{"id":2719,"date":"2016-12-06T14:28:19","date_gmt":"2016-12-06T14:28:19","guid":{"rendered":"http:\/\/www.vtei.cz\/?p=2719"},"modified":"2024-07-16T11:49:41","modified_gmt":"2024-07-16T10:49:41","slug":"characteristics-determining-the-stability-of-woody-debris-on-the-example-of-morava-river-in-the-litovelske-pomoravi","status":"publish","type":"post","link":"https:\/\/www.vtei.cz\/en\/2016\/12\/characteristics-determining-the-stability-of-woody-debris-on-the-example-of-morava-river-in-the-litovelske-pomoravi\/","title":{"rendered":"Characteristics determining the stability of woody debris on the example of Morava River in the Litovelsk\u00e9 Pomorav\u00ed"},"content":{"rendered":"

<\/i> This article is available in Czech only. For translation or more information on this topic, please contact author.<\/h4>\n

 <\/p>\n

Souhrn<\/h2>\n

\u0158\u00ed\u010dn\u00ed d\u0159evo je v\u00fdznamnou sou\u010d\u00e1st\u00ed vodn\u00edch tok\u016f, ve kter\u00fdch ovliv\u0148uje hydraulick\u00e9, morfologick\u00e9 i\u00a0biologick\u00e9 vlastnosti. Pro pot\u0159eby spr\u00e1vy vodn\u00edch tok\u016f a\u00a0omezen\u00ed povod\u0148ov\u00fdch rizik je d\u016fle\u017eit\u00e9 zn\u00e1t faktory, kter\u00e9 zvy\u0161uj\u00ed stabilitu d\u0159eva a\u00a0br\u00e1n\u00ed jeho odplaven\u00ed p\u0159i zv\u00fd\u0161en\u00fdch pr\u016ftoc\u00edch. Na z\u00e1klad\u011b anal\u00fdzy vlastnost\u00ed 160 kus\u016f \u0159\u00ed\u010dn\u00edho d\u0159eva v\u00a0koryt\u011b Moravy v\u00a0CHKO Litovelsk\u00e9 Pomorav\u00ed byly zji\u0161t\u011bny charakteristiky v\u00fdznamn\u00e9 pro stabilitu \u0159\u00ed\u010dn\u00edho d\u0159eva v\u00a0t\u00e9to oblasti. Anal\u00fdza byla provedena logistick\u00fdm regresn\u00edm modelem a\u00a0jako statisticky v\u00fdznamn\u00e9 identifikovala n\u00e1sleduj\u00edc\u00ed charakteristiky: d\u00e9lka kmene, v\u00fd\u0161ka ulo\u017een\u00ed v\u016f\u010di b\u011b\u017en\u00e9 hladin\u011b, m\u00edra zazemn\u011bn\u00ed, p\u0159\u00edtomnost ko\u0159en\u016f a\u00a0rozli\u0161en\u00ed, zda jde o\u00a0kmen \u017eiv\u00fd \u010di odum\u0159el\u00fd. Aplikac\u00ed modelu na konkr\u00e9tn\u00ed kus lze vypo\u010d\u00edtat modelovanou pravd\u011bpodobnost stability. Pokud je z\u00a0d\u016fvod\u016f zachov\u00e1n\u00ed p\u0159irozen\u00e9ho stavu koryta po\u017eadov\u00e1no ponech\u00e1n\u00ed stabiln\u00edch kus\u016f \u0159\u00ed\u010dn\u00edho d\u0159eva, m\u011bly by b\u00fdt tyto charakteristiky respektov\u00e1ny. Modelovan\u00e9 pravd\u011bpodobnosti korelovaly se subjektivn\u00edm odhadem stability s\u00a0koeficientem determinace R2<\/sup> = 0,57.<\/p>\n\"chalupska%e2%95%a0u_slat%e2%95%a0i_s%e2%95%a0iumava_1_uprava\"<\/a>\n

\u00davod<\/h2>\n

Term\u00ednem \u0159\u00ed\u010dn\u00ed d\u0159evo ozna\u010dujeme stromy a\u00a0jejich \u010d\u00e1sti, kter\u00e9 se p\u0159irozen\u00fdmi zp\u016fsoby nebo \u010dinnost\u00ed \u010dlov\u011bka dostaly do koryta vodn\u00edho toku a\u00a0vstupuj\u00ed zde do interakce s\u00a0vodn\u00edm proudem [1]. \u0158\u00ed\u010dn\u00ed d\u0159evo ovliv\u0148uje hydraulick\u00e9 podm\u00ednky v\u00a0koryt\u011b, formuje jeho tvar a\u00a0vytv\u00e1\u0159\u00ed pestrou nab\u00eddku stanovi\u0161\u0165 pro vodn\u00ed organismy [2]. D\u0159evo je dynamickou slo\u017ekou \u0159\u00ed\u010dn\u00edch krajin. Vstupuje do koryta r\u016fzn\u00fdmi p\u0159\u00edsunov\u00fdmi mechanismy, m\u011bn\u00ed svou polohu, je un\u00e1\u0161eno povodn\u011bmi a\u00a0znovu zachycov\u00e1no v\u00a0jin\u00fdch \u00fasec\u00edch koryta. V\u00a0pr\u016fb\u011bhu \u010dasu m\u016f\u017ee zcela podlehnout rozkladu, m\u016f\u017ee b\u00fdt vyplaveno do nivy nebo b\u00fdt poh\u0159beno v\u00a0\u0159\u00ed\u010dn\u00edch n\u00e1plavech. Pohyblivosti \u0159\u00ed\u010dn\u00edho d\u0159eva p\u0159i povodn\u00edch je v\u011bnov\u00e1na pozornost mimo jin\u00e9 z\u00a0d\u016fvodu \u00fadr\u017eby vodn\u00edch tok\u016f a\u00a0prevence povod\u0148ov\u00fdch rizik.<\/p>\n

Z\u00a0minul\u00fdch desetilet\u00ed existuje \u0159ada v\u00fdzkumn\u00fdch prac\u00ed, kter\u00e9 se mobilitou \u0159\u00ed\u010dn\u00edho d\u0159eva zab\u00fdvaj\u00ed. Deskriptivn\u00ed p\u0159\u00edstupy vych\u00e1zej\u00ed z\u00a0popisu mno\u017estv\u00ed a\u00a0charakteru d\u0159eva na lokalit\u011b, tvaru a\u00a0v\u00fdskytu akumulac\u00ed naplaven\u00e9ho d\u0159eva i\u00a0stabiln\u00edch kus\u016f v\u00a0jejich z\u00e1kladech [3\u20135]. Experiment\u00e1ln\u00ed studie zna\u010den\u00e9ho d\u0159eva v\u00a0ter\u00e9nu umo\u017enily posouzen\u00ed vlivu konkr\u00e9tn\u00edch pr\u016ftok\u016f v\u00a0definovan\u00fdch \u010dasov\u00fdch \u00fasec\u00edch. Zna\u010den\u00e9 kusy mohly b\u00fdt dohled\u00e1ny a\u00a0byly zkoum\u00e1ny parametry ovliv\u0148uj\u00edc\u00ed jejich transport [6\u201312]. Teorie o\u00a0vlastnostech mobilizace, transportu a\u00a0zachyt\u00e1v\u00e1n\u00ed idealizovan\u00fdch model\u016f \u0159\u00ed\u010dn\u00edho d\u0159eva byly ov\u011b\u0159ov\u00e1ny experiment\u00e1ln\u011b v\u00a0hydraulick\u00fdch laborato\u0159\u00edch i\u00a0v\u00a0ter\u00e9nu [13\u201315]. V\u00a0posledn\u00edch letech jsou sestavov\u00e1ny matematick\u00e9 modely popisuj\u00edc\u00ed dynamiku d\u0159eva v\u00a0r\u00e1mci dlouh\u00fdch \u00fasek\u016f tok\u016f nebo povod\u00ed. \u010cast\u00e9 jsou tak\u00e9 zcela praktick\u00e9 studie zab\u00fdvaj\u00edc\u00ed se n\u00e1chylnost\u00ed in\u017een\u00fdrsk\u00fdch staveb k\u00a0zachyt\u00e1v\u00e1n\u00ed vodou un\u00e1\u0161en\u00e9ho plovouc\u00edho materi\u00e1lu [16\u201318].<\/p>\n

V\u00fdzkumy proveden\u00e9 v\u00a0minulosti uk\u00e1zaly, \u017ee krom\u011b pr\u016ftokov\u00fdch charakteristik vodn\u00edho toku je stabilita \u0159\u00ed\u010dn\u00edho d\u0159eva ovliv\u0148ov\u00e1na p\u0159edev\u0161\u00edm jeho rozm\u011bry relativn\u011b k\u00a0\u0161\u00ed\u0159ce a\u00a0hloubce koryta. Stabilitu tak\u00e9 v\u00fdznamn\u011b zvy\u0161uj\u00ed zachoval\u00e9 ko\u0159eny, ko\u0159enov\u00fd bal a\u00a0v\u011btve. Poloha kmene v\u00a0koryt\u011b a\u00a0jeho orientace v\u016f\u010di proud\u011bn\u00ed jsou t\u00e9\u017e v\u00fdznamn\u00e9 parametry, kter\u00e9 nav\u00edc souvisej\u00ed se stabiliza\u010dn\u00edmi mechanismy b\u0159ehov\u00e9 vegetace a\u00a0stabilitou rozm\u011brn\u00fdch akumulac\u00ed \u0159\u00ed\u010dn\u00edho d\u0159eva. A\u010dkoliv jsou uveden\u00e9 poznatky obecn\u011b platn\u00e9, ka\u017ed\u00e1 lokalita je specifick\u00e1 sv\u00fdmi p\u0159\u00edrodn\u00edmi podm\u00ednkami, kter\u00e9 je t\u0159eba pro praktick\u00fd management \u0159\u00ed\u010dn\u00edho d\u0159eva zohlednit.<\/p>\n

V\u00a0t\u00e9to studii je pops\u00e1na identifikace parametr\u016f, kter\u00e9 jsou v\u00fdznamn\u00e9 pro stabilitu \u0159\u00ed\u010dn\u00edho d\u0159eva v\u00a0meandruj\u00edc\u00edm koryt\u011b v\u00a0podm\u00ednk\u00e1ch tvrd\u00e9ho luhu st\u0159edn\u00ed Evropy. Tato pr\u00e1ce vych\u00e1z\u00ed z\u00a0p\u0159edpokladu, \u017ee anal\u00fdzou st\u00e1vaj\u00edc\u00edho \u0159\u00ed\u010dn\u00edho d\u0159eva lze v\u00a0ur\u010dit\u00e9m typu p\u0159\u00edrodn\u00edho \u00fazem\u00ed pom\u011brn\u011b snadno ur\u010dit parametry, kter\u00e9 jsou v\u00a0dan\u00e9m prost\u0159ed\u00ed pro stabilitu d\u0159eva rozhoduj\u00edc\u00ed. Pokud dok\u00e1\u017eeme rozli\u0161it mezi d\u0159evem, kter\u00e9 se korytem v\u00a0minulosti pohybovalo, a\u00a0kusy \u0159\u00ed\u010dn\u00edho d\u0159eva, kter\u00e9 se vyskytuj\u00ed na p\u016fvodn\u00edm\u00a0m\u00edst\u011b sv\u00e9ho p\u0159\u00edsunu do koryta, m\u016f\u017eeme si ud\u011blat p\u0159edstavu o\u00a0charakteru d\u0159eva, kter\u00e9 je v\u00a0dan\u00e9m \u00fazem\u00ed stabiln\u00ed. D\u00edl\u010d\u00ed studovanou ot\u00e1zkou t\u00e9\u017e byla shoda mezi subjektivn\u00edm odhadem stability a\u00a0pravd\u011bpodobnost\u00ed stability vypo\u010d\u00edtan\u00e9 na z\u00e1klad\u011b anal\u00fdzy \u0159\u00ed\u010dn\u00edho d\u0159eva.<\/p>\n

Metody<\/h2>\n

Lokalita<\/h3>\n

Ter\u00e9nn\u00ed v\u00fdzkum prob\u011bhl na \u0159ece Morav\u011b u\u00a0Litovle na lokalit\u011b zvan\u00e9 Vrapa\u010d (49.7140\u00b0 N, 17.0340\u00b0 E). \u0158eka Morava je zde tokem sedm\u00e9ho \u0159\u00e1du podle Strahlera s\u00a0pr\u016fm\u011brnou \u0161\u00ed\u0159kou koryta 28 metr\u016f a\u00a0plochou povod\u00ed 2\u2009144\u2008km2<\/sup>. Pro bl\u00edzk\u00fd profil v\u00a0Moravi\u010danech ud\u00e1v\u00e1 \u010cesk\u00fd hydrometeorologick\u00fd \u00fastav pr\u016fm\u011brn\u00fd pr\u016ftok 17,1\u2008m3<\/sup>.s-1<\/sup>. Dno \u0159eky je p\u0159ev\u00e1\u017en\u011b \u0161t\u011brkov\u00e9, b\u0159ehy tvo\u0159\u00ed vrstvy povod\u0148ov\u00fdch hl\u00edn siln\u00e9 p\u0159es dva metry. Povrch \u0159\u00ed\u010dn\u00ed nivy se nach\u00e1z\u00ed v\u00a0nadmo\u0159sk\u00e9 v\u00fd\u0161ce kolem 230\u00a0m\u00a0n.\u00a0m. Neupraven\u00e9 koryto Moravy u\u00a0Vrapa\u010de prod\u011bl\u00e1v\u00e1 ji\u017e des\u00edtky let zrychlen\u00fd geomorfologick\u00fd v\u00fdvoj, kter\u00fd je n\u00e1sledkem zv\u00fd\u0161en\u00ed sp\u00e1du na kontaktu s\u00a0regulovan\u00fdm korytem nad Litovl\u00ed. Dne\u0161n\u00ed sklon koryta v\u00a0\u00faseku Vrapa\u010de je v\u00a0pr\u016fm\u011bru 2,4\u00a0\u2030. To je v\u00edce ne\u017e na zbytku Moravy mezi jezem \u0158imice a\u00a0Litovl\u00ed, jeho\u017e sklon je v\u00a0rozmez\u00ed 0,6\u20130,9\u00a0\u2030. Krom\u011b postupn\u00e9ho zahlubov\u00e1n\u00ed se zv\u00fd\u0161en\u00ed sp\u00e1du projevuje v\u00fdraznou bo\u010dn\u00ed eroz\u00ed a\u00a0rozmeandrov\u00e1n\u00edm koryta v\u00a0\u00faseku dlouh\u00e9m t\u00e9m\u011b\u0159 1 kilometr. Pr\u016fvodn\u00edm jevem rychl\u00e9ho \u00fastupu b\u0159eh\u016f v\u00a0n\u00e1razov\u00fdch oblouc\u00edch meandr\u016f je postupn\u00e9 sesouv\u00e1n\u00ed strom\u016f z\u00a0b\u0159ehov\u00e9ho porostu do koryta. Lokalita je charakteristick\u00e1 bohat\u00fdm p\u0159\u00edsunem m\u00edstn\u00ed (autochtonn\u00ed) d\u0159evn\u00ed hmoty i\u00a0akumulac\u00ed \u0159\u00ed\u010dn\u00edho d\u0159eva splaven\u00e9ho z\u00a0v\u00fd\u0161e polo\u017een\u00fdch \u00fasek\u016f toku. Druhov\u00e1 struktura \u0159\u00ed\u010dn\u00edho d\u0159eva odpov\u00edd\u00e1 tvrd\u00e9mu luhu, kter\u00fdm je \u0159eka obklopena. P\u0159evl\u00e1d\u00e1 v\u00a0n\u011bm jasan, dub, javor a\u00a0l\u00edpa. Z\u00a0ni\u017e\u0161\u00edch parti\u00ed nivy se do \u0159eky dost\u00e1vaj\u00ed \u010d\u00e1sti vrb a\u00a0v\u00fdjime\u010dn\u011b topol\u016f. Ze vzd\u00e1len\u011bj\u0161\u00edch \u010d\u00e1st\u00ed povod\u00ed jsou splavov\u00e1ny t\u00e9\u017e \u010d\u00e1sti ol\u0161ov\u00fdch kmen\u016f a\u00a0ojedin\u011ble zbytky jehli\u010dnan\u016f. V\u011bt\u0161ina \u0159\u00ed\u010dn\u00edho d\u0159eva na lokalit\u011b m\u00e1 p\u0159irozen\u00fd p\u016fvod (v\u00fdvraty, zlomy). Na lokalit\u011b byla v\u00a0roce 2003 provedena stabilizace \u0159\u00ed\u010dn\u00edho d\u0159eva v\u00a0podob\u011b ukotven\u00ed 70 vzrostl\u00fdch strom\u016f ocelov\u00fdmi lany [19]. Po roce 2010 v\u0161ak po\u010det zachoval\u00fdch kotven\u00fdch kmen\u016f klesl na minimum a\u00a0neovliv\u0148oval nijak v\u00fdznamn\u011b akumulaci naplaven\u00e9ho d\u0159eva nebo stabilitu okoln\u00edch nekotven\u00fdch kmen\u016f. N\u011bkolik zb\u00fdvaj\u00edc\u00edch kotven\u00fdch kmen\u016f, kter\u00e9 se na lokalit\u011b vyskytovaly v\u00a0roce 2014, nebylo do evidence \u0159\u00ed\u010dn\u00edho d\u0159eva zahrnuto.<\/p>\n

Popis \u0159\u00ed\u010dn\u00edho d\u0159eva<\/h3>\n

Pokud n\u00e1s zaj\u00edmaj\u00ed vlastnosti \u0159\u00ed\u010dn\u00edho d\u0159eva, kter\u00e9 je v\u00a0podm\u00ednk\u00e1ch dan\u00e9 lokality stabiln\u00ed, lze to zjistit anal\u00fdzou st\u00e1vaj\u00edc\u00edho d\u0159eva na lokalit\u011b, pokud dok\u00e1\u017eeme u\u00a0jednotliv\u00fdch kus\u016f ur\u010dit, zda se v\u00a0koryt\u011b pohybuj\u00ed, nebo z\u016fst\u00e1vaj\u00ed po dlouhou dobu na m\u00edst\u011b. S\u00a0jistou m\u00edrou zjednodu\u0161en\u00ed m\u016f\u017eeme autochtonn\u00ed \u0159\u00ed\u010dn\u00ed d\u0159evo v\u00a0t\u00e9to studii pova\u017eovat za kusy stabiln\u00ed p\u0159inejmen\u0161\u00edm za\u00a0obdob\u00ed n\u011bkolika posledn\u00edch let. Naopak naplaven\u00e9 d\u0159evo by m\u011blo b\u00fdt charakterizov\u00e1no vlastnostmi, kter\u00e9 jsou typick\u00e9 pro nestabiln\u00ed kusy b\u011b\u017en\u011b un\u00e1\u0161en\u00e9 povodn\u011bmi. Tato studie proto pova\u017euje kategorie autochtonn\u00ed\/naplaven\u00fd za analogick\u00e9 pro stabiln\u00ed\/nestabiln\u00ed \u0159\u00ed\u010dn\u00ed d\u0159evo. V\u00a0dal\u0161\u00edm textu tyto term\u00edny pou\u017e\u00edv\u00e1m s\u00a0v\u011bdom\u00edm t\u00e9to analogie.<\/p>\n

Tabulka 1. P\u0159ehled charakteristik pou\u017eit\u00fdch pro popis \u0159\u00ed\u010dn\u00edho d\u0159eva a\u00a0jejich hodnoty
\nTable 1. List of the characteristics used to describe the woody debris and their values<\/h5>\n
\"kozeny-tabulka-1-s-legendou\"<\/a><\/h6>\n

Popis \u0159\u00ed\u010dn\u00edho d\u0159eva prob\u00edhal na zm\u00edn\u011bn\u00e9 lokalit\u011b v\u00a0\u0159\u00edjnu 2014. Na \u00faseku dlouh\u00e9m p\u0159ibli\u017en\u011b 700 metr\u016f, zahrnuj\u00edc\u00edm \u010dty\u0159i meandrov\u00e9 oblouky, bylo pops\u00e1no vyskytuj\u00edc\u00ed se \u0159\u00ed\u010dn\u00ed d\u0159evo od d\u00e9lky dvou metr\u016f a\u00a0pr\u016fm\u011bru 10\u2008cm. D\u0159evo bylo popisov\u00e1no v\u00a0cel\u00e9m rozsahu koryta, tedy a\u017e po okraj b\u0159ehov\u00e9 hrany. Popis prob\u011bhl v\u00a0obdob\u00ed n\u00edzk\u00fdch pr\u016ftok\u016f, kdy bylo koryto z\u00a0v\u00edce ne\u017e 95\u2005% broditeln\u00e9. Pro ka\u017ed\u00fd jednotliv\u00fd kus \u0159\u00ed\u010dn\u00edho d\u0159eva byly zaznamen\u00e1ny charakteristiky velikosti, tvaru a\u00a0stupn\u011b rozkladu. D\u00e1le byla zaznamen\u00e1na poloha v\u00a0koryt\u011b, poloha v\u016f\u010di ostatn\u00edm kus\u016fm \u0159\u00ed\u010dn\u00edho d\u0159eva a\u00a0orientace v\u016f\u010di sm\u011bru proud\u011bn\u00ed. P\u0159ed statistickou anal\u00fdzou byl rozsah t\u0159\u00edd kategori\u00e1ln\u00edch prom\u011bnn\u00fdch agregov\u00e1n, abychom z\u00edskali t\u0159\u00eddy s\u00a0dostate\u010dn\u00fdmi \u010detnostmi pozorov\u00e1n\u00ed. P\u0159ehled popisn\u00fdch charakteristik je uveden v\u00a0tabulce 1<\/em>.<\/p>\n

U\u00a0ka\u017ed\u00e9ho kusu bylo rovn\u011b\u017e posouzeno, zda jde o\u00a0kus naplaven\u00fd, nebo ulo\u017een\u00fd v\u00a0m\u00edst\u011b sv\u00e9ho vzniku (tzv.\u00a0autochtonn\u00ed). Ur\u010den\u00ed bylo provedeno pomoc\u00ed vzhledu kmene (nap\u0159.\u00a0zp\u016fsob a\u00a0vy\u0161\u0161\u00ed m\u00edra abraze u\u00a0naplaven\u00fdch kus\u016f) a\u00a0jeho pozice (nap\u0159.\u00a0v\u00fdskyt v\u00a0akumulaci zjevn\u011b naplaven\u00e9ho d\u0159eva, na \u0161t\u011brkov\u00fdch lavic\u00edch mimo potenci\u00e1ln\u00ed m\u00edsta p\u0159\u00edsunu, p\u0159i b\u0159ez\u00edch bez stromov\u00e9 vegetace\u00a0atd.) nebo na z\u00e1klad\u011b stop v\u00a0ter\u00e9nu (nap\u0159.\u00a0bezprost\u0159edn\u011b souvisej\u00edc\u00ed b\u0159ehov\u00e1 n\u00e1tr\u017e). Identifikaci autochtonn\u00edch kmen\u016f usnad\u0148ovala skute\u010dnost, \u017ee u\u00a0meandruj\u00edc\u00edch tok\u016f je hlavn\u00edm p\u0159\u00edsunov\u00fdm mechanismem b\u0159ehov\u00e1 eroze [1]. P\u0159\u00edklady \u0159\u00ed\u010dn\u00edho d\u0159eva vyskytuj\u00edc\u00edho se na lokalit\u011b p\u0159ibli\u017euje obr.\u00a01<\/em>. Mezi naplaven\u00e9 byly za\u0159azeny kmeny, kter\u00e9 se posunuly alespo\u0148 o\u00a025 metr\u016f od pravd\u011bpodobn\u00e9ho m\u00edsta p\u0159\u00edsunu do koryta. Rozli\u0161en\u00ed na autochtonn\u00ed a\u00a0naplaven\u00e9 \u0159\u00ed\u010dn\u00ed d\u0159evo bylo provedeno bez ohledu na \u010das, kter\u00fd uplynul od p\u0159em\u00edst\u011bn\u00ed naplaven\u00e9ho kusu nebo p\u0159\u00edsunu autochtonn\u00edho kusu ze b\u0159ehu. To znamen\u00e1, \u017ee jako naplaven\u00fd byl ozna\u010den i\u00a0kmen, kter\u00fd se p\u0159esunul p\u0159i jedn\u00e9 povod\u0148ov\u00e9 epizod\u011b v\u00a0minulosti a\u00a0v\u00a0nov\u00e9 pozici ji\u017e z\u016fstal po \u0159adu let.<\/p>\n\"kozeny-1\"<\/a>\n

Obr.\u00a01. P\u0159\u00edklady \u0159\u00ed\u010dn\u00edho d\u0159eva v\u00a0\u0159ece Morav\u011b na lokalit\u011b Vrapa\u010d v\u00a0CHKO Litovelsk\u00e9 Pomorav\u00ed; A\u2013C: naplaven\u00e9 kusy, D\u2013F: autochtonn\u00ed kusy
\nFig. 1. Examples of wood in Morava River at the Vrapa\u010d locality, Litovelsk\u00e9 Pomorav\u00ed PLA; A\u2013C: allochthonous pieces, D\u2013F: autochthonous pieces<\/h6>\n

Posledn\u00ed zaznamen\u00e1vanou charakteristikou byl subjektivn\u00ed odhad stability popisovan\u00e9ho kusu, tedy odolnost v\u016f\u010di odplaven\u00ed p\u0159i povodn\u00edch. Hlediskem pro posouzen\u00ed byly informace z\u00a0literatury a\u00a0v\u00edcelet\u00e1 zku\u0161enost autora s\u00a0dynamikou \u0159\u00ed\u010dn\u00edho d\u0159eva na lokalit\u011b. Tato charakteristika souvis\u00ed do zna\u010dn\u00e9 m\u00edry s\u00a0popisovan\u00fdmi vlastnostmi \u0159\u00ed\u010dn\u00edho d\u0159eva, ale jejich souhrnn\u00e9 hodnocen\u00ed pro dan\u00fd kus bylo rychl\u00fdm subjektivn\u00edm odhadem. Anal\u00fdza tohoto odhadu prob\u00edhala proto odd\u011blen\u011b od ostatn\u00edch, objektivn\u011b hodnocen\u00fdch charakteristik. Subjektivn\u00ed odhad stability napodobuje ter\u00e9nn\u00ed odhad, kter\u00fd by na z\u00e1klad\u011b m\u00edstn\u00ed zku\u0161enosti mohl prov\u00e1d\u011bt nap\u0159\u00edklad pracovn\u00edk pov\u011b\u0159en\u00fd spr\u00e1vou vodn\u00edho toku.<\/p>\n

Statistick\u00e1 anal\u00fdza ter\u00e9nn\u00edch dat<\/h3>\n

Anal\u00fdza dat spo\u010d\u00edvala v\u00a0ur\u010den\u00ed nez\u00e1visl\u00fdch popisn\u00fdch charakteristik, kter\u00e9 jsou v\u00fdznamn\u00e9 pro rozli\u0161en\u00ed mezi autochtonn\u00edmi a\u00a0naplaven\u00fdmi kusy \u0159\u00ed\u010dn\u00edho d\u0159eva (prom\u011bnn\u00e1 AUTOCH). Vzhledem k\u00a0povaze z\u00edskan\u00fdch dat, kdy z\u00e1visl\u00e1 prom\u011bnn\u00e1 nab\u00fdv\u00e1 pouze dvou hodnot (autochtonn\u00ed\/naplaven\u00fd, resp. stabiln\u00ed\/nestabiln\u00ed), byla pro statistickou anal\u00fdzu zvolena logistick\u00e1 regrese. Logistick\u00fd regresn\u00ed model popisuje pravd\u011bpodobnost P<\/em> zkouman\u00e9ho jevu za podm\u00ednek dan\u00fdch hodnotami nez\u00e1visl\u00fdch prom\u011bnn\u00fdch x<\/em>. <\/em>Vyjad\u0159ujeme jej rovnic\u00ed:<\/p>\n\"kozeny-vzorec-1\"<\/a>\n\n\n\n\n\n
kde<\/td>\nP<\/em><\/td>\nje<\/td>\npravd\u011bpodobnost, \u017ee nastane zkouman\u00fd jev,<\/td>\n<\/tr>\n
<\/td>\nx<\/em><\/td>\n<\/td>\nvyjad\u0159uje hodnoty nez\u00e1visl\u00fdch prom\u011bnn\u00fdch,<\/td>\n<\/tr>\n
<\/td>\n\u03b20<\/sub> <\/em>a\u00a0\u03b2i<\/em><\/td>\n<\/td>\nkoeficienty regresn\u00edho modelu.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

Ekvivalentn\u011b lze tento model vyj\u00e1d\u0159it rovnic\u00ed:<\/p>\n\"kozeny-vzorec-2\"<\/a>\n

Proto\u017ee hodnoty koeficient\u016f \u03b2<\/em> jsou v\u00fdsledkem logistick\u00e9ho regresn\u00edho modelu, m\u016f\u017eeme po dosazen\u00ed hodnot prom\u011bnn\u00fdch x<\/em> vypo\u010d\u00edtat pravd\u011bpodobnost P<\/em>, s\u00a0jakou nastane zkouman\u00fd jev [20].<\/p>\n

V\u00fdb\u011br vhodn\u00fdch prom\u011bnn\u00fdch a\u00a0tvorba logistick\u00e9ho regresn\u00edho modelu prob\u00edhala v\u00a0n\u011bkolika kroc\u00edch. V\u00a0prvn\u00ed f\u00e1zi byla z\u00e1visl\u00e1 prom\u011bnn\u00e1 AUTOCH modelov\u00e1na v\u017edy jen pomoc\u00ed jedn\u00e9 z\u00a0popisn\u00fdch charakteristik. V\u00fdsledek tohoto tzv.\u00a0univariantn\u00edho modelu nazna\u010dil, kter\u00e9 z\u00a0popisn\u00fdch charakteristik jsou v\u00fdznamn\u00e9 pro za\u0159azen\u00ed kusu mezi autochtonn\u00ed \u010di naplaven\u00e9 \u0159\u00ed\u010dn\u00ed d\u0159evo. Z\u00e1rove\u0148 byly zhodnoceny p\u0159\u00edpadn\u00e9 z\u00e1vislosti mezi popisn\u00fdmi charakteristikami pomoc\u00ed graf\u016f, korela\u010dn\u00ed matice a\u00a0hodnot VIF (variance inflation factor). Do dal\u0161\u00ed anal\u00fdzy byly za\u0159azeny pouze charakteristiky, kter\u00e9 univariantn\u00ed model ur\u010dil jako v\u00fdznamn\u00e9 pro stabilitu \u0159\u00ed\u010dn\u00edho d\u0159eva a\u00a0z\u00e1rove\u0148 tyto prom\u011bnn\u00e9 nebyly\u00a0v\u00a0korelaci. Z\u00a0v\u00fdb\u011bru v\u00fdznamn\u00fdch a\u00a0vz\u00e1jemn\u011b nekoreluj\u00edc\u00edch popisn\u00fdch charakteristik byl sestaven logistick\u00fd regresn\u00ed model. Abychom z\u00edskali men\u0161\u00ed skupinu popisn\u00fdch charakteristik se silnou z\u00e1vislost\u00ed ke stabilit\u011b, postupn\u011b byly z\u00a0modelu vylu\u010dov\u00e1ny prom\u011bnn\u00e9 s\u00a0nejni\u017e\u0161\u00ed hodnotou v\u00fdznamnosti (p\u00a0\u2013 hodnota). Do v\u00fdsledn\u00e9ho modelu pak byl za\u0159azen v\u00fdb\u011br nez\u00e1visl\u00fdch prom\u011bnn\u00fdch s\u00a0o\u010dek\u00e1van\u00fdm siln\u00fdm vlivem na ur\u010den\u00ed stability kmene. Fin\u00e1ln\u00ed model byl ov\u011b\u0159en pomoc\u00ed anal\u00fdzy rezidu\u00ed.<\/p>\n\"kozeny-2\"<\/a>\n

Obr.\u00a02. Popisn\u00e9 charakteristiky \u0159\u00ed\u010dn\u00edho d\u0159eva pro kusy naplaven\u00e9 (N) a\u00a0autochtonn\u00ed (A); krabicov\u00fd diagram zn\u00e1zor\u0148uje medi\u00e1n, 1\u20133 kvartil (krabi\u010dka), maxima a\u00a0minima (vousy), odlehl\u00e9 hodnoty jsou zobrazeny kole\u010dky; pro kategori\u00e1ln\u00ed prom\u011bnn\u00e9 jsou \u0161\u00ed\u0159ky sloupc\u016f \u00fam\u011brn\u00e9 po\u010dtu hodnot
\nFig. 2. Descriptive characteristics of wood for pieces of driftwood (N) and autochthonous (A) wood pieces; the box and whisker plot shows the median, 1st<\/sup>\u20133rd<\/sup> quartile (box), maximum and minimum values (whisker), outliers are displayed as circles; for discrete variables the width of columns is proportional to the number of values<\/h6>\n

Subjektivn\u011b hodnocen\u00e1 stabilita byla analyzov\u00e1na odd\u011blen\u011b od ostatn\u00edch prom\u011bnn\u00fdch, tedy pouze v\u00a0univariantn\u00edm modelu. D\u00e1le byla porovn\u00e1na s\u00a0v\u00fdsledky logistick\u00e9ho regresn\u00edho modelu a\u00a0stanovena m\u00edra korelace subjektivn\u00edho hodnocen\u00ed a\u00a0pravd\u011bpodobnost\u00ed po\u010d\u00edtan\u00fdch modelem. Statistick\u00e1 anal\u00fdza byla provedena v\u00a0programu R [21].<\/p>\n

V\u00fdsledky<\/h2>\n

Na lokalit\u011b bylo pops\u00e1no celkem 160 kus\u016f \u0159\u00ed\u010dn\u00edho d\u0159eva o\u00a0d\u00e9lce od 2 do 32 metr\u016f. Toto mno\u017estv\u00ed p\u0159edstavovalo objem 134,31\u2008m3<\/sup> na plo\u0161e p\u0159ibli\u017en\u011b 1,96\u2008ha. Mezi autochtonn\u00ed kmeny bylo za\u0159azeno 96 kus\u016f, naplaven\u00fdch kus\u016f bylo 64. Jen 13 p\u0159\u00edpad\u016f tvo\u0159ily \u017eiv\u00e9 nebo do\u017e\u00edvaj\u00edc\u00ed stromy, ostatn\u00ed \u0159\u00ed\u010dn\u00ed d\u0159evo tvo\u0159ily odum\u0159el\u00e9 kusy. Rovn\u00fdch 100 kmen\u016f m\u011blo zachoval\u00e9 ko\u0159eny. P\u0159ibli\u017en\u011b dv\u011b t\u0159etiny \u0159\u00ed\u010dn\u00edho d\u0159eva se vyskytovaly v\u00a0akumulac\u00edch, tedy ve vz\u00e1jemn\u00e9m kontaktu s\u00a0alespo\u0148 dv\u011bma dal\u0161\u00edmi kmeny. U\u00a0v\u011bt\u0161iny kus\u016f lze p\u0159edpokl\u00e1dat pom\u011brn\u011b dobrou mechanickou pevnost, proto\u017ee v\u011bt\u0161ina \u0159\u00ed\u010dn\u00edho d\u0159eva spadala do kategorie kmen\u016f sice ji\u017e bez k\u016fry, ale dosud s\u00a0pevn\u00fdm d\u0159evem. Hodnoty \u010detnost\u00ed a\u00a0pr\u016fm\u011brn\u00e9 hodnoty pro jednotliv\u00e9 sledovan\u00e9 parametry jsou uvedeny v\u00a0tabulce 1<\/em>. Minima a\u00a0maxima spojit\u00fdch prom\u011bnn\u00fdch lze d\u00e1le ode\u010d\u00edst z\u00a0graf\u016f na obr.\u00a02<\/em>.<\/p>\n\"kozeny-3\"<\/a>\n

Obr.\u00a03. Modelovan\u00e9 pravd\u011bpodobnosti stability pro odum\u0159el\u00fd kus \u0159\u00ed\u010dn\u00edho d\u0159eva v\u00a0t\u0159\u00edd\u011b rozkladu 0\u20133; grafy zn\u00e1zor\u0148uj\u00ed r\u016fzn\u00e9 kombinace d\u00e9lky, v\u00fd\u0161ky ulo\u017een\u00ed, zazemn\u011bn\u00ed a\u00a0p\u0159\u00edtomnosti ko\u0159en\u016f
\nFig. 3. The stability model for a\u00a0deadwood piece in the decay class 0\u20133; the graphs show various combinations of length, log elevation, burial and presence of roots<\/h6>\n

Tabulka 2<\/em> uv\u00e1d\u00ed v\u00fdsledek univariantn\u00edho testov\u00e1n\u00ed v\u00fdznamnosti jednotliv\u00fdch prom\u011bnn\u00fdch pro ur\u010den\u00ed kategorie autochtonn\u00ed\/naplaven\u00fd kus. Statistick\u00e1 v\u00fdznamnost byla jednotliv\u011b prok\u00e1z\u00e1na u\u00a0v\u0161ech prom\u011bnn\u00fdch, krom\u011b stupn\u011b rozkladu, kde anal\u00fdza jen t\u011bsn\u011b nedos\u00e1hla 5% hladiny statistick\u00e9 v\u00fdznamnosti. Vztahy mezi popisn\u00fdmi charakteristikami a\u00a0prom\u011bnnou AUTOCH jsou zobrazeny v\u00a0grafech na obr.\u00a02<\/em>. Autochtonn\u00ed kusy \u0159\u00ed\u010dn\u00edho d\u0159eva se vyzna\u010dovaly v\u00fdrazn\u011b v\u011bt\u0161\u00ed d\u00e9lkou doprov\u00e1zenou m\u00edrn\u011b vy\u0161\u0161\u00edm pr\u016fm\u011brem kmene. Byly obvykle ulo\u017eeny v\u00a0\u00farovni hladiny nebo pod n\u00ed, tedy n\u00ed\u017ee ne\u017e kmeny naplaven\u00e9. St\u0159edn\u00ed m\u00edra zazemn\u011bn\u00ed se u\u00a0autochtonn\u00edch kmen\u016f pohybovala okolo 20\u2005%, kde\u017eto u\u00a0kmen\u016f naplaven\u00fdch byl kontakt s\u00a0materi\u00e1lem dna nebo b\u0159eh\u016f v\u00fdjimkou. Mezi autochtonn\u00edmi kmeny byl vy\u0161\u0161\u00ed pod\u00edl \u017eiv\u00fdch strom\u016f se zachovalou korunou a\u00a0ko\u0159enov\u00fdm syst\u00e9mem. I\u00a0kdy\u017e se asi dv\u011b t\u0159etiny z\u00a0celkov\u00e9ho po\u010dtu kus\u016f \u0159\u00ed\u010dn\u00edho d\u0159eva vyskytovaly v\u00a0akumulac\u00edch, autochtonn\u00ed kmeny se nach\u00e1zely dvakr\u00e1t \u010dast\u011bji samostatn\u011b ne\u017e kmeny naplaven\u00e9. T\u00e9m\u011b\u0159 60\u2005% autochtonn\u00edch kmen\u016f se nach\u00e1zelo v\u00a0poloze bl\u00edzk\u00e9 sm\u011bru proud\u011bn\u00ed se \u0161pi\u010dkou sm\u011b\u0159uj\u00edc\u00ed po proudu. Mezi naplaven\u00fdmi kmeny se tato vlastnost vyskytovala jen asi v\u00a020\u2005% p\u0159\u00edpad\u016f. Naprost\u00e1 v\u011bt\u0161ina \u0159\u00ed\u010dn\u00edho d\u0159eva se sice vyskytovala v\u00a0rozsahu vymezen\u00e9m \u0161\u00ed\u0159\u00ed hladiny za b\u011b\u017en\u00fdch pr\u016ftok\u016f, pokud se ale n\u011bjak\u00e9 kusy vyskytovaly na b\u0159ehu, byly to \u010dast\u011bji kusy naplaven\u00e9. Naplaven\u00e9 kusy tak\u00e9 \u010dast\u011bji ne\u017e autochtonn\u00ed spadaly do vy\u0161\u0161\u00edch t\u0159\u00edd stupn\u011b rozkladu, a\u010dkoliv p\u0159ev\u00e1\u017en\u00e1 v\u011bt\u0161ina kus\u016f bylo d\u0159evo mechanicky pevn\u00e9. Subjektivn\u011b byla stabilita v\u011bt\u0161iny naplaven\u00fdch kus\u016f hodnocena zn\u00e1mkami 1\u20133, kde\u017eto mezi autochtonn\u00edmi kmeny p\u0159eva\u017eovaly t\u0159\u00eddy 6\u20139, tedy velmi stabiln\u00ed poloha. Statistick\u00e1 v\u00fdznamnost t\u00e9to z\u00e1vislosti byla z\u00e1rove\u0148 jedna z\u00a0nejvy\u0161\u0161\u00edch ve srovn\u00e1n\u00ed s\u00a0ostatn\u00edmi charakteristikami.<\/p>\n

Tabulka 2. Statistick\u00e1 v\u00fdznamnost z\u00e1vislosti prom\u011bnn\u00e9 AUTOCH na popisn\u00fdch charakteristik\u00e1ch v\u00a0postupn\u00fdch kroc\u00edch statistick\u00e9 anal\u00fdzy a\u00a0regresn\u00ed koeficienty v\u00fdsledn\u00e9ho modelu (hladiny v\u00fdznamnosti p\u00a0\u2013 hodnoty: [***] p < 0,001; [**] 0,001 < p < 0,01; [*] 0,01 < p < 0,05; [nesig.] p > 0,05)
\nTable 2. Statistical significance of the AUTOCH variable on the descriptive characteristics in the sequential steps of statistical analysis and the regression coefficients of the resulting model (significance levels of p\u00a0\u2013 value: [***] p < 0.001; [**] 0.001 < p < 0.01; [*] 0.01 < p < 0.05; [nesig.] p > 0.05)<\/h5>\n\"kozeny-tabulka-2\"<\/a>\n

I\u00a0kdy\u017e byly popisn\u00e9 charakteristiky zaznamen\u00e1v\u00e1ny jako nez\u00e1visl\u00e9 prom\u011bnn\u00e9, je z\u0159ejm\u00e9, \u017ee n\u011bkter\u00e9 prom\u011bnn\u00e9 spolu mohly korelovat. Korela\u010dn\u00ed matice uk\u00e1zala v\u00fdznamn\u011bj\u0161\u00ed korelaci mezi prom\u011bnn\u00fdmi NAD_HL a\u00a0ZAZEM (Pearson\u016fv korela\u010dn\u00ed koeficient = – 0,48) a\u00a0mezi prom\u011bnn\u00fdmi DELKA a\u00a0PRUMER (0,5). U\u00a0ostatn\u00edch spojit\u00fdch prom\u011bnn\u00fdch byl korela\u010dn\u00ed koeficient ni\u017e\u0161\u00ed ne\u017e 0,2. Prom\u011bnn\u00e1 PRUMER nebyla nakonec do v\u00fdsledn\u00e9ho modelu za\u0159azena. Druh\u00fdm p\u0159\u00edstupem k\u00a0odhalen\u00ed korelac\u00ed mezi popisn\u00fdmi charakteristikami byl v\u00fdpo\u010det hodnoty VIF (variance inflation factor). Ve v\u0161ech p\u0159\u00edpadech nedosahoval hodnoty 3, co\u017e je obecn\u011b pova\u017eov\u00e1no za v\u00fdsledek, kter\u00fd nesv\u011bd\u010d\u00ed o\u00a0v\u00fdznamn\u00e9 korelaci mezi prom\u011bnn\u00fdmi.<\/p>\n

Popisn\u00e9 charakteristiky byly d\u00e1le testov\u00e1ny logistick\u00fdm regresn\u00edm modelem. Jak je vid\u011bt z\u00a0p\u0159ehledu v\u00a0tabulce 2<\/em>, statisticky v\u00fdznamnou popisnou charakteristikou byla d\u00e9lka kmene, m\u00edra zazemn\u011bn\u00ed, stupe\u0148 rozkladu a\u00a0p\u0159\u00edtomnost ko\u0159en\u016f. T\u011bsn\u011b za hranic\u00ed 5% statistick\u00e9 v\u00fdznamnosti byl vyhodnocen pr\u016fm\u011br kmene. Abychom sn\u00ed\u017eili po\u010det prom\u011bnn\u00fdch v\u00a0modelu a\u00a0omezili je na statisticky v\u00fdznamn\u00e9 z\u00e1vislosti, upravili jsme model postupn\u00fdm odeb\u00edr\u00e1n\u00edm prom\u011bnn\u00fdch s\u00a0nejni\u017e\u0161\u00ed statistickou v\u00fdznamnost\u00ed. Z\u00e1rove\u0148 byl do modelu vlo\u017een prvek interakce mezi prom\u011bnn\u00fdmi NAD_HL a\u00a0ROZKLAD2tr, kter\u00e1 byla zji\u0161t\u011bna p\u0159i kontrole z\u00e1vislost\u00ed mezi prom\u011bnn\u00fdmi. Interakce je jev, kdy se vlivem jedn\u00e9 prom\u011bnn\u00e9 m\u011bn\u00ed sm\u011brnice modelovan\u00e9ho vztahu u\u00a0jin\u00e9 prom\u011bnn\u00e9. V\u00a0na\u0161em p\u0159\u00edpad\u011b to konkr\u00e9tn\u011b znamenalo, \u017ee mechanicky odoln\u00e9 kmeny z\u00a0t\u0159\u00edd rozkladu 0\u20133 byly \u010dast\u011bji hodnoceny jako autochtonn\u00ed, pokud se nach\u00e1zely pod hladinou. Pro \u010d\u00e1ste\u010dn\u011b zetlel\u00e9 d\u0159evo v\u00a0t\u0159\u00edd\u00e1ch rozkladu 3,5\u20135 byla z\u00e1vislost opa\u010dn\u00e1\u00a0\u2013 autochtonn\u00ed kmeny byly ulo\u017eeny v\u00fd\u0161e nad hladinou ne\u017e naplaven\u00e9 kusy. V\u00a0kategorii rozkladu 3,5\u20135 se nach\u00e1zelo jen 18 kmen\u016f, tedy 11\u2005% celkov\u00e9ho po\u010dtu. Logick\u00e9 vysv\u011btlen\u00ed interakce m\u016f\u017ee vych\u00e1zet z\u00a0podm\u00ednek na lokalit\u011b. Naplaven\u00e9 d\u0159evo je v\u00a0koryt\u011b t\u00e9m\u011b\u0159 v\u017edy ulo\u017eeno v\u00a0akumulac\u00edch, kter\u00e9 jsou za b\u011b\u017en\u00fdch pr\u016ftok\u016f nad hladinou. Naopak v\u011bt\u0161ina autochtonn\u00edch kmen\u016f se vyskytuje ve vod\u011b v\u00a0\u00farovni hladiny nebo pod n\u00ed. Toto d\u0159evo z\u00e1rove\u0148 spad\u00e1 nej\u010dast\u011bji do t\u0159\u00eddy rozkladu 0\u20133, proto\u017ee pono\u0159en\u00e9 d\u0159evo se jen pomalu rozkl\u00e1d\u00e1. Naplaven\u00e9 d\u0159evo v\u00a0akumulac\u00edch dob\u0159e vysych\u00e1, co\u017e op\u011bt br\u00e1n\u00ed rychl\u00e9mu rozkladu. Jin\u00e1 je situace na b\u0159ez\u00edch. Autochtonn\u00ed i\u00a0naplaven\u00e9 d\u0159evo zde m\u016f\u017ee le\u017eet v\u00a0r\u016fzn\u00e9 v\u00fd\u0161i nad hladinou a\u00a0z\u00e1rove\u0148 je ulo\u017eeno ve vlhku, co\u017e napom\u00e1h\u00e1 rozkladu. Za vznikem popsan\u00e9 interakce tedy mohou st\u00e1t vlastnosti \u0159\u00ed\u010dn\u00edho d\u0159eva ulo\u017een\u00e9ho na b\u0159ez\u00edch, kde nen\u00ed rozd\u00edl ve v\u00fd\u0161kov\u00e9m ulo\u017een\u00ed mezi autochtonn\u00edmi a\u00a0naplaven\u00fdmi kusy a\u00a0z\u00e1rove\u0148 jsou v\u0161echny zde ulo\u017een\u00e9 kmeny n\u00e1chyln\u011bj\u0161\u00ed k\u00a0rozkladu.<\/p>\n\"kozeny-4\"<\/a>\n

Obr.\u00a04. Korelace subjektivn\u00edho odhadu stability a\u00a0modelovan\u00e9 pravd\u011bpodobnosti stability kus\u016f \u0159\u00ed\u010dn\u00edho d\u0159eva
\nFig. 4. Correlation of the subjective estimate of stability and modelled probabilities of wood piece stability<\/h6>\n

Do v\u00fdsledn\u00e9ho modelu popisuj\u00edc\u00edho vliv popisn\u00fdch charakteristik na ur\u010den\u00ed kmene jako autochtonn\u00ed \u010di naplaven\u00fd byly zahrnuty prom\u011bnn\u00e9 NAD_HL, ZAZEM, ZIVY, KORENY, DELKA, ROZKLAD2tr a\u00a0zm\u00edn\u011bn\u00e1 interakce NAD_HL: ROZKLAD2tr. Koeficienty \u03b2<\/em> a\u00a0statistickou v\u00fdznamnost jednotliv\u00fdch parametr\u016f modelu zobrazuje tabulka 2<\/em>. Pomoc\u00ed vykreslen\u00ed r\u016fzn\u00fdch typ\u016f rezidu\u00ed proti nez\u00e1visl\u00fdm prom\u011bnn\u00fdm a\u00a0odhadnut\u00fdm hodnot\u00e1m byla ov\u011b\u0159ena platnost modelu. V\u00a0\u017e\u00e1dn\u00e9m z\u00a0p\u0159\u00edpad\u016f nedoch\u00e1zelo k\u00a0v\u00fdrazn\u00e9mu odklonu od nuly. Jak bylo pops\u00e1no v\u00fd\u0161e, autochtonn\u00ed kmeny m\u016f\u017eeme pova\u017eovat za stabiln\u00ed \u0159\u00ed\u010dn\u00ed d\u0159evo, naplaven\u00e9 kmeny jsou naopak nestabiln\u00ed pohyblivou slo\u017ekou \u0159\u00ed\u010dn\u00edho d\u0159eva. Koeficienty \u03b2<\/em> a\u00a0skute\u010dn\u00e9 hodnoty prom\u011bnn\u00fdch x<\/em> pro jednotliv\u00e9 kusy \u0159\u00ed\u010dn\u00edho d\u0159eva jsme dosadili do rovnice modelu (2) a\u00a0z\u00edskali tak pravd\u011bpodobnost stability pro ka\u017ed\u00fd jednotliv\u00fd kus \u0159\u00ed\u010dn\u00edho d\u0159eva na lokalit\u011b. V\u00a091\u2005% p\u0159\u00edpad\u016f se vypo\u010dten\u00e1 pravd\u011bpodobnost stability shodovala s\u00a0ozna\u010den\u00edm kmene za autochtonn\u00ed kus. Modelovan\u00e1 stabilita p\u0159edstavuje hodnotu pravd\u011bpodobnosti, s\u00a0jakou bude kmen stabiln\u00ed v\u00a0podm\u00ednk\u00e1ch pr\u016ftok\u016f, kter\u00e9 se na lokalit\u011b vyskytly v\u00a0obdob\u00ed p\u0159edch\u00e1zej\u00edc\u00edm ter\u00e9nn\u00edmu sb\u011bru dat.<\/p>\n

V\u00fdsledn\u00fd model logistick\u00e9 regrese umo\u017e\u0148uje v\u00fdpo\u010det pravd\u011bpodobnosti odplaven\u00ed pro kus \u0159\u00ed\u010dn\u00edho d\u0159eva s\u00a0konkr\u00e9tn\u00edmi parametry. Nap\u0159\u00edklad kmen o\u00a0d\u00e9lce 15 metr\u016f ulo\u017een\u00fd v\u00a0\u00farovni hladiny z\u016fstane na sv\u00e9m m\u00edst\u011b s\u00a0pravd\u011bpodobnost\u00ed 41\u2005%. Pokud bude m\u00edt nav\u00edc zachoval\u00e9 ko\u0159eny, stoupne tato pravd\u011bpodobnost na 78\u2005%. Bude\u00ad\u00ad\u2011li nav\u00edc v\u00a0kontaktu se sedimentem na 20 procentech sv\u00e9ho povrchu, v\u00fdsledn\u00e1 pravd\u011bpodobnost stability stoupne na 91\u2005%. To je pravd\u011bpodobnost odplaven\u00ed kusu \u0159\u00ed\u010dn\u00edho d\u0159eva s\u00a0medi\u00e1nov\u00fdmi vlastnostmi typick\u00e9ho stabiln\u00edho kmene. Oproti tomu b\u011b\u017en\u00fd nestabiln\u00ed kmen o\u00a0d\u00e9lce \u0161est metr\u016f, bez ko\u0159en\u016f, voln\u011b ulo\u017een\u00fd ve v\u00fd\u0161ce 0,5\u2008m nad hladinou z\u016fstane p\u0159i povodni na p\u016fvodn\u00ed pozici jen s\u00a0pravd\u011bpodobnost\u00ed 3,4\u2005%. Uveden\u00e9 pravd\u011bpodobnosti plat\u00ed pro odum\u0159el\u00e9 kmeny ve t\u0159\u00edd\u011b rozkladu 0\u20133. Pro \u010d\u00e1ste\u010dn\u011b zetlel\u00e9 kmeny ve t\u0159\u00edd\u011b rozkladu 3,5\u20135 by se v\u00fdsledn\u00e1 pravd\u011bpodobnost vlivem interakce NAD_HL:ROZKLAD2tr zv\u00fd\u0161ila na 17\u2005% u\u00a0kr\u00e1tk\u00e9ho nestabiln\u00edho kusu, zat\u00edmco u\u00a0dlouh\u00e9ho kmene s\u00a0ko\u0159eny by z\u016fstala t\u00e9m\u011b\u0159 stejn\u00e1 (93\u2005%). Pro odum\u0159el\u00e9 \u0159\u00ed\u010dn\u00ed d\u0159evo ve t\u0159\u00edd\u011b rozkladu 0\u20133 jsou pravd\u011bpodobnosti stability pro r\u016fzn\u00e9 kombinace d\u00e9lky, p\u0159\u00edtomnosti ko\u0159en\u016f, ulo\u017een\u00ed nad hladinou a\u00a0m\u00edry zazemn\u011bn\u00ed vyneseny v\u00a0grafech na obr.\u00a03<\/em>.<\/p>\n

Vypo\u010dten\u00e9 pravd\u011bpodobnosti jsme d\u00e1le pou\u017eili pro korelaci mezi modelovanou pravd\u011bpodobnost\u00ed a\u00a0subjektivn\u00edm ur\u010den\u00edm stability v\u00a0prom\u011bnn\u00e9 STAB_SUBJ. Modelovan\u00e9 hodnoty korelovaly se subjektivn\u00edm odhadem s\u00a0koeficientem determinace R2<\/sup> = 0,57 (obr.\u00a04<\/em>).<\/p>\n\"kozeny-5\"<\/a>\n

Obr.\u00a05. Pr\u016ftok na m\u011b\u0159ic\u00ed stanici Morava\u00ad\u2011Moravi\u010dany v\u00a0obdob\u00ed let 2004\u20132014 (zdroj dat: \u010cHM\u00da)
\nFig. 5. Discharge at the Morava\u00ad\u2011Moravi\u010dany gauging station in the period 2004\u20132014 (source of data: Czech hydrometeorological institute)<\/h6>\n

Diskuse<\/h2>\n

Metoda pro v\u00fdpo\u010det stability \u0159\u00ed\u010dn\u00edho d\u0159eva pou\u017eit\u00e1 v\u00a0t\u00e9to studii je pom\u011brn\u011b netradi\u010dn\u00ed. V\u011bt\u0161ina prac\u00ed, kter\u00e9 se zab\u00fdvaly ter\u00e9nn\u00edm v\u00fdzkumem dynamiky \u0159\u00ed\u010dn\u00edho d\u0159eva, pracovala s\u00a0daty o\u00a0mobilit\u011b d\u0159eva zji\u0161t\u011bn\u00e9 rozd\u00edlem v\u00a0dohled\u00e1n\u00ed zna\u010den\u00fdch kus\u016f p\u0159ed a\u00a0po povodni [9, 10, 22\u201325]. Tyto v\u00fdsledky jsou sice interpreta\u010dn\u011b jednozna\u010dn\u00e9, ale jejich z\u00edsk\u00e1n\u00ed je obt\u00ed\u017en\u00e9: d\u0159evo je pot\u0159eba nejprve ozna\u010dit a\u00a0po povodni dohledat. To je \u010dasov\u011b n\u00e1ro\u010dn\u00e9 a\u00a0nar\u00e1\u017e\u00ed to na \u0159adu probl\u00e9m\u016f\u00a0\u2013 po povodni se da\u0159\u00ed dohledat jen \u010d\u00e1st zna\u010den\u00fdch kmen\u016f, co\u017e plat\u00ed i\u00a0pro kmeny, kter\u00e9 z\u016fstaly na m\u00edst\u011b. To m\u016f\u017ee v\u00fdznamn\u011b zkreslovat v\u00fdsledky. Tento probl\u00e9m byl v\u00a0ned\u00e1vn\u00fdch letech \u0159e\u0161en pomoc\u00ed sofistikovan\u00fdch zp\u016fsob\u016f zna\u010den\u00ed RFID \u010dipy [26], nebo dokonce pomoc\u00ed vys\u00edla\u010dek se zabudovan\u00fdm\u00a0GPS p\u0159ij\u00edma\u010dem [12]. Pou\u017eit\u00ed takov\u00fdch metod je v\u0161ak st\u00e1le pracn\u00e9 a\u00a0ne\u00fam\u011brn\u011b drah\u00e9. Hod\u00ed se sp\u00ed\u0161e pro z\u00e1kladn\u00ed v\u00fdzkum obecn\u00fdch z\u00e1konitost\u00ed pohybu d\u0159eva p\u0159i zv\u00fd\u0161en\u00fdch vodn\u00edch stavech. Tyto \u201ep\u0159\u00edm\u00e9 metody\u201c nav\u00edc vypov\u00eddaj\u00ed o\u00a0mobilit\u011b d\u0159eva p\u0159i jedn\u00e9 konkr\u00e9tn\u00ed povod\u0148ov\u00e9 ud\u00e1losti, jej\u00ed\u017e zobecn\u011bn\u00ed je obt\u00ed\u017en\u00e9. Z\u00edsk\u00e1n\u00ed dat z\u00a0p\u0159\u00edm\u00fdch metod je limitov\u00e1no d\u00e9lkou trv\u00e1n\u00ed v\u00fdzkumn\u00fdch projekt\u016f. Pokud nemaj\u00ed v\u00fdzkumn\u00edci v\u00fdjime\u010dn\u00e9 \u0161t\u011bst\u00ed, zachyt\u00ed se sv\u00fdm vzorkem zna\u010den\u00e9ho d\u0159eva jen b\u011b\u017enou povode\u0148 odpov\u00eddaj\u00edc\u00ed \u010dasov\u00e9mu \u00faseku v\u00fdzkumu. Oproti tomu zde popsan\u00e1 metoda je zalo\u017eena na jednor\u00e1zov\u00e9m popisu d\u0159eva vyskytuj\u00edc\u00edho se na lokalit\u011b. Obejde se bez zna\u010den\u00ed i\u00a0zp\u011btn\u00e9ho dohled\u00e1v\u00e1n\u00ed d\u0159eva po povodni. Integruje vliv povodn\u00ed za del\u0161\u00ed \u010dasov\u00e9 obdob\u00ed. Jak dlouh\u00fd tento \u010dasov\u00fd \u00fasek m\u016f\u017ee b\u00fdt, z\u00e1vis\u00ed zejm\u00e9na na rychlosti z\u00e1niku \u0159\u00ed\u010dn\u00edho d\u0159eva na lokalit\u011b. Pokud se v\u00a0minulosti vyskytla v\u00fdznamn\u00e1 povod\u0148ov\u00e1 ud\u00e1lost, je mo\u017en\u00e9 naplaven\u00e9 d\u0159evo analyzovat po dobu jeho \u017eivotnosti. V\u00a0podm\u00ednk\u00e1ch t\u00e9to studie odhadujeme tuto dobu nejm\u00e9n\u011b na p\u011bt let, ale velmi pravd\u011bpodobn\u011b se setk\u00e1v\u00e1me se d\u0159evem star\u0161\u00edm, zvl\u00e1\u0161t\u011b mezi autochtonn\u00edmi kusy. Na obr.\u00a05<\/em> je hydrogram bl\u00edzk\u00e9ho profilu Morava\u00ad\u2011Moravi\u010dany, kter\u00fd je reprezentativn\u00ed pro Moravu na po\u010d\u00e1tku CHKO Litovelsk\u00e9 Pomorav\u00ed. V\u00a0obdob\u00ed p\u011bti let p\u0159ed proveden\u00edm v\u00fdzkumu pro\u0161lo lokalitou n\u011bkolik p\u0159ibli\u017en\u011b jednolet\u00fdch a\u00a0jedna p\u011btilet\u00e1 povode\u0148. V\u00a0cel\u00e9m desetilet\u00ed, kter\u00e9 graf zobrazuje, se vyskytla nejv\u011bt\u0161\u00ed povode\u0148 v\u00a0roce 2006\u2008s\u00a0dobou opakov\u00e1n\u00ed 10\u201320\u00a0let [27]. \u00davahy o\u00a0tom, jestli \u0159\u00ed\u010dn\u00ed d\u0159evo v\u00a0roce 2014 mohlo vypov\u00eddat je\u0161t\u011b o\u00a0povodni z\u00a0roku 2006, jsou re\u00e1ln\u00e9. Jak bylo pops\u00e1no v\u00fd\u0161e, kotven\u00e9 kmeny, kter\u00e9 se na lokalit\u011b vyskytovaly od roku 2003, se do dne\u0161n\u00ed doby prakticky rozpadly a\u00a0podobn\u00fd osud mohl stihnout i\u00a0zbytky kmen\u016f naplaven\u00fdch do vy\u0161\u0161\u00edch parti\u00ed koryta p\u0159i povodni roku 2006. Na druhou stranu \u0159ada zahrani\u010dn\u00edch prac\u00ed z\u00a0m\u00edrn\u00e9ho klimatick\u00e9ho p\u00e1su uv\u00e1d\u00ed dobu zdr\u017een\u00ed d\u0159eva v\u00a0koryt\u011b v\u00a0des\u00edtk\u00e1ch let [4, 28]. Podobn\u011b M\u00e1\u010dka\u00a0aj. [29] byli schopni v\u00a0roce 2009 rozpoznat naplaven\u00e9 a\u00a0autochtonn\u00ed \u0159\u00ed\u010dn\u00ed d\u0159evo v\u00a0akumulac\u00edch z\u00a0povodn\u00ed 2002 a\u00a02006.<\/p>\n

Popsanou metodu je mo\u017en\u00e9 uplatnit pouze na homogenn\u00edm \u00faseku \u0159eky po str\u00e1nce p\u0159\u00edsunu \u0159\u00ed\u010dn\u00edho d\u0159eva. Jestli\u017ee porovn\u00e1v\u00e1me naplaven\u00e9 d\u0159evo s\u00a0autochtonn\u00edm, m\u011bla by existovat alespo\u0148 hypotetick\u00e1 mo\u017enost naplaven\u00ed kmen\u016f stejn\u00fdch rozm\u011br\u016f, jako je autochtonn\u00ed d\u0159evo. V\u00a0podm\u00ednk\u00e1ch lokality byl mo\u017en\u00fd p\u0159\u00edsun vzrostl\u00fdch strom\u016f z\u00a0\u00faseku 500 metr\u016f neupraven\u00e9ho koryta k\u00a0mostu Mlade\u010d\u00ad\u2011Nov\u00e9 Z\u00e1mky a\u00a0d\u00e1le z\u00a0asi 2,5\u2008km dlouh\u00e9ho \u00faseku \u010d\u00e1ste\u010dn\u011b upraven\u00e9ho koryta a\u017e k\u00a0jezu \u0158imice. V\u00a0b\u0159ehov\u00fdch porostech se vyskytuje podobn\u00fd vzrostl\u00fd porost jako na Vrapa\u010di, i\u00a0kdy\u017e intenzita b\u0159ehov\u00e9 eroze je d\u00edky stabilizaci b\u0159eh\u016f ni\u017e\u0161\u00ed. V\u00a0charakteristik\u00e1ch polohy hodnot\u00ed metoda vlastnosti naplaven\u00e9ho d\u0159eva \u201epo naplaven\u00ed\u201c. V\u00a0tom se odli\u0161uje od metod p\u0159\u00edm\u00e9ho sledov\u00e1n\u00ed, kdy je hodnocena poloha kusu p\u0159ed odplaven\u00edm. Lze ale p\u0159edpokl\u00e1dat, \u017ee v\u011bt\u0161ina naplaven\u00e9ho d\u0159eva pat\u0159ila mezi naplaven\u00e9 d\u0159evo i\u00a0v\u00fd\u0161e v\u00a0povod\u00ed a\u00a0uvolnila se z\u00a0podobn\u00fdch pozic, jak\u00e9 byly popisov\u00e1ny na zkouman\u00e9 lokalit\u011b. P\u0159i hodnocen\u00ed v\u00fdznamu d\u00e9lky kmene nen\u00ed uva\u017eov\u00e1na mo\u017en\u00e1 fragmentace b\u011bhem povodn\u011b. Takto vznikl\u00e9 kr\u00e1tk\u00e9 kusy by m\u011bly ni\u017e\u0161\u00ed pravd\u011bpodobnost stability ne\u017e kmen p\u016fvodn\u00ed. P\u0159esto byl tento vliv zanedb\u00e1n, zejm\u00e9na proto, \u017ee k\u00a0hodnocen\u00ed mo\u017enosti fragmentace jednotliv\u00fdch kus\u016f nebyly z\u00edsk\u00e1ny \u017e\u00e1dn\u00e9 \u00fadaje. Lze p\u0159edpokl\u00e1dat, \u017ee k\u00a0rozl\u00e1m\u00e1n\u00ed doch\u00e1z\u00ed \u010dast\u011bji u\u00a0\u0159\u00ed\u010dn\u00edho d\u0159eva ve vy\u0161\u0161\u00edm stupni rozkladu. Na lokalit\u011b v\u0161ak p\u0159evl\u00e1dalo d\u0159evo nezetlel\u00e9, co\u017e byl druh\u00fd d\u016fvod, pro\u010d nebyla mo\u017enost fragmentace zohledn\u011bna.<\/p>\n

Podobn\u00fd p\u0159\u00edstup p\u0159i hodnocen\u00ed akumulac\u00ed d\u0159eva pou\u017eili M\u00e1\u010dka\u00a0aj. [29] v\u00a0NP Podyj\u00ed. Naplaven\u00e9 kusy d\u0159eva se vyzna\u010dovaly statisticky v\u00fdznamnou krat\u0161\u00ed d\u00e9lkou a\u00a0men\u0161\u00edm pr\u016fm\u011brem ne\u017e kmeny autochtonn\u00ed. Oproti autochtonn\u00edm kus\u016fm postr\u00e1daly naplaven\u00e9 kusy ko\u0159eny i\u00a0v\u011btve. Abbe a\u00a0Montgomery [5] popsali stabiln\u00ed tzv.\u00a0kl\u00ed\u010dov\u00e9 kusy akumulac\u00ed pomoc\u00ed jejich d\u00e9lky a\u00a0pr\u016fm\u011bru b\u00e1ze kmene. Na z\u00e1klad\u011b t\u011bchto dvou parametr\u016f byly stabiln\u00ed kusy dob\u0159e odli\u0161iteln\u00e9 od naplaven\u00e9ho \u0159\u00ed\u010dn\u00edho d\u0159eva. D\u00e9lka stabiln\u00edch kus\u016f byla obvykle v\u011bt\u0161\u00ed ne\u017e polovina \u0161\u00ed\u0159ky koryta. Pr\u016fm\u011br b\u00e1ze t\u011bchto kmen\u016f byl obvykle v\u011bt\u0161\u00ed ne\u017e polovina hloubky koryta.<\/p>\n

V\u00fdznam rozm\u011br\u016f kmene pro stabilitu byl ji\u017e pops\u00e1n v\u00a0mnoha studi\u00edch z\u00a0ter\u00e9nu i\u00a0hydraulick\u00fdch laborato\u0159\u00ed [6, 10, 13]. P\u0159i zobecn\u011bn\u00ed je d\u00e9lka kmene obvykle uv\u00e1d\u011bna relativn\u011b vzhledem k\u00a0\u0161\u00ed\u0159ce koryta. Kmeny s\u00a0hodnotou bl\u00edzko 1 a\u00a0del\u0161\u00ed jsou pova\u017eov\u00e1ny za velmi stabiln\u00ed. V\u00a0tomto se na\u0161e v\u00fdsledky shoduj\u00ed, nebo\u0165 nejdel\u0161\u00ed popsan\u00e9 kmeny s\u00a0d\u00e9lkami 25\u201330 metr\u016f s\u00a0vysokou pravd\u011bpodobnost\u00ed stability se shoduj\u00ed d\u00e9lkou s\u00a0\u0161\u00ed\u0159\u00ed koryta, kter\u00e1 je v\u00a0pr\u016fm\u011bru 28\u2008m. Oproti v\u00fdsledk\u016fm mnoha jin\u00fdch studi\u00ed nebyla zji\u0161t\u011bna p\u0159\u00edli\u0161 v\u00fdznamn\u00e1 z\u00e1vislost stability na pr\u016fm\u011bru kmene. Je to patrn\u011b zp\u016fsobeno t\u00edm, \u017ee hloubka vody na lokalit\u011b b\u011bhem zv\u00fd\u0161en\u00fdch pr\u016ftok\u016f p\u0159esahuje pr\u016fm\u011br nejsiln\u011bj\u0161\u00edch vyskytuj\u00edc\u00edch se kus\u016f. Siln\u011bj\u0161\u00ed z\u00e1vislost by mohla b\u00fdt prok\u00e1z\u00e1na, pokud by byly m\u011b\u0159eny tlou\u0161\u0165ky b\u00e1z\u00ed kmen\u016f nebo nejl\u00e9pe pr\u016fm\u011br ko\u0159enov\u00e9ho syst\u00e9mu [5]. Rovn\u011b\u017e m\u00edra zazemn\u011bn\u00ed a\u00a0p\u0159\u00edtomnost ko\u0159en\u016f jsou obecn\u011b zn\u00e1m\u00e9 znaky podporuj\u00edc\u00ed stabilitu kmene [5, 10, 11]. Ob\u011b charakteristiky jsou sou\u010d\u00e1st\u00ed fin\u00e1ln\u00edho modelu. Oproti tomu p\u0159\u00edtomnost koruny (siln\u00fdch v\u011btv\u00ed) a\u00a0paraleln\u00ed orientace s\u00a0proudem byly statisticky v\u00fdznamn\u00e9 pouze jednotliv\u011b, ale nikoliv v\u00a0logistick\u00e9m regresn\u00edm modelu v\u00a0\u201ekonkurenci\u201c ostatn\u00edch charakteristik.<\/p>\n

Ulo\u017een\u00ed naplaven\u00fdch kus\u016f v\u00a0akumulaci \u0159\u00ed\u010dn\u00edho d\u0159eva bylo charakteristick\u00e9 pro rozli\u0161en\u00ed mezi autochtonn\u00edmi a\u00a0naplaven\u00fdmi kusy, ale statisticky v\u00fdznamn\u00e9 pouze v\u00a0univariantn\u00edm modelu. Tato vlastnost vych\u00e1z\u00ed z\u00a0povahy naplaven\u00e9ho d\u0159eva, kter\u00e9 se p\u0159i povodni hromadn\u011b zachyt\u00e1v\u00e1 na p\u0159ek\u00e1\u017ek\u00e1ch [3]. Tento vztah nelze jednodu\u0161e interpretovat obr\u00e1cen\u011b jako z\u00e1vislost, tedy \u017ee d\u0159evo ulo\u017een\u00e9 uvnit\u0159 akumulac\u00ed je n\u00e1chyln\u011bj\u0161\u00ed k\u00a0odplaven\u00ed. Stabilita akumulace jako celku je z\u00e1visl\u00e1 na tzv.\u00a0kl\u00ed\u010dov\u00fdch kusech, rozm\u011brn\u00fdch kmenech, kter\u00e9 le\u017e\u00ed v\u00a0jej\u00edm z\u00e1kladu [5]. Ostatn\u00ed sou\u010d\u00e1sti akumulace, zejm\u00e9na tzv.\u00a0voln\u00e9 kusy, se mohou p\u0159i zv\u00fd\u0161en\u00fdch vodn\u00edch stavech snadno mobilizovat. Stabilita jednotliv\u00fdch prvk\u016f akumulace se tedy m\u016f\u017ee zna\u010dn\u011b li\u0161it, i\u00a0kdy\u017e akumulace samotn\u00e1 je stabiln\u00ed po \u0159adu let. Rozli\u0161en\u00ed jednotliv\u00fdch kus\u016f \u0159\u00ed\u010dn\u00edho d\u0159eva mezi funk\u010dn\u00ed prvky akumulac\u00ed v\u0161ak nebylo v\u00a0t\u00e9to studii provedeno.<\/p>\n

Jako novinka mezi identifikovan\u00fdmi vlastnostmi ovliv\u0148uj\u00edc\u00edmi stabilitu kmene se jev\u00ed v\u00fdznam \u00farovn\u011b ulo\u017een\u00ed kmene v\u016f\u010di b\u011b\u017en\u00e9 hladin\u011b. Interpretace je z\u0159ejm\u00e1. Kmeny vyskytuj\u00edc\u00ed se po v\u011bt\u0161inu roku pod hladinou jsou nas\u00e1kl\u00e9 vodou, jejich specifick\u00e1 hustota je vy\u0161\u0161\u00ed ne\u017e hustota vody, co\u017e v\u00fdznamn\u011b sni\u017euje n\u00e1chylnost k\u00a0odplaven\u00ed. Oproti tomu d\u0159evo naplaven\u00e9 na vrcholech akumulac\u00ed \u010di \u0161t\u011brkov\u00fdch lavic snadno vysych\u00e1 a\u00a0zvy\u0161uje tak pravd\u011bpodobnost dal\u0161\u00edho odplaven\u00ed. Hustota d\u0159eva je b\u011b\u017en\u011b uva\u017eov\u00e1na p\u0159i modelov\u00e1n\u00ed pohybu jednotliv\u00fdch kus\u016f, obvykle v\u00a0kombinaci s\u00a0pr\u016fm\u011brem kmene [13]. Kombinace t\u011bchto hodnot se pou\u017e\u00edv\u00e1 k\u00a0stanoven\u00ed tzv.\u00a0kritick\u00e9 hloubky, p\u0159i kter\u00e9 doch\u00e1z\u00ed k\u00a0odplaven\u00ed voln\u00fdch kus\u016f d\u0159eva. Tento vztah sice dob\u0159e funguje v\u00a0m\u011blk\u00fdch a\u00a0sklonit\u00fdch toc\u00edch ni\u017e\u0161\u00edch \u0159\u00e1d\u016f [10], v\u00a0hlubok\u00fdch meandruj\u00edc\u00edch korytech se v\u0161ak uplat\u0148uj\u00ed i\u00a0jin\u00e9 mechanismy stabilizace [30]. Zv\u00fd\u0161en\u00ed specifick\u00e9 hustoty d\u0159eva zde p\u016fsob\u00ed jako pozitivn\u00ed zp\u011btn\u00e1 vazba.<\/p>\n

Podobn\u00fdm zp\u016fsobem jako v\u00a0t\u00e9to studii pou\u017eili pro anal\u00fdzu pohyblivosti \u0159\u00ed\u010dn\u00edho d\u0159eva logistick\u00fd regresn\u00ed model nap\u0159.\u00a0Van der Nat\u00a0aj. [8] na \u0159ece Tagliamento v\u00a0It\u00e1lii, Wohl a\u00a0Goode [10] na horsk\u00fdch potoc\u00edch v\u00a0Coloradu a\u00a0Merten\u00a0aj. [11] na toc\u00edch v\u00a0okol\u00ed Ho\u0159ej\u0161\u00edho jezera v\u00a0USA. Prom\u011bnn\u00e9 fin\u00e1ln\u00edch model\u016f se li\u0161\u00ed podle metodik t\u011bchto prac\u00ed a\u00a0popisovan\u00fdch charakteristik \u0159\u00ed\u010dn\u00edho d\u0159eva. Van der Nat\u00a0aj. [8] popsali modelem stabilitu d\u0159evn\u00edch akumulac\u00ed v\u00a0z\u00e1vislosti na velikosti povodn\u011b, velikosti akumulace a\u00a0jej\u00ed poloze. Model Wohl a\u00a0Goode [10] zahrnoval statisticky v\u00fdznamn\u00e9 parametry: d\u00e9lka kmene, pr\u016fm\u011br kmene a\u00a0typ ulo\u017een\u00ed (nap\u0159\u00ed\u010d p\u0159es koryto, voln\u00fd, pod\u00e9l b\u0159ehu). Merten a\u00a0kol. [11] vyhodnotil v\u00a0modelu sedm v\u00fdznamn\u00fdch faktor\u016f, mezi kter\u00e9 pat\u0159ila d\u00e9lka kmene, zazemn\u011bn\u00ed a\u00a0p\u0159\u00edtomnost ko\u0159en\u016f. Podobn\u011b jako v\u00a0t\u00e9to studii pou\u017eil upraven\u00fd vzorec logistick\u00e9 regrese pro vykreslen\u00ed k\u0159ivek pravd\u011bpodobnosti stability pro dan\u00fd parametr.<\/p>\n

Modelovan\u00e9 hodnoty pravd\u011bpodobnosti samoz\u0159ejm\u011b nemohou b\u00fdt p\u0159ij\u00edm\u00e1ny s\u00a0nekritickou d\u016fv\u011brou. Ka\u017ed\u00fd model plat\u00ed pouze v\u00a0podm\u00ednk\u00e1ch dat, ze kter\u00fdch byl vytvo\u0159en. I\u00a0kdy\u017e p\u0159ipust\u00edme, \u017ee p\u0159\u00edrodn\u00ed podm\u00ednky na lokalit\u011b Vrapa\u010d jsou reprezentativn\u00ed pro hlavn\u00ed koryto Moravy v\u00a0Litovelsk\u00e9m Pomorav\u00ed, modelovan\u00e9 hodnoty by m\u011bly b\u00fdt pou\u017eity sp\u00ed\u0161e pro demonstraci relativn\u00edho v\u00fdznamu jednotliv\u00fdch popisn\u00fdch charakteristik. Takov\u00fd p\u0159\u00edklad m\u016f\u017ee m\u00edt v\u00fdznam p\u0159i vzd\u011bl\u00e1v\u00e1n\u00ed pracovn\u00edk\u016f pov\u011b\u0159en\u00fdch \u00fadr\u017ebou koryta. Z\u00a0tohoto pohledu je zaj\u00edmav\u00fd vztah zn\u00e1zorn\u011bn\u00fd na obr.\u00a04<\/em>. Pom\u011brn\u011b dobr\u00e1 m\u00edra korelace mezi subjektivn\u011b ur\u010denou stabilitou \u0159\u00ed\u010dn\u00edho d\u0159eva a\u00a0jej\u00edch modelovan\u00fdch hodnot navozuje my\u0161lenku, \u017ee by k\u00a0z\u00e1kladn\u00edmu odhadu stability sta\u010dilo kvalifikovan\u00e9 posouzen\u00ed pracovn\u00edka s\u00a0m\u00edstn\u00ed znalost\u00ed.<\/p>\n

Z\u00e1v\u011br<\/h2>\n

Metodou ter\u00e9nn\u00edho popisu \u0159\u00ed\u010dn\u00edho d\u0159eva byly z\u00edsk\u00e1ny charakteristiky typick\u00e9 pro stabiln\u00ed (autochtonn\u00ed) kmeny a\u00a0nestabiln\u00ed (naplaven\u00e9) \u0159\u00ed\u010dn\u00ed d\u0159evo. Jak vyplynulo z\u00a0univariantn\u00edho testov\u00e1n\u00ed jednotliv\u00fdch prom\u011bnn\u00fdch, stabilitu \u0159\u00ed\u010dn\u00edho d\u0159eva ovliv\u0148uje \u0159ada charakteristik rozm\u011br\u016f, tvaru, mechanick\u00fdch vlastnost\u00ed a\u00a0ulo\u017een\u00ed d\u0159eva v\u00a0koryt\u011b. Pomoc\u00ed logistick\u00e9ho regresn\u00edho modelu byly jako charakteristiky nejv\u00fdznamn\u011bj\u0161\u00ed pro stabilitu ur\u010deny: d\u00e9lka kmene, v\u00fd\u0161ka ulo\u017een\u00ed v\u016f\u010di b\u011b\u017en\u00e9 hladin\u011b, m\u00edra zazemn\u011bn\u00ed, p\u0159\u00edtomnost ko\u0159en\u016f a\u00a0zda jde o\u00a0kmen \u017eiv\u00fd \u010di odum\u0159el\u00fd. Pokud je z\u00a0d\u016fvod\u016f zachov\u00e1n\u00ed p\u0159irozen\u00e9ho stavu koryta po\u017eadov\u00e1no ponech\u00e1n\u00ed stabiln\u00edch kus\u016f \u0159\u00ed\u010dn\u00edho d\u0159eva, m\u011bly by b\u00fdt tyto charakteristiky respektov\u00e1ny. Podobn\u011b lze vyu\u017e\u00edt analogii se stabiln\u00edmi kmeny p\u0159i instalaci z\u00e1m\u011brn\u011b vkl\u00e1dan\u00e9ho d\u0159eva do koryt vodn\u00edch tok\u016f. Kone\u010dn\u011b modelovan\u00e9 pravd\u011bpodobnosti stability mohou slou\u017eit k\u00a0lep\u0161\u00edmu pozn\u00e1n\u00ed dynamiky \u0159\u00ed\u010dn\u00edho d\u0159eva na konkr\u00e9tn\u00ed lokalit\u011b a\u00a0z\u00edsk\u00e1n\u00ed citu pro ter\u00e9nn\u00ed odhad stability \u0159\u00ed\u010dn\u00edho d\u0159eva.<\/p>\n

Pod\u011bkov\u00e1n\u00ed<\/h3>\n

D\u011bkuji pracovn\u00edk\u016fm Spr\u00e1vy CHKO Litovelsk\u00e9 Pomorav\u00ed, zejm\u00e9na pan\u00ed Olze \u017dern\u00ed\u010dkov\u00e9 za vst\u0159\u00edcnost p\u0159i zaji\u0161t\u011bn\u00ed ter\u00e9nn\u00edho v\u00fdzkumu a\u00a0Jan\u011b Timkov\u00e9 z\u00a0firmy DataTalk za pomoc se statistickou anal\u00fdzou.<\/em><\/p>\n","protected":false},"excerpt":{"rendered":"

Wood is an important part of streams, affecting their hydraulic, morphological and biological features. For the stream management and flood risk reduction it is important to know the characteristics that increase the stability of wood and prevent its flushing at high water stages. <\/p>\n","protected":false},"author":8,"featured_media":2643,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[91,86],"tags":[674,676,677,675,673],"coauthors":[597],"class_list":["post-2719","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-applied-ecology","category-hydraulics-hydrology-and-hydrogeology","tag-floods","tag-litovelske-pomoravi-pla","tag-logistic-regression","tag-morava-river","tag-woody-debris"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/posts\/2719","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/users\/8"}],"replies":[{"embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/comments?post=2719"}],"version-history":[{"count":2,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/posts\/2719\/revisions"}],"predecessor-version":[{"id":30360,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/posts\/2719\/revisions\/30360"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/media\/2643"}],"wp:attachment":[{"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/media?parent=2719"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/categories?post=2719"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/tags?post=2719"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/www.vtei.cz\/en\/wp-json\/wp\/v2\/coauthors?post=2719"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}