Red rains as major contributors of nutrients and alkalinity to terrestrial ecosystems at Montseny (NE Spain)

O sis, 6: 215-229 (1991) Red ains as majo con ibu o s o nu ien s and alkalini y o e es ial ecosys ems a Mon seny (NE Spain) Anna Á ila and Fe an Roda Cen e de Rece ca Ecolbgica i Aplicacions Fo es als. Uni e si a Au bnoma de Ba celona. 08193 Bella e a. Spain. Keywo ds: acid ain, a mosphe ic deposi ion, calcium, elemen inpu s, Saha an dus . Abs ac . Du ing wo yea s (1983-1985), 8 ed ains we e collec ed in he Mon seny moun ains. The suspended ma e con ained in ed ains has i s sou ce in he Saha a dese and co esponds o ine sil s. Due o he calci e con en o dus esponsible o he ed ains, hese ains a e alkaline wi h a olume weigh ed mean pH o 7.7 e sus 4.7 o non- ed ains. Es ima ed HCO; alkalini y inpu due o ed ains is 0.21 keqlhaly and o al annual HCO; alkalini y inpu is 0.24 keqlhaly . To ha igu e, one has o add a leas 0.24 keqlhaly due o he alkalini y o calci e con ained in ed dus . Fo he same pe iod, hyd ogen inpu in bulk deposi ion amoun s o 0.19 keqlhaly . These esul s indica e ha acidi y due o non- ed ain espisodes is neu alized on an annual basis, wi h ed ains being he majo neu alizing agen . Ion concen a ions in ed ains a e high, which makes hem an impo an sou ce o nu ien s e en hough hey only accoun o 5.4% o annual p eci i a ion. Red ains a e pa icula ly en iched wi h ca2+ ha ing a P+ olume-weigh ed mean Ca concen a ion 14 imes highe han non- ed ains and deli e ing 46% o he annual inpu s o soluble ca2+ in bulk p ecipi a ion a Mon seny. Resum. Pape de les pluges de ang en I'apo a mos e ic de nu ien s i d'alcalini a al Mon seny. Du an dos anys (1983-1985) es a en ecolli 8 pluges de ang a La Cas anya (Mon seny). La ma k ia pa iculada de les pluges de ang s'o igina al dese del Saha a, i co espon a llims ins. Degu al con ingu de calci a d'aques s llims, les pluges de ang són alcalines, amb un pH mig de 7.7, en on de 4.7 pe la es a de les pluges. L'en ada anual es imada d'alcalini a en o ma de bica bona dissol a les pluges de ang és de 0.21 keqlhalany, d'un o al es ima de 0.24 keqlhalany pe o es les pluges. A aques a xi a cal a egi almenys 0.24 keqlhalany degu al con- ingu de calci a a la ase sblida de les pluges de ang. Du an el ma eix pe iode, I'en ada d'hi- d ogenions a la deposició global és de 0.19 keqlhalany. Aques s esul a s indiquen que, a escala anual, les pluges de ang apo en su icien alcalini a com pe neu ali za I'acidesa de la es a de les pluges. Les concen acions ibniques són ele ades a les pluges de ang, i aques es s6n pe an , un ac o impo an en la deposició a mos k ica de nu ien s, enca a que ep esen en en p omig només un 5.4% de la p ecipi ació anual. Les pluges de ang són pa icula men iques en calci: la se a concen ació mi jana ponde ada d'aques elemen és 14 egades més g an que la de la es a de les pluges, i con ibueixen amb un 46% a les en ades anuals de calci en la deposició global del Mon seny. In oduc ion Acid ain has been he subjec o a as esea ch e o a emp ing o cla i y he ole o incoming a mosphe ic acidi y in he al e a ion o biogeochemical cycles in e es ial and aqua ic ecosys ems (Ad iano & Ha as 1989, D ablos & Tollan 1980, Hu chinson & Ha as 1980, Reuss & Johnson 1986). Ne e heless, he e a e ela i ely ew s udies s essing he ac ha pa icula a eas equen ly ecei e ain e en s o high pH (PH well abo e 5.6, he pH alue o pu e wa e equilib a ed wi h a mosphe ic CO*). Alkalini y esponsible o his pH inc ease is mainly due o dissolu ion o alkaline componen s o aeolian soil dus . Khemani e al. (1987) ound in India ha highly alkaline pH alues in cloud and ainwa e we e associa ed o high concen a ions o dus pa icula es in he a mosphe e. Bua -Ména d e al. (1974) epo ed obse a ions o Ca2' en ichmen in ma ine ae osols o e he A lan ic wich we e a ibu ed o soil dus om he Saha a and he Kalaha i dese s. Loye-Pilo e al. (1986) obse ed in Co sica ha ain wi h high concen a ions o dus had an impo an Ca2' en ichmen and Gigh pH alues. Dus collec ed in Co sica was aced back o o igina e in he Saha a dese . As a consequence, geog aphic a eas in he icini y o a id o denuda ed egions a e subjec o occasional episodes in which he a mosphe e is loaded wi h Ca2+-MgZ ae osols o high neu alizing capaci y. The Medi e anean is a pa icula example o such an a ea, being in luenced by dus anspo om No h A ica, as has been widely epo ed (Büche & Lucas 1984, Ches e e al. 1984, Loye-Pilo e al. 1986, Ma in-Vide & Mo eno 1985). Be game i e al. (1989) ha e iden i ied se e a1 seasonal pa e ns o a mosphe ic ci cula ion ha p oduce anspo o A ican dus o e he Medi e anean. Saha an dus anspo o sou he n Eu ope is e ealed by he occasional occu ence o episodes o so-called ed ain o ed snow. A Ba celona, he a e age annual equency o ed ain e en s was 2.8 o he pe iod 1944-1974, being concen a ed mos ly in sp ing and summe (Ma in-Vide & Mo eno 1985). Red ain episodes ha e been s udied om a me eo ological o mine alogical poin o iew (Büche e al. 1983, Büche & Lucas 1972, 1975, Clemen e al. 1972), bu he chemis y o ed ains is poo ly documen ed. The pu pose o his pape is o desc ibe he chemis y o ed- ains collec ed in a u al a ea o NE Spain and o e alua e hei con ibu ion o he annual inpu o nu ien s and in he neu aliza ion o acidic deposi ion. To ha pu pose we p esen da a o a wo-yea pe iod o weekly sampling o bulk p ecipi a ion in he Mon seny moun ains, dis inguishing ain e en s con aining ed dus om e en s wi hou ed dus . S udy a ea The s udy si e is loca ed wi hin he Es ació Biolbgica de La Cas anya I RED RAINS AS CONTRIBUTORS OF NUTRIENTS AND ALKALINITY 217 (41°46'N, 2'21'E) in he Mon seny moun ains, a p edominan ly o es ed massi which lies 25-30 km om he Medi e anean coas in no heas Spain. Ag icul u a1 land lie o he eas and wes o he moun ain, i s nea es indus ial conu ba ion being ha o Ba celona, 40 km o he SSW. The li hology o Mon seny is mainly o med by low g ade me omo phic schis s and g anodio i es. The si e o p ecipi a ion sampling is loca ed in a o es clea ing nea he ou le o a small expe imen al wa e shed (TM9) wich is ou inely moni o ed o chemis y by ou labo a o y wi hin he scope o a b oade s udy o nu ien cycling in holm-oak (Que cus ilex L.) o es s (A ila 1988, A ila & Rodi 1988, Fe és e al. 1984, Roda 1983, Roda e al. 1990). The al i ude a he si e o p ecipi a ion sampling is 700 m a.s.1. Mean annual p ecipi a ion measu ed o 7 yea s a he si e is 873 mm. Mean annual ai empe a u e is es ima ed o be a ound 9°C. Vege a ion o he wa e shed and on su ounding slopes is a dense, undis u bed o es o holm oak. Me hods Bulk p ecipi a ion o analysis was sampled wi h 4 collec o s, each consis ing o a polye hylene unnel 19 cm in diame e connec ed by a loop o a 10 L polye hylene bo le, ollowing he Hubba d B ook design (Likens e al., 1977). Collec o s a e pe manen ly open o he a mosphe e, so he measu ed lux is bulk deposi ion which includes we deposi ion and g a i a o y sedimen a ion o pa icles. Weekly, ainwa e samples (and occasionally snow collec ed by a 63-cm diame e bucke ) we e collec ed, and collec o s we e eplaced by labo a o y cleaned ones. The cleaning p ocedu e consis ed o a de e gen wash o he disassembled pa s o he collec o , a inse wi h abundan ap wa e ollowed by soaking in 10% dilu ed comme cial HCl, and a inse wi h abundan dis illed wa e . Final conduc i i y o dis illed wa e un h ough he assembled collec o was checked o ensu e ha no con amina ion o HCl ocu ed. A plug o nylon mesh, also washed wi h dis illed wa e and checked o conduc i i y, was placed in he unnel neck o p e en insec s and la ge deb is om eaching he collec ing essel. In equen samples wi h insec s o deb is we e disca ded. Usually, collec o s we e bulked by pai s o gi e wo eplica es o analysis, bu o small ains (up o 5 mm) all collec o s we e bulked o gi e a single sample. Occasionally, he samples o he ou collec o s we e analyzed. The amoun o p ecipi a ion was measu ed by a s anda d ain-gauge a weekly in e als. The day o collec ion, samples we e aken o he labo a o y, whe e pH and conduc i i y we e measu ed a e equilib a ing he sample o he labo a o y empe a u e. pH was measu ed po en iome ically using a combined elec ode which was calib a ed wi h pH bu e s o 7 and 9, o 4 and 7 depending on he sample PH. Conduc i i y was measu ed a e calib a ion o he conduc ime e wi h a 100 pS/cm (a 20°C) solu ion. All ed ains we e il e ed excep o one (sample collec ed on 2 May 1985). Majo ions we e analyzed on ozen aliquo s. Sodium and po assium we e measu ed by a omic emission spec ome y, calcium and magnesium by a omic abso p ion spec ome y, and C1-, NO;, So; and NH4 ' by ion ch oma og aphy. Di e ences be ween collec o eplica es we e nea ly always less han 10%. In cases o highe di e ences, incohe en alues could be easily iden i ied and we e disca ded. P ecission and accu acy we e es ed in all analy ical uns and we e ound o be sa is ac o y (A ila 1988). Mean pH was calcula ed by ans o ming indi idual pH alues o hyd ogen concen a ion, weigh ing hese by p ecipi a ion olume and econ e ing o PH. We ha e no alkalini y da a o he s udy pe iod, so o ob ain a en a i e igu e we ha e es ima ed he HCO; concen a ions om he equilib ium eac ions o he ca bonic acid-ca bona e sys em: and me ging he wo equa ions, Knowing he pH o indi idual ain samples, HCO; can be ound om he abo e ela ionship. We ha e aken an a mosphe ic pC02 o 340 ppm as e lec ing ha o he ainwa e samples, and alues o kh and kl a labo a o y empe a u e (20°C): kh=10-'.41 and kl= 10-638 (D e e 1982). F om ha we ob ain: [HCO;] =5.13/[H'] (concen a ions in peq L-') This ela ionship was applied o he H' concen a ion o each sample, and HCOJ es ima ed alues we e hus ob ained. These we e hen weigh ed by olume o ob ain a HCO; olume-weigh ed mean concen a ion. Resul s and discussion Resul s epo ed he e e e o he wo-yea pe iod om 1 Augus 1983 o 31 July 1985. Du ing his pe iod 87 bulk p ecipi a ion weekly samples we e collec ed, 8 o which we e iden i ied as ed ains by isual obse a ion o a colou ed solid esidue on unnel walls. O he 8 ed ains, 3 occu ed in he i s yea ( hose collec ed on 16 Ap il 1984, 6 May 1984 and 25 June 1984), and 5 in he second (collec ed on 12 No embe 1984, 25 Ap il 1985, 2 May 1985, 21 June 1985 and 19 July 1985). The ed ain on 19 July 1985 was e y RED RAINS AS CONTRIBUTORS OF NUTRIENTS AND ALKALINITY 219 small (0.1 mm) gi ing no enough olume o chemical analyses, and ha o 16 Ap il 1984 (1.3 mm) was disca ded because o p obable con amina ion; he e o e, only he chemis y o he emaining 6 pe iods o ed ain will be conside ed he e (Table 1). As hey we e weekly samples, hey may ha e included mo e han one indi idual ed ain because ed ains end o be con iguous (Ma in-Vide and Mo eno 1985). The di e ences in p ecipi a ion chemis y ound in his s udy be ween ed and non- ed ains a e hus conse a i e es ima es. Amoun o p ecipi a ion The annual amoun o p ecipi a ion a TM9 was 1052.5 mm and 873.0 mm o 1983-1984 and 1984-1985 espec i ely. The wo yea mean was 962.8 mm. The 8 weekly samples con aining ed ains deli e ed a o al o 103.9 mm, o 52.0 mm y -l. This amoun s o 5.4% o he mean p ecipi a ion ecei ed a he si e in his pe iod. The e was a nine old di e ence in he amoun o ed Table 1. Volume-weigh ed mean concen a ion in ed ains and non- ed ains o he wo-yea pe iod 1983-1985 a Mon seny. Conduc i i y in yS cm-I a 20 OC; concen- a ions in yeq L-'; p ecipi a ion in mm y -l. Non- ed Red non- ed Ion Red ain ain ain Conduc i i y 73.3 18.5 4.0 PH 7.76 4.68 - H+ 0.02 21.1 10-~ Na' 98.6 21.6 4.6 K+ 12.3 2.43 5.1 CaZc 549 38.5 14.3 Mg2' 47.7 9.5 5.0 Mg* 26.7 4.6 5.8 NH: 19.7 26.6 0.7 NO3- 36.1 19.0 1.9 SO:- 140 50.5 2.8 SO: 128 47.9 2.7 C1- 126 26.0 4.8 HCO~-" 416 3.3 126 P ecipi a ion 51.2~ 908' 0.06 n 6 72 - * M~~' and SO:- concen a ions a e co ec ing o sea-sal ae osol wi h Na' as e- e ence ion. " es ima ed om equilib iuin eac ions o he ca bona e sys em, ha ing conside ed a cons an pCOz a 340 ppm and kl and kh alues a 20 "C. non-analyzed and non-conside ed e en s o alled a u he 1.4 mm y -I. non-analyzed e en s o alled a u he 3 mm y -l. ains be ween he wo yea s: 10.6 mm in 1983-1984 s. 93.3 mm in 1984-1985, accoun ing espec i ely o 1% and 10% o he annual p ecipi a ion. The g ea e amoun o ed ain eco ded in he second yea esul ed om wo la ge- olume e en s (50 mm collec ed on 12 No embe 1984 and 31 mm collec ed on 25 Ap il 1985). These wo s o ms accoun oge he o 88% o he o al p ecipi a ion o ed ains in he second yea . The e o e, he e is a hgh in e annual a iabili y in ed ain amoun o p ecipi a ion depending on he p ecipi a ion deli e ed by pa icula e en s. The amoun o p ecipi a ion pe pe iod does no di e signi ican ly be ween pe iods o ed ain and pe iods o non- ed ain ( es a e loga i hmic ans o ma ion, P>0.05). Solu es in ed ain The olume weigh ed mean concen a ions o majo ionic componen s in ed ain and non- ed ain samples collec ed a TM9 a e gi en in Table 1. Excep o NH4, which is he sole ion o ha e simila concen a ions in bo h ype o ains, and o H' which dec eases in ed ains, all ions ha e much g ea e concen a ions in ed ains. Di e ences a e signi ican o all ions (P<0.005 o conduc i i y, PH, Na', K', Ca2', Mg2', SO:' and C1-; P=0.09 o NO;) excep o NH: (P=0.22) when aplying a co a iance analysis wi h amoun o p ecipi a ion as a co a ia e. Sodium, K', Mg2' and C1- means in ed ains a e g ea e by a ac o o 4 o 5 han non- ed ains, while o NO; and SO:- his ac o is 2 and 3 espec i ely (Table 1). Howe e , he ion wi h he highes di e ence is Ca2', whose mean concen a ion in ed ains is 14 imes ha o non- ed ains. The abo e men ioned highe ion concen a ions a e e lec ed in conduc i i y, which consequen ly is g ea e in ed ains. As s a ed be o e, H' concen a ion is e y low in ed ains, gi ing a olume-weigh ed mean pH o 7.76. This alue con as s s ongly wi h he mean pH o non- ed ains o 4.68 (Table 1). The empo al a ia ion o pH in bulk deposi ion a TM9 is p esen ed in Figu e 1, ed ains di e ing om non- ed ains because o hei highe pH alues. Red ains a e cha ac e is ically e y concen a ed in all ions, bu hey show also a high a iabili y in concen a ions, wi h coe icien s o a ia ion (%) o 35.2 o conduc i i y, 56.0 o Na', 56.1 o K', 42.9 o Ca2', 25.8 o Mg2', 68.1 o NH:, 49.7 o NO;, 55.8 o SO:- and 54.0 o C1-, o he 6 ed ain e en s a TM9. Red ains could be dis inc ly di e en ia ed om non- ed ains by means o a p incipal componen analysis appyied o he ma ix o he 9 analyzed a iables (PH, Na', K', Ca2', Mg2', NH:, NO'!, S04 and C1-) and 85 alid weekly samples. The g aphical exp ession o he analysis is show in Figu e 2, whe e he h ee i s componen s accoun o 91.7% o ainwa e ionic a ia ion. Red- ain samples o m a dis inc g oup when indi idual sco es a e ep esen ed o he i s componen (iden i ied as mine aliza ion o ichness in Ca2', Mg2', K', S042- and NO i) and he hi d componen (de ined by pH RED RAINS AS CONTRIBUTORS OF NUTRIENTS AND ALKALINI= 221 Figu e 1. Tempo al a ia ion o bulk p ecipi a ion pH a TM9. Red ains a e ep esen ed by open ci cles. Figu e 2. Resul s o p incipal cqmponen analysis o bulk p ecipi a ion chemis y o all he weekly samples; de ined by hei pH and majo ion concen a ions. Sco es o indi idual ain samples de ined by he i s and hi d componen s. Red ains a e ep esen ed by illed ci cles. and NH4 and iden i ied as acidi y-alkalini y o ain). Red ains a e di e en ia ed om he es o ains because o hei high mine aliza ion and high pH alues (Fig. 2) as is e lec ed by esul s in Table 1. A his momen , i is wo h s essing ha ed- ains we e de ined by usual ecogni ion o colou ed ma ks o deposi ed dus . I is possible ha , because o he weekly sampling schedule, some ed ains alling a he beginning o he pe isd and ollowed by non- ed ains we e missclassi ied, o con e sely, ed ains a he end o he pe iod could con ain non- ed ains collec ed a he beginning o he pe iod. In bo h cases, he mixing would p oduce a dilu ion o he usual high concen a ions o ed ains, and p obably, a dec ease in he sample PH. These samples, and o he s in luenced by dus ae osols bu no lea ing isual ma ks, would all in he icini y o he ed ain g oup in he p incipal componen analysis, and hey migh be as well some o he poin s o ela i ely high pH in Figu e 1 which a e no signaled as ed ains. No only do ed ains di e widely om non- ed ains in hei ionic concen a ions, bu also hei ela i e ionic p opo ions a e qui e dis inc . Pa icula ly in e es ing is he a io o non-sea sal Ca2' o non-sea sal M~~+ (also called excess Ca and excess Mg and ep esen ed by Ca* and Mg*). In ac , as only 2% o Ca2' in bulk p ecipi a ion a Mon seny comes om sea sal ae osols, we ha e compu ed excess ca2' only o consis ency wi h ~g~+. A Mon seny, non- ed ains ha e a ema kably cons an Ca*/Mg* a io o abou 8, on an equi alen basis. This gi es a high linea co ela ion be ween Ca* and Mg* (Fig. 3). This cons ancy o he Ca*/Mg* a io implies ha ei he (i) all majo con inen al sou ces o soluble Ca2' and M~~' in non- ed ains a Mon seny yield a Ca*/Mg* a io o abou . 8, o (ii) di e en sou ces wi h a ying p opo ions o soluble Ca2' and M~~+ always combine in such a manne o gi e he obsemed Ca*/Mg* a io. In con as , ed ains a e clea ly se apa om non- ed ains in a sca e g am o Ca* e sus Mg* (Fig. 3). In p opo ion o Mg*, ed ains a e much iche in ca2' han non- ed ains, hei Ca*/Mg* a io be ween olume-weig ed means ising o 21 on an equi alen basis. This sha p di e ence in ionic p opo ions p o ides u he e idence ha he sou ce a eas o ai -bo ne ma e a e di e en o ed and non- ed ains. Two o he s udies o bulk deposi ion in Ca alonia ha e dis inguished be ween ed and non ed ain e en s. One was ca ied a a nea by s a ion named TM5 which is loca ed in culmina1 hea hlands in he Mon seny and is only 1.35 km away om TM9. The o he one was ca ied a P ades (Ta agona), 120 km o he WSW om ou s udy si e. A TM5, en ed ains we e iden i ied by isual obse a ion o eddish ma ks on collec o s be ween Janua y 1982 and Decembe 1984. O hese, only he a i hme ic mean concen a ion is a ailable (Belillas 1989) which is gi en in Table 2. Fo compa ison, he a i hme ic mean concen a ion o he 6 ed ain e en s collec ed a TM9 is gi en in Table 2. Red ains in bo h s a ions show simila ea u es in: 1) basic pH and 2) e y high ionic concen a ions, pa icula y - - --. . - . . - - -. . .. - - RED RAINS AS CONTRIBUTORS OF NUTRIENTS AND ALKALINITY 223 1 Excess Mg (peq/L) Figu e 3. Sca e g am o non-sea sal ('excess') Ca concen a ion e sus non-sea sal Mg concen a ion in bulk deposi ion a TM9. Red ains a e ep esen ed by illed ci cles. Poin s no ma ked as ed ains bu wich a e well abo e he eg ession line could be ed ains no iden i ied as such in he ield. Reg ession has been compu ed only wi h non- ed ains. Table 2. A i hme ic mean chemis y o ed ains in h ee locali ies o no heas e n Spain. Concen a ions in peq L-'. Conduc i i y in pS cm-' a 20 OC. P ecipi a ion in mm. Mon seny P adesc Va iable TM9" TMS~ Bosc Poble PH 7.66 6.24 7.07 Na 75.3 148 57.7 K' 16.6 26.1 8.9 CaZ 708 750 199 MgZ 53.1 91.2 20.7 NH: 25.3 62.7 24.9 NO; 52.2 110 22.6 SO,'- 21 1 317 67.2 C1- 97.1 178 53.3 HCO; - 291d 201 P ecipi a ion 102.3 281 312 n 6 10 6 " This s udy. Augus 1983 o July 1985. Belillas 1989. Janua y 1982 o Decembe 1984. ' Piñol 1990. Sep embe 1986 o June 1988. Mean o 5 samples.