logging all occurred subsequently. The activities of the settlers and their descendants resulted in clear-cutting of most of the watershed as well as conversion of a substantial portion to pasture and arable land in the late 1800s and early 1900s. The subsequent abandonment of agricultural land led to replacement by natural succession and was coupled with forest plantations such as P. strobus in the 1930s (16). Since the 1980s, a dramatic decline in forest growth has been recorded, coupled with a large decline in calcium in various components of the ecosystem, including the forest floor. The causes of this ecosystem change come from several factors such as natural factors, disease, nutrient limitation, and pollution effects. The emergent hypothesis is that the deficiency of available calcium in the soil is a result of the effects of acid rain (29).

The collection of the sediment core and the construction of the age-depth model are described in ref. 17.

XRF Method and Calibration.

The 102 sediment samples were dried at 60 °C until no further mass loss was observed, ground to pass through a <40-µm sieve, and analyzed for the 11 major elements traditionally listed as oxides (Na, Mg, Al, Si, P, S, K, Ca, Ti, Mn, and Fe) and 6 trace elements (Cu, Zn, As, Pb, Rb, and Mo) by handheld wavelength-dispersive XRF spectrometry (Bruker Tracer III). The major elements were measured during 3 min per sample at 15 kV with a vacuum attached to the instrument, limiting the background noise. The trace elements were measured during 2 min per sample at 40 kV, with a yellow filter increasing the magnitude of the trace element spectra. Data were first recorded in counts and then were calibrated with the mudrock calibration (30), which is a combination of 200 standards of similar matrix (grain-size, homogeneity) and composition, to obtain quantitative data (Figs. S1 and S2 and Table S1). Of the 102 samples, 70 were measured three times for major and three times for trace elements. Because iron is quantified by both the major and trace element protocols, we chose the value with the smallest SD. Analytical error was ±2% for both major and trace elements.

BSI was calculated as the ratio of elements that form the basic cations (Na, Ca, Mg, and K) to those elements and Al (which produces an acidic cation, Al3+). Thus, this index reflects the balance between base and acidic cations in a sample. A decrease in this index will be due to an increase in the acidic cation and/or a decrease of base cations, and decreases in the index can be interpreted as increased acidification of the system.

To test the potential exchange of cations within the sediment core, the exchangeable Ca in the sediment as well as K, Mg, and Na were assessed (Table S2). The analysis was run on 2-g sediment samples, previously dried and powdered. Cations (Ca2+, K+, Mg2+, and Na+) were extracted by ammonium acetate solution (1 M, pH 7.0) following the method of Brown (31). A low-sodium filter paper was used, and the filtrate was analyzed by an Inductively Coupled Plasma Spectrometer, Model 720-ES Inductively Coupled Plasma Optical Emission Spectrometer (Varian Australia Pty Ltd.).

XRD Method.

Thirty of the 102 sediment samples from the Mirror Lake core were analyzed for mineral composition by XRD. Each sample powder was pressed into a sample holder and analyzed at 40 kV, 20 mA volume with a copper X-ray source, Kα of 1.54. Data are first presented as count peaks on a 2θ angle scale, from 2 to 80°. The mineral composition was assessed by the software HighScore Plus, version 4.0. Because we were interested in the dominant mineralogy in the sediment, we did not target trace minerals and we stopped the peak assignation when the mineral composition explained at least 70% of the peaks. Because XRD reveals the same mineral phases within the 30 samples tested but does not allow quantification of their proportion, we performed a normative calculation based on oxide element concentrations for five samples randomly distributed through the core. First, 10 samples, which were previously dried and powdered, were heated to 1000 °C for 2 h (Table S3). This procedure removed organic matter, composition water from clays, hydrated minerals, and the CO2 from carbonates. The resultant material thus reflected only crystalline-form minerals. On these samples, 12 element concentrations from XRF after a loss-on-ignition (LOI) at 1,000 °C were transformed to oxide-form concentrations (SiO2, TiO2, Al2O3, FeO, MnO, MgO, CaO, Na2O, K2O, P2O5, Cr2O3, and NiO) to match mineral composition according to the valence of the chemical formula of each mineral (Table S4).

Nitrogen Isotopes.

To understand the relationship among calcium (Ca), nitrogen (N), and carbon (C) better, we used previously published elemental and isotopic data for N that were produced at the Stable Isotope Lab at the University of Regina using standard methods on a Thermoquest (Finnigan-MAT) Delta Plus mass spectrometer interfaced with a Carlo Erba NC2500 elemental analyzer. Analytical error was <0.1‰ for δ15N (more information in ref. 17).

Statistical Analysis.

PCAs were conducted to assess the relationship among elements and to compare Ca trends to other nutrients over two different periods: (i) from 873 to 1800 CE and (ii) from 1600 to 2000 CE PCA from 873 to 1800 CE was rotated on correlations with a Varimax technique to strengthen contrasts. Examining the relationships among elements over the first time period (1), the first PCA axis describes those elements that are decreasing in concentration from 876 to 1800 CE (Ca, Mn, K) versus those that increase over this period (Al, Si; Fig. S2). A Kruskal–Wallis rank-sum test was used to test whether elemental concentrations during the seven identified events in the first period differed significantly from concentrations at other times. A breakpoint analysis was used to identify (i) the overall period of the occurrence of the seven events during period 1 and (ii) the date when the 11 elements displaying an increase in concentration during period 2 reached 95% of their upper asymptote. All statistics were computed in R v. 3.1.2 (32).

Acknowledgments

We thank Wyatt Oswald, Peter Leavitt, Bruce Kaiser, Pamela Kempton, Colleen Barbe, Emily Mellicant, Matthew Brueseke, Justin Hynicka, and Paul Bukaveckas for discussion and assistance. We are grateful for support from the Novus Research Coordinating Network (Grant NSF-DEB-1145815). Funding for long-term data on precipitation, stream water, and lake chemistry was provided by the National Science Foundation, including the Long Term Research in Environmental Biology and Long-Term Ecological Research programs, and The Andrew W. Mellon Foundation. The data reported in this paper are tabulated in Supporting Information. This is a contribution of the Hubbard Brook Ecosystem Study. The Hubbard Brook Experimental Forest is administered by the USDA Forest Service, Northern Experiment Station.

Footnotes

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1604909113/-/DCSupplemental.

References

  • 1.Gruber N, Galloway JN. An Earth-system perspective of the global nitrogen cycle. Nature. 2008;451(7176):293–296. doi: 10.1038/nature06592. [DOI] [PubMed] [Google Scholar]
  • 2.Porder S, Chadwick OA. Climate and soil-age constraints on nutrient uplift and retention by plants. Ecology. 2009;90(3):623–636. doi: 10.1890/07-1739.1. [DOI] [PubMed] [Google Scholar]
  • 3.Hedin LO, Vitousek PM, Matson PA. Nutrient losses over four million years of tropical forest development. Ecology. 2003;84(9):2231–2255. [Google Scholar]
  • 4.Newman GS, Hart SC. Shifting soil resource limitations and ecosystem retrogression across a three million year semi-arid substrate age gradient. Biogeochemistry. 2015;124(1-3):177–186. [Google Scholar]
  • 5.Menge DNL, Hedin LO, Pacala SW. Nitrogen and phosphorus limitation over long-term ecosystem development in terrestrial ecosystems. PLoS One. 2012;7(8):e42045–e42045. doi: 10.1371/journal.pone.0042045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Johnson MO, Gloor M, Kirkby MJ, Lloyd J. Insights into biogeochemical cycling from a soil evolution model and long-term chronosequences. Biogeosciences Discuss. 2014;11(4):5811–5868. [Google Scholar]
  • 7.Balaria A, Johnson CE, Groffman PM. Effects of calcium treatment on forest floor organic matter composition along an elevation gradient. Can J Res. 2014;44(8):969–976. [Google Scholar]
  • 8.Battles JJ, Fahey TJ, Driscoll CT, Jr, Blum JD, Johnson CE. Restoring soil calcium reverses forest decline. Environ Sci Technol Lett. 2013;1(1):15–19. [Google Scholar]
  • 9.Mainwaring DB, Maguire DA, Perakis SS. Three-year growth response of young Douglas-fir to nitrogen, calcium, phosphorus, and blended fertilizers in Oregon and Washington. For Ecol Manage. 2014;327(2014):178–188. [Google Scholar]
  • 10.Likens GE, et al. The biogeochemistry of calcium at Hubbard Brook. Biogeochemistry. 1998;41(2):89–173. [Google Scholar]
  • 11.Likens GE, Wright RF, Galloway JN, Butler TJ. Acid rain. Sci Am. 1979;241(4):43–51. [Google Scholar]
  • 12.Likens GE. Biogeochemistry of a Forested Ecosystem. 3rd Ed Springer Science & Business Media; New York: 2013. [Google Scholar]
  • 13.Odén S. The Acidification of Air and Precipitation and Its Consequences on the Natural Environment. State National Science Research Council; Washington, DC: 1968. [Google Scholar]
  • 14.Brady NC, Weil RR. The Nature and Properties of Soils. Prentice-Hall Inc.; Upper Saddle River, NJ: 1996. [Google Scholar]
  • 15.Oldfield F, et al. Terrestrial and aquatic ecosystem responses to late Holocene climate change recorded in the sediments of Lochan Uaine, Cairngorms, Scotland. Quat Sci Rev. 2010;29(7):1040–1054. [Google Scholar]
  • 16.Likens GE. An Ecosystem Approach to Aquatic Ecology: Mirror Lake and Its Environment. Springer-Verlag; New York: 1985. [Google Scholar]
  • 17.McLauchlan KK, Craine JM, Oswald WW, Leavitt PR, Likens GE. Changes in nitrogen cycling during the past century in a northern hardwood forest. Proc Natl Acad Sci USA. 2007;104(18):7466–7470. doi: 10.1073/pnas.0701779104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Likens GE, Driscoll CT, Buso DC. Long-term effects of acid rain: Response and recovery of a forest ecosystem. Sci Pap Ed. 1996;272(5259):244–245. [Google Scholar]
  • 19.Kylander ME, Ampel L, Wohlfarth B, Veres D. High-resolution X-ray fluorescence core scanning analysis of Les Echets (France) sedimentary sequence: New insights from chemical proxies. J Quat Sci. 2011;26(1):109–117. [Google Scholar]
  • 20.Kaushal S, Binford M. Relationship between C:N ratios of lake sediments, organic matter sources, and historical deforestation in Lake Pleasant, Massachusetts, USA. J Paleolimnol. 1999;22(4):439–442. [Google Scholar]
  • 21.Matthews JA, Briffa KR. The “Little Ice Age”: Re-evaluation of an evolving concept. Geogr Ann, Ser A. 2005;87(1):17–36. [Google Scholar]
  • 22.Houlton BZ, et al. Nitrogen dynamics in ice storm-damaged forest ecosystems: Implications for nitrogen limitation theory. Ecosystems (N Y) 2003;6(5):431–443. [Google Scholar]
  • 23.Markewitz D, Richter DD, Allen HL, Urrego JB. Three decades of observed soil acidification in the Calhoun Experimental Forest: Has acid rain made a difference? Soil Sci Soc Am J. 1998;62(5):1428–1439. [Google Scholar]
  • 24.Perakis SS, et al. Forest calcium depletion and biotic retention along a soil nitrogen gradient. Ecol Appl. 2013;23(8):1947–1961. doi: 10.1890/12-2204.1. [DOI] [PubMed] [Google Scholar]
  • 25.Blum JD, et al. Mycorrhizal weathering of apatite as an important calcium source in base-poor forest ecosystems. Nature. 2002;417(6890):729–731. doi: 10.1038/nature00793. [DOI] [PubMed] [Google Scholar]
  • 26.Johnson CE, Driscoll CT, Blum JD, Fahey TJ, Battles JJ. Soil chemical dynamics after calcium silicate addition to a Northern hardwood forest. Soil Sci Soc Am J. 2014;78(4):1458–1468. [Google Scholar]
  • 27.Schaller M, Blum JD, Hamburg SP, Vadeboncoeur MA. Spatial variability of long-term chemical weathering rates in the White Mountains, New Hampshire, USA. Geoderma. 2010;154(3):294–301. [Google Scholar]
  • 28.Tian D, Niu S. A global analysis of soil acidification caused by nitrogen addition. Environ Res Lett. 2015;10(2):24019. [Google Scholar]
  • 29.Winter TC, Likens GE. Mirror Lake: Interactions Among Air, Land, and Water. Univ of California Press; Berkeley, CA: 2009. [Google Scholar]
  • 30.Rowe H, Hughes N, Robinson K. The quantification and application of handheld energy-dispersive x-ray fluorescence (ED-XRF) in mudrock chemostratigraphy and geochemistry. Chem Geol. 2012;324-325(2012):122–131. [Google Scholar]
  • 31.Brown JR. Recommended Chemical Soil Test Procedures for the North Central Region. Missouri Agricultural Experiment Station, University of Missouri–Columbia; Columbia, MO: 1998. [Google Scholar]
  • 32.R Core Team . R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing; Vienna, Austria: 2014. [Google Scholar]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES