You are on page 1of 30


Bioclimate-vegetation interrelations along the Pacific rim of North America

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Manuel Peinado1* · Gustavo Díaz1 · José Delgadillo2 · Francisco Manuel Ocaña-Peinado3 · Miguel Ángel Macías4 · Juan Luis Aguirre1 · Alejandro Aparicio1

Cátedra de Medio Ambiente, Universidad de Alcalá, Alcalá de Henares, Spain. 2Facultad de Ciencias, Universidad Autónoma de Baja California, Ensenada, BC, Mexico. 3Departamento de Estadística e Investigación Operativa, Universidad de Granada, Granada, Spain. 4Departamento de Ciencias Ambientales, Universidad de Guadalajara, Guadalajara, JAL, Mexico. Corresponding author:

ABSTRACT This study was designed to examine relationships between climate and vegetation of the Pacific rim of North America, from the Mediterranean deserts of California to Alaska’s boreal taiga. Relations were inferred from temperature and rainfall data recorded at 457 weather stations and by sampling the vegetation around these stations. Climate data were used to construct climatograms, calculate forty one variables and detect main latitudinal and longitudinal gradients. In order to identify the best functions able to relate our variables, polynomial and non-polynomial regressions were performed. The k-means algorithm was the clustering method used to validate the variables that could best support our bioclimatic classification. The variable that best fitted our classification was finally used to prepare a discriminatory key for bioclimates. Across this extensive area three macrobioclimates were identified, Mediterranean, Temperate and Boreal, within which we were able to distinguish nine bioclimates. Finally, we relate the different types of potential natural vegetation to each of these bioclimates and describe their floristic composition and physiognomy. Keywords Bioclimatology · Boreal forests · Mediterranean Vegetation · Plant Formations · Temperate Rainforests · Zonobiomes


29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78

1. Introduction Vegetation science, like any other science, uses classification to understand the laws of Nature, and organize knowledge. Bioclimatic classification schemes attempt to relate meteorological data to the geographic distribution areas of living organisms, mainly of single plant species or plant communities. World zones showing marked climatic gradients can be the best laboratory to assess whether the vegetation’s distribution follows a climatic pattern, as reflected by meteorological data. The Pacific coast of North America with a general climate that is driven by oceanic currents (the Humboldt Current in the south, the Aleutian-California current in the north) is an area of particular bioclimatic interest owing to its drastic north-south gradient embracing tundras, coastal rainforests, temperate and Mediterranean forests, Mediterranean scrubs, and deserts. Although this is well-known and mentioned in practically every general geobotanical survey (for references see [1] and [2]), we still lack a detailed description of the bioclimatic patterns and transitions that occur along this dramatic gradient and its impacts on the distribution of vegetation types. Several published reports exist on the bioclimatology of North America [3, 4, 5, 6]. These surveys have been based on the classification into zonobiomes of Walter [7] and on the successive bioclimatic schemes proposed by Rivas-Martínez, which have led to a worldwide bioclimatic classification system [8]. Several authors have also described the vegetation of the area examined here, but so far no investigation within a given study area has tried to relate the types of vegetation identified in field work to climate data provided by the meteorological stations of the selected area. Following Walter’s and Rivas-Martinez’s concepts, but using real field data, the main objective of the present study was to identify the climate variables that would best discriminate existing bioclimates and potential natural vegetation along the Pacific rim of North America between the Mexico-United States border in California and Alaska. A bioclimate was here defined as an eclectic biophysical model described by means of climate variables and vegetation types [4], mainly those regarded as potential natural vegetation (PNV), i.e., the climatophilous plant community that would become established if all successional sequences were completed without human interference under the present climate and edaphic conditions [9]. As support for this bioclimatic interpretation, two statistical methods –regression analysis and cluster analysis– were used to determine the lowest possible number of simple climate factors that could define and predict the vegetation distribution changes observed. 2. Materials and methods 2.1. Study area Overlooking the Pacific Ocean, the study area covers an airline distance of approximately 4,100 km from Cook Inlet, AK (the northernmost weather station examined was Skwentna, at 61º58’N) to southern CA (where the southernmost station was Campo, at 32º37’N). Its western limit is Port Heiden, AK, at 158º 37’W, while the easternmost site examined was 116º25’W, in Borrego Desert, CA. The whole area forms part of the largest and highest of North American physiographic systems, the Pacific Border System [1], which is the backdrop for most of the ocean’s shores. Two parallel belts of mountains dominate the area. In the north, the Alaska, Chugach and Saint Elias ranges of AK and British Colombia (BC), the BC Coastal Ranges and the Insular Mountains of the islands of Queen Charlotte and Vancouver, constitute a seaward fringe of peaks. To the south, the Coast Ranges between northern CA, Oregon (OR) and Washington (WA) (including the Olympics) dominate the outer coastal topography. These coastal mountains act as effective barriers to the moisture-laden westerly winds, and rainshadow plateaus, depressions and valleys form downwind (Fig. S1 as supplementary material at From the Fraser River Valley, in southeast BC, southwards, across WA and OR, the piedmonts of the Cascades define the eastern boundary.


79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131

According to the macrobioclimate (MB) classification system used in our previous studies for western North American zonobiomes, the study area shows three broad MBs: Boreal, Temperate and Mediterranean [3]. The central zone of the Pacific coast, between OR and BC, shows a Temperate MB, characterized by mild wet winters, cool relatively dry summers, and a long frostfree season. In the northwestern corner of BC and especially along the coast of southeastern AK, the winters are colder and a very oceanic variant of the Boreal MB prevails. Zones of highly continental Boreal MB occur in the lee of the Coastal Ranges in northwestern BC and interior AK. The interior climate is very continental: summers are short but relatively warm; winters are long, extremely cold, and dry. Also, although precipitation is light, evaporation is minimal and permafrost impedes drainage, so bogs and wetlands are common [10]. During all seasons, the prevailing westerly winds in the study area are moisture-laden due to their journey over relatively warm seas. In winter, the land is colder than the ocean, and precipitations along coastal lowlands are frequent. In the southern part of the study area, the land along the coast is warmer than the ocean during the summer and, consequently, when the wind reaches the low coastal area there is little or no precipitation and the Mediterranean MB dominates. Dice [11] included the boreal and temperate climate zones within the Hudsonian (continental boreal), Sitkan and Aleutian (oceanic boreal) and Oregonian (temperate) biotic provinces. The winter rain zone corresponds to the Californian Region, whose provinces, Northern California and Southern California, are included in the study area. For a more detailed phytogeographical classification, see [12, 13]. 2.2. Climate and vegetation Before conducting fieldwork, a geographic information system (GIS) was designed using ArcMap 9.3 software and the digital terrain model (DTM) of the University of California, Davis ( Into this GIS, we entered all the available weather stations providing climatological normals [14] for the study area. To avoid large deviations due to continentality and altitude effects, an essential criterion in the final selection process was that every station had to be less than 100 km away from the sea and 1,000 m below sea level. Four hundred and fifty-seven weather stations fulfilled these requirements (Fig. 1), and climate data for each station were compiled from [15] for the US stations, and [16] for the Canadian ones. Next, by combining DTM and satellite images ( three variables were calculated for each station: 1) distance to the shoreline in a straight line always in a westerly direction; 2) orientation with respect to prevailing moisture-laden westerlies; and 3) orographic position. Accordingly, the stations were then grouped into three categories (Fig. 2): COAST, grouping stations on coastal plains, approximately at sea level, directly accessed by wet fronts. One hundred and sixty-two stations included in this category show the highest rainfall records (average P: 2088 mm); WIND, describing stations on the windward slopes of the Coastal Ranges and Cascades. Rainfall in these 60 stations is lower than in the previous group (average P: 1495 mm), but in areas with less rain (Fig. 2: Mediterranean and Submediterranean) this type or orographic rainfall prevails; and LEE, stations influenced by the rainshadow. Rainfall records in those 235 stations are significantly lower (average P: 946 mm). Four maps showing DTM and station categories are available as supplementary material (Maps S1 to S4). Using climate data from weather stations, several parameters and indices were calculated (Table 1). Besides the indices calculated by us, we also used some of those included in the classification schemes of Holdridge [17], Rivas-Martínez [8], Thornthwaite’s index calculated by the method of Dingman [18], and the climatograms of Walter and Lieth [7] which were constructed using the program BIOCLIMA (Alcaraz pers. comm.). Climate data, variables and indices for each station are available as electronic supplementary material (Table S1). Each weather station was initially assigned to a particular PNV using both bibliographical sources and our field data acquired since 1989. In subsequent fieldwork, this classification into vegetation types was checked for 440 of the weather stations. Fieldwork was conducted at sites near each station selected by examining satellite images ( to ensure the presence of natural vegetation that was relatively well preserved. These sites were visited from


and geological substrate. In order to identify the best regression function able to relate our variables. It is the simplest and most commonly used algorithm that employs a squared error criterion. The adjusted R-squared coefficient (R2) was the criterion used to select the best regression in each case. it calculates a partition of patterns in k clusters [23]. Once we had defined the groups identified as BIOs through prior climate analyses and fieldwork. These seasonal changes are also reflected in the increasing latitudinal and longitudinal gradients observed in OTI2w. The variable that best fitted our hypothesis was finally used to prepare a discriminatory key for MBs and BIOs. Ts2 and Tp decrease with increasing latitude and decreasing longitude. At each site. Most dramatic effects were detected for the interior boreal stations of Alaska. 2. %Psu. OTI3 and OTI4 (Fig. 3a and 3b. Results Despite the influence of local factors such as altitude.3. through a mathematical function. satellite images. whereas %Pwi and %Psp rains decrease (Fig. M. %Pau and %Pcm1 increase as latitude increases. Statistical analysis Climate data. the best regression detected was using polynomial functions. non-hierarchical data clustering method suitable for classifying large amounts of data into a given number . our results reveal several latitudinal and longitudinal gradients in the study area (Tables 2 and 3): 1. In a limited number of cases no polynomial functions to relate variables appeared. slope. Polynomial and no polynomial functions. Since statistical analysis should be conducted when employing heuristics to estimate the probability of incorrect outcomes [22].132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 2003 to 2011. 3d). each station was definitively assigned to one PNV. Among the different CLA techniques. According to the dominant species and vegetation structure. Provided with a set of n numeric objects (number of stations in our case) and an integer number k (BIOs or MBs in our case). which showed a seasonal rhythm whereby summer or autumn were the rainiest (Pau > Psu > Pwi > Psp or Psu >Pau > 4 . OTI2. the k-means algorithm is a partitioned. In most of our analysis. The former emphasizes gradients of continuous variation while the latter produces discontinuous groups. The most significant trend in rainfall detected was its seasonal pattern. Two temperature gradients exist whereby T. 3. 2. These regression types have been used to describe phenomena in many scientific fields [21]. the relationship between the independent variable x (climate variables in our study) and the dependent variable y (latitude and longitude in our study). the stronger the relationship between the variables considered. parameters and indices in Table 1 were used as variables in the statistical tests: Regression Analysis and Cluster Analysis (CLA). following the basic life forms of Raunkiaer adapted by Ellenberg and MuellerDombois [19]. the PNV was inferred using bibliographical sources. Data on vegetation were obtained by plotless sampling. the goal of CLA was to identify the climate variable or set of climate variables that could best distinguish some groups from others and to confirm our hypothesis of separation into BIOs. Hence. aerial photograms and by comparing their weather data to those of the stations visited Plant nomenclature follows [20]. aspect and orography on the rainfall and temperature data recorded for each sampled station. together with their respective R2. whereas Tn increases approximately north of 58º (Figs. The higher the value of this coefficient. we first subjectively selected a stand according to the homogeneity of its physical features. aspect. taking censuses of vascular plants at each stand. The environmental data collected for each stand were elevation. polynomial and non-polynomial regressions were performed. m. through CLA we checked the indices and algorithms heuristically obtained by us in previous non-statistical analyses. type of soil. Tsu. For the 17 stations that could not be sampled in the field. 3c). Regression analysis is a common statistical technique to express. We recorded the dominant species and their corresponding ecophysiognomic characters. are represented in Tables 2 and 3. vegetation structure and species dominance. for abbreviations of climate variables see Table 1).

Of less importance is the mismatch in CLA 2. Main features of the PNV types are summarized in Table 4. Picea sitchensis. The other 95 stations that show two dry months outside California share two main features: a) Arbutus menziesii.728 mm in Little Port Walter. Thuja plicata or Tsuga heterophylla. the nine bioclimates discerned in the study area can be distinguished by seven variables (Table 7) that are statistically reliable (Table 8). whereas it should lack typical temperate species such as Abies grandis. Precipitation and temperature gradients cause a similar bioclimatic gradient (Fig. Besides. the first step in our statistical analysis was to find an algorithm capable of discriminating Mediterranean and Temperate MBs. 6) reflected by different types of PNV (Fig. Mediterranean. in which winter is always the rainiest season and summer the driest (Fig. In Mediterranean-Temperate zonoecotones we used a double criterion to consider a station as Mediterranean: a) climatically. due to 9 temperate and boreal stations of the bordering oceanic zone between British Columbia and Alaska. is characterized by. All those Californian stations fulfilled these conditions. 3. whose climatic PNV consists mainly of sclerophyllous woody plants. at least. 4).6 mm in Pearblossom. respectively [7]. A dichotomous key to PNV based on climate variables is provided as Table 5.182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 Pwi > Psp) contrasting with the rest of the study area. There. and b) the PNV should be dominated or co-dominated by evergreen trees or shrubs along with some conifers and late. it should show at least two consecutive dry-warm months. 7). It can be observed that even within a high percentage of correct classifications most deviations between real results and CLA results occur in two cases: a) In the case of MB. rainfall is so high and summer drought so brief that the forests can be regarded as ZE IV/V [7: 166]. 4. which are regarded as the typical sclerophyllous vegetation of Mediterranean California. Of the initial 457 weather stations. the Mediterranean MB. subzonobiomes and zonoecotones can be related to some of the bioclimates described in this article. According to CLA. when trying to differentiate in frontier zones between Dry and Humid bioclimates (CLA 5 and 6). when trying to numerically distinguish between Temperate Dry (Submediterranean) and Mediterranean-Humid (CLA 1 and 3) stations. Inland. Walter distinguishes two climate areas or subzonobiomes: Boreal Cold-Oceanic and Boreal Cold-Continental. sclerophyllous region of western North America. A. In Table 8. One hundred and fifty-seven of these are in California. the general gradient of increasing rainfall is masked by the rainshadow effect in the lee of the coastal mountains (Fig. Hence. the winter-rain region extends northward to British Columbia through the interior valleys of Oregon. stations described as “correctly classified” are those found to group together in the CLA with those expected according to our prior hypothesis of belonging to a given MB or bioclimate. 5). always in the lee of the coastal mountains. Alaska). menziesii) and 5 . phytogeographical. b) In the case of bioclimate. Within this last ZB. and ecological features as Table 6. 252 show at least two consecutive dry-warm months. Pseudotsuga menziesii var. Walter’s classification includes two zonoecotones ZE V-VIII and ZE IV-V. as discussed below. to 5. By definition. detected in the study area and. Discussion Seasonal patterns of rainfall support the zonobiomes defined by Walter [7] for the study area: ZB IV (Mediterranean). and other key based on floristic. The Mediterranean zone of North America spreads from the south of Oregon along the entire California coast until the northwestern corner of Baja California [3]. two consecutive dry months during the warmest period in the year. California. and VIII (Boreal). menziesii (hereafter P. Temperate and Boreal. Zonobiomes clearly correspond to the three MBs.or drought deciduous hardwoods. considered the typical Mediterranean. amabilis. Although there is a clear increase in precipitation with latitude (P ranges from 135. Washington and the Puget Sound. ZB V (Warm-Temperate). a month is defined as dry if its precipitation is less than twice the temperature (P < 2T) both measured in mm and ºC.

Tn>22. Its presence is reflected by the appearance of Picea glauca forests and Picea mariana muskegs. and understory species typical of temperate areas are common under the forest canopy.8 in the Temperate MB. such areas are characterized by the presence of A. showing a Boreal Oceanic climate according to Walter’s classification or Mesothermal after Köppen’s system. Numerically. whose physiognomy is that of a savannah or a woodland dominated by A.233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 Quercus garryana are always the dominant trees. As a result. Around these 95 stations outside CA. plicata co-dominate. These were included within the Boreal Continental bioclimate. showing a Boreal Continental climate according to Walter’s classification. The flora of these areas has a floristic composition that is intermediate between temperate and boreal. rainshadow zones with warm. In the 79 northernmost and coldest stations Tn>1. Accordingly. and b) they lie in the lee of the Coastal and Vancouver ranges. P. Thuja plicata. menziesii and Q. 2.8 is best at discriminating both groups of stations. precipitation during the warmest summer months is higher than in the first group. sclerophyllous species are common under the forest canopy. menziesii. wet winters. Puget Trough and interior valleys of Southwestern Oregon. this time discrimination between the two groups is possible using OTI2w values. dry summers and mild. garryana. whose distribution areas correspond to that of a “modified Mediterranean climate” [24]. menziesii and Q. menziesii. T. PNV in this province is dominated by typical temperate trees such as Abies amabilis. is a wide region in which the Temperate climate dominates until we reach the border between OR and CA. Thus. the Arbutus menziesii-Quercus garryana landscape corresponds to the drier climate of the Willamette Valley. all stations showing an OTI3<1. the best indices for discriminating boreal and temperate stations are negative (Tn) and positive (Tp) temperatures along with rainfall percentages during the warmest seasons (%Pcm1 and %Psu). The area spreads from coastal AK to approximately mainland Dixon Entrance. In the first group. In Oregon and Washington. i. The true boreal zone is recognizable in the climate diagrams as the point where the duration of the period with a daily average temperature of more than 10ºC drops below 120 days and the cold season lasts longer than six months [7]. menziesii and Q. an OTI2w>2 indicates a PNV with clear floristic temperate affinities. with the absence of A. the OTI3 is higher and a threshold of 1. garryana. Thuja plicata or Tsuga heterophylla. whereas temperate stations in which OTI3>1. The remaining 40 northern stations can be divided into two groups. In this second group. Oregonian. Tsuga heterophylla are absent. Hudsonian.8 are here considered as Mediterranean.. The first group includes 24 stations in cold 6 . In the second group A. When these latter stations are compared with the rest of the temperate stations. and temperate trees such as Picea sitchensis. A second question concerns the discrimination between temperate and boreal stations at the borders between Alaska and British Columbia. or Microthermal after Köppen’s classical system. This climate arises in the lee of the BC Coastal Ranges north of 58º. 26]. spreading from Dixon Entrance southward. In British Columbia. three phytogeographical provinces coexist [13]: 1. Sitka spruce forests constitute the PNV of this rainy and cold boreal coast. which corresponds to Walter’s ZE IV-V. the PNV can be divided into two groups. Picea sitchensis. and we included those with an OTI3>1.8 but OTI2w<2 were included in the Temperate Dry or Submediterranean bioclimate. in the lee of the Coastal Ranges [25. Sitkan. and was referred to as “boreo-temperate” in our analysis of the chionophilous vegetation of western North America [27]. heterophylla and T. These stations were ascribed to the Temperate Humid bioclimate. 3. From British Columbia to Alaska.e. mariana. There is no sharp boundary but a transitional ZE VI-VIII intercalated between the two [7]. At 39 of these stations. garryana. located in inland Alaska zones colonized by taiga forests dominated or co-dominated by Picea glauca and/or P. where the border with the temperate Oregonian province is found [11]. just north of the Queen Charlotte Islands.

the PNV comprises Quercus agrifolia forests and woodlands: AGRI-1. garryana forests and woodlands in wetter areas. the PNV comprises Q. or Q. a Tropical Arid zone lying in an area where Larrea tridentata dominates or co-dominates in every zonal community. In only one station showing the highest continentality index (COI>18). and AGRI-2. soils. in the driest parts of the Central Valley. the PNV is 7 . but because both show lower summer rainfall values and their PNV is dominated by Abies amabilis. as in the Southern Californian and Martirense provinces. When precipitation increases and TEV decreases (OEI>7). heterophylla and P. mainly in the coldest (THI<230) northern areas of Oregon. In semiarid areas. and DOUG-2. in such areas. wiliszenii (WILIS). When OTI<2. in ultramafic soils. summer rainfall is three times lower and temperatures cooler. 30]. menziesii. The second group comprises the remaining southern stations distributed in the southernmost part of the Alaska Pandhale and along the British Columbia Coastal Ranges. plicata. Towards the coast. At both stations. where temperatures rise (THI>230). These forests. which can be reached by the powerful root system of the dominant tree [29. These stations were assigned to the Boreal Oceanic bioclimate. Four of these located in southeastern California appear in the Mohave Desert. douglasii savannahs (DOUG-1-SAV) seem to require deep phreatic layers. In areas under maritime influence (COI<13). which never occur in the first group. the most northwestern extension of the Sonoran Desert. douglasii forests and woodlands: DOUG-1 thriving in normal. whose PNV is dominated by typical temperate species such as Picea engelmannii..e. Within the Mediterranean MB there are four bioclimates. they are local seral brushlands on drier slopes with shallow soils. Inland. T. inhabiting colder areas (Tp<190) of the Southern Californian mountains and lowlands from Point Conception northwards. Washington and British Columbia. The threshold 2 for OTI seems to be the best limit for discriminating Semiarid from Dry and Humid bioclimates. Six stations sharing the highest values of aridity (ARI>5 and OEI<2) were included within the Mediterranean Arid bioclimate. This second group of stations was included in the Temperate Humid bioclimate.284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 regions (Tp<76) with the highest or among the highest summer rainfall spread across the northern Alaska Pandhale. woodlands and savannahs were here included within the vegetation type GARRY. i. T. In dry or humid areas (OTI>2). In continental areas (COI>13). Also included in this group were two subalpine stations located above 850 msl in the British Columbia Coastal Ranges. is the climatic climax in zones with Mediterranean Semiarid bioclimate and constitute degradation stages of sclerophyllous or mixed evergreen forests in areas of higher rainfall of both the Southern Californian and Martirense provinces [28]. Chaparral. they can be easily included within the Temperate MB and their PNV is floristically related to other temperate forests. In interior areas of the Northern Californian province. There are two types of PNV in Mediterranean Semiarid areas: chaparrals and savannahs of Quercus douglasii. on warm lowlands (Tp>190) of the Southern Californian province. and their PNV (ATRI) is dominated by shrubs of the genus Atriplex. in the rainshadow of the Southern Coastal Ranges. while Q. chaparrals are the true PNV. chaparral is seral to the forest vegetation. not ultramafic. the PNV chiefly corresponds to Q. not the natural climax. Otherwise. Under the Mediterranean Dry bioclimate the PNV consists of oak forests and woodlands distributed according to continentality. the PNV is dominated by Q. the evergreen sclerophyllous shrubland that dominates the cismontane side of coastal mountains ranges from about San Francisco south to Ensenada in Baja California. garryana savannahs in the continental and drier valleys of central Oregon and Washington. The remaining two Mediterranean Arid stations are located further north. a feature of the climatophilous vegetation of the cold deserts of the Great Basin region [5]. chaparrals (CHAPA-4) mostly pervade rocky places with shallow soils. Quercus agrifolia woodlands thrive only in gullies and canyons with access to additional soil moisture. chaparrals may represent what remains after human activity has destroyed the trees. the Mediterranean Humid bioclimate occurs.

and SITKA-2. supported by many Californian elements. forests of Sitka spruce form a long narrow band adjacent to the Pacific ocean. redwoods (SEQU-1) begin to appear. and SEQU-2. 8 . agrifolia do not flourish. The range of Q. closely related to the Oregonian floristic element. which receive less summer fog than the slopes sustaining redwood forests. The most conspicuous feature of the temperate PNV of the Pacific Northwest with regard to other temperate forests of the world is the confining of deciduous trees to younger forests. and Quercus chrysolepis forests constitute the PNV in normal (CHRYS-2) and ultramafic soils (CHRYS-3). the temperature is cool and there is high precipitation and frequent fogs. Only a mountain station in the Southern Coastal Ranges supports mixed forests dominated by the endemic Pseudotsuga macrocarpa (CHRYS-1). mainly by LITHO. Inland. There. Floristic and bioclimatic data suggest separating the Sitka spruce forests into two associations: SITKA-1. Most of the floristic assemblages associated with redwoods can be divided into two groups. agrifolia forests (AGRI-3). and oakwoods. linked to the Californian floristic element [13]. those of AGRI-3 sustain a large group of differential taxa most of which also occur in northern forests such as LITHO and SEQU-1. of the Mediterranean Humid bioclimate. Since the northernmost areas occupied by the SITKA-2 association are largely inaccessible. the latter grouping pervading forests from northern California to approximately Glacier Bay National Park in southeastern Alaska. which appears on the Klamath and Siskiyou mountains on the northern California and southern Oregon coasts. coastal trees such as L. From 35°48' (at Salmon Creek. Unlike the southern oakwoods. such that between this latitude and 41° there exists a topographical mosaic. the transitional ecotone with the boreal SITKA-1 should lie somewhere between Glacier Bay and the Malaspina Glacier. while the boreal SITKA-1 has been called a “subpolar rain forest” [33]. from west to east and from low to high altitude. a Lithocarpus densiflorus-Arbutus menziesii forest. from south to north. At the southern limits of their distribution area. redwood groves commence as isolated patches linked to riverbanks and shady canyons. of the Temperate MB. agrifolia ends north of San Francisco Bay. This area is roughly coincidental with that of the redwoods (SEQU1). temperate. and Griffin [29] in northern California. As noted above. sitchensis on the westernmost Kenai Peninsula of Alaska. In fact. the PNV is dense Q.336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 mixed evergreen forests (LITHO). the northern group. This floristically intermediate composition between Mediterranean and temperate redwoods is consistent with Walter’s map [7:12] in which the coastal border between California and Oregon is drawn as the ZE IV-V. In the warmest areas across the windward slopes of the Coastal Ranges in central and northern California. in contrast. differentiated by some Oregonian elements. Although the PNV of the temperate areas is unified by climate and physiography and has an evergreen coniferous component. with redwoods settling on northern or ocean exposed slopes where the fog effect is ecologically important. where the SITKA-1 area begins and spreads northward to the sea-level timberline of P. the northern limit of its range cannot be determined exactly. These groups support two floristic and climatic redwoods: SEQU-1. but principally from southern Oregon to the Gulf of Alaska. and the dominance of giant conifers that escaped decimation during Pleistocene glaciation [32]. and the southern group. Walter [7] described this area as ZE V-VIII. refer to a Lithocarpus densiflorus-Arbutus menziesii forest that substitutes redwood forests according to a decreasing-humidity gradient. it varies from the coast inland. From northern California. the classic broad sclerophyll forest described by Whittaker [31]. LITHO is essentially a redwood border forest occurring mainly on sunny or leeward slopes. California). thriving on sunny or leeward slopes. both Whittaker [31] in the Siskiyou and Klamath. riverbanks and frequently disturbed areas. boreal. Since the SITKA-2 distribution range occupies the area showing the greatest precipitation in western North America. the designations “perhumid rain forest” or “temperate rain forest” are generally used for these forests. densiflorus and Q. where maritime influences are maximal. and oak forests are replaced there by other mixed evergreen forests.

dominates low and medium elevations from northern Vancouver Island to Glacier Bay in northern southeast Alaska. J.. Rivas-Martínez. sitchensis dominate down to sea level. the so-called “seasonal rainforest” [33].387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 Away from direct oceanic influence. in the lee of the Kenai coastal mountains. [5] S. REFERENCES [1] L. “North American Terrestrial Vegetation. or “perhumid rainforest”. in which Abies amabilis is common and sometimes co-dominant with T. Tsuga mertensiana and P. 1. North American boreal and western temperate forest vegetation. F. mariana muskegs. 1993. I. heterophylla and T. In permafrost-free. 2000. D. L. [2] M. Aguirre and J. Sánchez-Mata and M. D.” In: Flora of North America Editorial Committee. 5-148. although the timberline can be as low as 200 m above sea level. PICEA-2 alternates with BETUL. appears the Boreal Oceanic bioclimate. Vol. G. 2007. Along the rainier and foggier coasts of Alaska. 191. [3] M. ENGEL. [6] S. in the Cascades. zonoecotones and azonal vegetation along the Pacific coast of North America. PICEA-2 forests occupy zones where the ground is seasonally frozen. Ocaña-Peinado. Acknowledgments This research was supported by grants from the Franklin Institute of North American Studies (University of Alcalá). which is replaced inland eastwards. “Zonobiomes. windward of the Alaska-Chugach Ranges from Kenai south to approximately Yakutat Bay. 2. where the PNV mainly corresponds to a mixed deciduous-coniferous forest (PICEA-1). 5. Delgadillo. No. Vol. and only reaches the Pacific coast at Cook Inlet. Peinado. Costa. 10. Billings. New York. In contrast. A. plicata.” Itinera Geobotanica. Flora of North America: North of Mexico. M. Barbour and W. J. in the lee of the British Columbia Coastal Ranges) replace PSEU. which thrives in an area that is permafrost-free [10]. pp. pp. A. “Syntaxonomical synopsis of the North America natural potential vegetation communities. which correspond to the “temperate rainforests” so-called by American ecologists. 5-316. Aguirre. when OTI2w>2. subalpine forests (AMAB. along with many other typical species of temperate forests were lacking in relevés taken in the SITKA-1 boreal forests. Brouillet and R. At high elevations. very dry soils (lithic regosols). menziesii. The PNV corresponds to Sitka spruce forests (SITKA-1). Vol. The Boreal Continental bioclimate arises in the lee of the British Columbia coastal mountains north of 58º.” Plant Ecology. 1999. “Climate and Physiography. Rivas-Martínez. This bioclimate spans most interior areas of British Columbia and Alaska. Vol. which grow scattered among prostrate shrubs and over a dense carpet of fruticose lichens and xeric mosses. a woodland almost exclusively comprised of multistemmed and rather stunted Betula neoalaskana. by the Submediterranean forests harboring Arbutus menziesii and Thuja plicata (ARTHU).” Plant Ecology. 221-252. 263-281. From Oregon northward to southern British Columbia a less moist variant of PSEU predominates. 212. Within the Boreal MB there are two bioclimates. Vol. 15-46. [4] M. The most humid variant. J. “Bioclimates and vegetation along the Pacific basin of Northwestern Mexico. 1997. 9 . pp. mainly on rocky outcrops of the highest elevations. M.” 2nd ed. Whetstone. Its presence is reflected by the appearance of Picea glauca forests and P. pp. Oxford University Press. most of the temperate zone of the lowlands toward the interior of the coastal strip occupied by SITKA-2 is covered by mixed evergreen forests (PSEU). pp. D. mainly in cryosols developed on north-facing slopes and higher altitudes. Peinado. Cambridge. Macías. The authors thank Ana Burton. Cambridge University Press. plicata and P. 12. Delgadillo and M. who polished our English. L.” Itinera Geobotanica. Eds. Macías. T.

Avena and C. R. “Worldwide bioclimatic classification system”. [19] H. pp. NRCS (2011) The PLANTS Database. J. 29-47. 75. 2009. 2002. Delgadillo. [11] L. Macías. Rivas-Sáenz. R. [16] NCDIA. Delgadillo and M. “R. M. Fisher.” Journal of Climate and Applied Meteorology. R.” Anales Jardín Botánico de Madrid. Grattan-Guinness. “Natural vegetation of Oregon and Washington. F. pp. pp.” 2nd ed. Ann Arbor. Springer-Verlag.” Plant Ecology. Accessed June 2012. Aguirre and J. S. 2002. Penas Merino. “Vegetation of the Earth and Ecological Systems of the Geobiosphere. Williams. W. “An approach to adjusting climatological time series for discontinuous inhomogeneities. 26. 2002.B. Dyrness. Washington. [8] S. G. Berlin.” U.” 2003- http://climate.. [13] M. A. http://plants. Mueller-Dombois. pp. [15] NOAA. Peinado. 27-60. [20] USDA. L. 1-638. R. [14] T. S. 1943.” Northwest Science. [22] Z. 2. Technical Report 040.” 3rd ed. 1985. A.” Prentice Hall. Statistical Methods for Research Workers. R. “The biotic provinces of North America. L.gc. Vol. de la Cruz M. L. 2000. Mexico. J. Vol. Thysell and A. “Quercus garryana communities in the Puget Sound. 1974. [10] M. Rivas-Martínez.usda. pp.html [17] L. Michalewicz and D. National Oceanic and Atmospheric Administration. Oregon State University Press. Fogel. Ed. “How To Solve It: Modern Heuristics. Aguirre. “U. pp. Amsterdam. Prentice Hall. Department of Commerce. Erickson and D. British Columbia Ministry of Forests. 129. Monthly Climate Normals 1971-2000. Landmark Writings in Western Mathematics: Case Studies. Blasi C. 2001. 66. 1. Tatham and W. Karl and C. 2nd ed. R. Ricotta. Meidinger. 2005. Vol. Dice. Carey. “A phytosociological survey of the boreal forest (Vaccinio-Piceetea) in North America. Franklin and C. “A phytosociological and phytogeographical survey of the coastal vegetation of western North America. [9] C. Edwards. S. Corvallis. Dingman.435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 [7] H. Carranza. 1988.” Upper Saddle River. J.” Instituto Interamericano de Cooperación para la Agricultura. [21] A. Berlin. R. 5. “Physical Hydrology. (2007) Garry oak (Quercus garryana) plant communities in British Columbia: a guide to identification. 1744-1763.” University of Michigan Press. “Are potential vegetation maps a meaningful alternative to neutral landscape models?” Applied Vegetation Science. Holdridge.” In: I. “Multivariate Data Analysis with readings. B. F. [26] D. 1987. Peinado. Vol. [18] S. “National Climate Data and Information Archive. No. National Climatic Center. A. 1982. 1998.” John Wiley and Sons.” Plant Ecology. [12] M.” Springer Verlag. San José de Costa Rica. L. 2011. Black. Macías. 2008. Victoria. V. 10 . Part I: plant communities of Baja California. Walter. and A. M. L. J. Vol. Peinado. Elsevier. “Aims and methods of vegetation ecology. Ellenberg and D.weatheroffice. 151194. Asheville. 271–275. Vol. “Fitogeografía de la costa del Pacífico de Norteamérica. 196. 219-235. pp 1640-1940. Global Geobotany. New York. Vol. Aguirre and M. “Ecología basada en zonas de vida. Hair. R. [24] W. Englewood Cliffs. [23] J. [25] J.

F. “Vegetation formations and associations of the zonobiomes along the North American Pacific coast. California Native Plant Society. Journal of Vegetation Science. Davis.” Vegetatio. Departamento de Biología Vegetal. pp. 123-135. [29] D. “Oak forest. pp. 79-94. [28] M. 279-338. 1994.” Ph. Aguirre and J. pp. J.479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 [27] M. pp. 1995. [33] J. Álvarez.” Ecological Monographs. “Vegetation of the Siskiyou Mountains. [31] R. Universidad de Alcalá. Griffin. I. [32] R. 1960. Delgadillo. 1380-1386. MacKinnon. 1994. F. D. F. 204. [30] J. 114. Peinado. 383-416. “Major plant communities of warm North American deserts”. L. 1988. Oregon and California. L. “Evergreen coniferous forests of the Pacific Northwest. Vol. H. 2nd ed. H. Pojar and A. Major. Alcaraz. Franklin. 1995. pp. Terrestrial Vegetation of California. Thesis. 11 . Waring and J. Vol. 6. fitogeográfico y fitosociológico del suroeste de Norteamérica Estados Unidos y México.” Science Vol. Alcalá de Henares.” Lone Pine Publishers. Whittaker. Alcaraz. “Introducción al conocimiento bioclimático. Peinado. J.” In: M. Barbour and J. Aguirre and J. 1979. 30. Delgadillo.. “Plants of coastal British Columbia. Vancouver. G. Eds.

months in which P ≥ 2T Average maximum temperature of the coldest month Average minimum temperature of the coldest month Ombro-Evapotranspiration Index (OEI = 10(Pp/TEV)) Ombrothermal index (OTI = Pp/Tp) Ombrothermal index of the two driest consecutive months of the year Ombrothermal index of the three driest consecutive months of the year Ombrothermal index of the three driest consecutive months of the year more the previous month Ombrothermal index of the two warmest consecutive months of the year Yearly average precipitation Autumn precipitation (September + October + November) Percentage autumn precipitation Precipitation of the warmest four-month season in the year Percentage precipitation of the warmest four-month season Precipitation of the four-month season before Pcm1 Percentage precipitation of the four-month season before Pcm1 Precipitation of the four-month season after Pcm1 Percentage precipitation of the four-month season after Pcm1 Total precipitation for the humid months Percentage precipitation of the humid months Percentage precipitation from June to September Positive precipitation.496 Table 1 Climate data. parameters and indices ARI BIOT COI HBIO HEV Hm M m OEI OTI OTI2 OTI3 OTI4 OTI2w P Pau %Pau Pcm1 %Pcm1 Pcm2 %Pcm2 Pcm3 %Pcm3 Phm %Phm: %Pj-s Pp: Ps2 Psp %Psp Psu %Psu Pwi %Pwi SEPI T TEV THI Tmax Tmin Tn Aridity index (ARI = TEV/P) Holdrige biotemperature Continentality Index (COI = Tmax-Tmin) Holdrige evapotranspiration index (HEVI = HEV/P) Holdrige yearly evapotranspiration Humid months. sum of the mean monthly temperatures of those months whose mean 12 . total precipitation of those months whose mean temperature is higher than 0ºC Precipitation of the two warmest consecutive months of the year Spring precipitation (March + April + May) Percentage spring precipitation Summer precipitation (June + July + August) Percentage summer precipitation Winter precipitation (December + January + February) Percentage winter precipitation Seasonal precipitation index (SEPI = (%Pwi + %Psp) / (%Psu + %Pau)) Mean yearly temperature Thornthwaite yearly evapotranspiration Thermicity Index (THI = 10(T + m + M)) Mean temperature of the warmest month Mean temperature of the coldest month Negative temperature.

period or number of months of the year whose monthly temperature > 35 ºC 13 .Tp Ts2 Tsu VGP temperature is lower than 0ºC Positive temperature. sum of the mean monthly temperatures of those months whose mean temperature is higher than 0ºC Mean temperature of the two warmest consecutive months of the year Summer mean temperature (June + July + August) Vegetative growth.

62 16.38 22.47 180.04 -0.05 31.93 -0.07 1.00 135.01 level.20 10.97 0.60 9.12 705.60 -0.65 0.94 0.80 34.30 5.38 16.40 40.59 7.50 0.94 0.23 0.90 41.45 25.78 8.15 4.04 1.05 511.49 3.45 20.80 74.48 458.88 -0.88 0.60 OTI3 2.65 8.69 -0.67 SEPI 2.01 191.11 3.35 41.60 3.50 1.45 0.80 2.78 0.06 10.79 41.54 12.55 OTI4 3.71 0.90 42.65 3.53 13.00 0.30 33.68 0.77 -0.60 11.86 0.27 0.96 -0.25 17.32 21. parameters and indices: summary of the main statisticals values.30 7.85 56.30 1705.20 65.43 1386.41 11.60 497 498 14 .03 12.68 139.88 1.00 10.29 0.54 -0.20 11.68 9.13 15.66 51.50 9.74 0.33 4.85 22.98 8.30 7.05 42.15 1215.80 3.10 14.87 868.15 13.79 Mean 95% Confidence Interval for Mean: Lower Bound 95% Confidence Interval for Mean: Upper Bound Median Lower Quartile Upper Quartile Maximum Minimum Variance Standard deviation Correlation coefficient (Latitude) Correlation coefficient (Longitude) Tp TEV HEV 134.02 0.25 620.70 1378.50 55.75 1.10 0.15 HEVI 1.66 7.26 0.29 111.57 Tn 3.75 31.90 10.45 107.70 10.59 0.14 7.00 135.91 206.00 12.40 14.57 3.91 1.46 3.83 P 1301.28 10.65 689.99 682.34 0. Mean 95% Confidence Interval for Mean: Lower Bound 95% Confidence Interval for Mean: Upper Bound Median Lower Quartile Upper Quartile Maximum Minimum Variance Standard deviation Correlation coefficient (Latitude) Correlation coefficient (Longitude) T 10.84 0.13 33.82 11.96 14.77 7.77 2.34 139.27 16.15 1.53 -0.83 M 8.79 20.30 658.56 0.00 -0.35 %Psu 9.59 OTI2 2.72 1.82 50.00 21.79 1139.93 21.24 1.40 68.33 10.63 11.00 5.21 1300.50 Pp 1218.88 0.24 9.00 60.54 0.85 1.03 18.29 2.05 8.70 -0.83 %Pau %Pcm1 %Pcm2 %Pcm3 26.80 3.75 OEI ARI 19.39 0.22 0.00 0.50 11.50 20.39 134.00 0.13 0.00 94.60 35.00 14.60 863662.74 167.88 315.01 8.53 -0.60 Ts2 %Pwi %Psp 36.58 0.82 24.56 3.11 175.57 14.33 0.70 -24.75 22.80 22.00 83.70 270.00 3.63 1.98 50.02 5728.86 1104.76 43.20 0.65 -437.04 3.96 -0.11 142.01 COI 14.90 59.33 35.20 4.46 190.46 8.00 119.38 3.26 -0.12 7.20 40.09 261.95 -0.08 21.23 16.77 7.17 -0.20 14.89 -0.64 0.32 2.53 0.82 -0.96 -0.13 35.96 -0.30 16.59 6.44 83.51 0.20 0.60 4.56 10.80 1.74 32561.68 199.80 929.12 -0.45 1329.52 0.58 3.41 524.91 27905.70 58.05 0.39 494.83 -0.96 -0.08 1.20 69.81 -0.78 -0.91 0.70 14.48 11.07 15.30 48.74 -0.17 4.38 7.55 HBIO 11. All correlations are significant at the 0.98 0.35 22.35 47.59 27.04 0.93 2.38 35.53 8.77 4.44 238.30 21.95 176.71 0.36 1.45 1758.78 5.00 2.37 0.47 -0.60 4.79 m 0.40 2349.05 3.62 49.67 0. For abbreviations of climate variables see Table 1.36 4.14 2.00 12.85 13.55 3.60 796061.35 2.20 1.34 660.03 20.76 -0.93 9.96 15.74 2.69 52.97 36.60 298.29 26.30 3.94 25.77 OTI THI OTI2w 11.00 22.09 -0.57 0.62 1.26 48.50 19.71 0.77 Tsu 17.40 593.23 34.90 50.80 7.6 -3.70 638.90 26.55 9.97 2.24 14.47 2.02 12.10 0.Table 2 Climate data.35 26.68 -0.76 -0.28 124.15 4.90 824.71 2.00 30.20 29.10 21.00 12.41 Hm 8.85 -0.10 5728.93 17.74 14.75 11.19 10.1 892.00 0.48 0.14 57010.20 3.82 -0.33 62.37 130.18 672.00 51.80 23.80 30.80 35.70 -15.73 10.30 1.41 0.53 0.16 -0.50 1136.69 -0.87 -0.

02102x) y = 101.7966/x) y = 1/(0.1108x -0.4303 – 374.0307x) y = -3560.1318x+0.622/x) y = 34664.3336 82.1247 68.1557x) 2 2 2 4 2 3 R² 92.9232 57.0161 70.70002x -0.52807 .0257 3 y = 1845.255x+7.25x+6.7752x –0.524 92.2398x+0.5434x y = 34.0331 2 3 y = 2421.5317x -0.0306x2 y = 1426.49172x -0.1709x –0.468 83.35-39. All the regression functions are significant at the 0.5547/x) y = 1/(7.0012x y = 1/(0.8323 91.9974 + 0.7919x –0.86238 – 904.0078x +0.415-8.5739 89.22-184.125x+0.1123x+0.71x+6.0968 70.0670 – 7.9074 15 .8236 75.56-49.1685 84.8-159.1369 85.9104 .0994x +0.1645 67.05 level.4858x2 -0.0257 76.03777x +0. For abbreviations of the variables see Table 1 Variables M T m Tsu Ts2 Latitude y = 45.00006x y = 4267.9853x y = (22.2376x 2 92.2089 76.00006x y = 21852.4978 – 0.278-0.0078 – 0.8848x+0.4898/x) 2 3 78.4277 4 4 4 y = exp (4.08241x) y = 5198.0524x -0.46-306.499 Table 3 Regression fit equations and adjusted R².422 – 1.6185 67.0061x 2 2 2 3 3 R² 76.4620x –0.2532 79.6292 67.8359 68.3258x –0.2939x y = 109.0004x3 y = 172.0011x y = 1591.7-708.0136x-0.5307x y = 25.8887x + 0.57-2.32214/x) y = 1/(0.0476x3 y = exp (8.13-151.92-28.4587x+0.0008x y = (-3.7/x 2 2 %Pwi %Psp %Psu %Pau %Pcm1 %Pcm2 Tp TEV HEV HEVI HBIO SEPI THI OTI2w y = -53.6928 .0184x y = 75.0485 95.0727x +0.3810x -0.0006x y = 3735.647 – 13520.6036x-0.88.6208 70.1538 – 16.03785x +0.56x+5.1068 86.8547 86.0125 62.03501x +0.72-95.83/x) y = 382.00001x y = 1/(0.578x+7.0003x y = 42.843/x) y = exp (8.6308x + 0.8179x y = 35.6x+4.405 .0095x 2 2 2 3 3 3 2 3 2 3 4 65.2574-0.0583+2.8128– 387.0.88+74.486/x) y = 1959.4-674.0.68-38.9009– 399.8384 .1987 – 16.0005x 2 3 4 y = 23933.1903x2 -0..9365 79.2638x - y = 1/(0.9105 – 0.0136 – 1.0693 – 0.1664 OTI2 OTI3 Tn 78.63-76.1749 91.0003x 93.0016x 2 3 57.28x+1.6748x –0.0041x 2 y = 5664.5-706.00006x y = 23986.0769 82.0458 70.562/x) y = exp (8.8567 2 3 Longitude y = 639.6746 72.5881 67.250715x) y = exp (7.8659-2.0007x y = 1/(0.3732x y = exp (4.2055 y = 1715.9437 y = 52.9547x+0.0173x y = 117.1187 92.4544 – 0.708x+7.9796x+0.5-648.678x+0.976-6.0405x –0.9967 52.6353 90.6552 93.8902x+0.2848 85.3933 81.

Zonobiome IV. Zonobiome VII. Xerohalophilous shrub growing on slightly saline soils. Boreal spruce forest on lowlands. Zonobiome VII. Zonobiome IV. Northern Californian province. Sonoran microphyllous desert. Chaparral. Mediterranean Arid. Zonobiome IV. Boreal Continental. Zonobiome VIII. Redwood forests. Mediterranean Humid. Temperate Humid. Quercus douglasii forests. Mediterranean Dry. Southern Californian province. 16 . Mediterranean Humid. Chaparral. Southern Californian province (Venturan sector). Mediterranean Semiarid. Temperate Humid. Mediterranean Humid. Mediterranean Humid. Mixed Hardwood Forest on ultramafic soils. Northern Californian province. Mediterranean (and Tropical) Arid. Temperate coniferous forests. Sitka spruce forests. Oregon oak forests and woodlands. Quercus agrifolia woodlands (Southern Californian and Martirense provinces). Mixed Hardwood forests. Martirense province. Mediterranean Semiarid. Mixed Hardwood Forest. Mediterranean Dry (forests and woodlands) and Semiarid (savannahs). Zonobiome VIII. Zonobiome IV. Zonobiome IV. Mediterranean Dry. Zonobiome IV. Zonobiome VIII. woodlands and savannahs. Temperate Humid. Arbutus menziesii-Thuja plicata forests. Boreal Continental. Redwood forests. Boreal Oceanic. Southern Californian province (Diegan and Riversidian sectors). Zonobiome VII. Zonobiome IV. Chaparral. Mohavean microphyllous desert. Mixed Hardwood Forest. Zonobiome IV. Mediterranean Humid. Mediterranean Semiarid. Quercus agrifolia woodlands (Southern Californian mountains and drier areas of the Northern Californian province). Zonobiome IV. Chaparral. Zonobiome IV. Temperate Dry. Zonobiome VII. AGR-1 AGR-2 AGR-3 AMAB ARTHU ATRI BETUL CHAP-1 CHAP-2 CHAP-3 CHAP-4 CHRY-1 CHRY-2 CHRY-3 DOUG-1 DOUG-2 ENGEL GARRY LITHO MID-1 MID-2 PICEA-1 PICEA-2 PSEU SEQU-1 SEQU-2 SITKA-1 SITKA-2 WILISZ Mediterranean Dry (on Semiarid areas may thrives in gullies and canyons with additional soil moisture). Zonobiome VII. Temperate Humid. Quercus douglasii forests and woodlands on ultramafic soils. Mediterranean Arid. Zonobiome IV. Temperate Cold. Zonobiome IV. Sitka spruce forests. Zonobiomes III. Mixed evergreen forest. Dwarf birch scrub on highlands. Subalpine forest. Zonobiome VII. IV. Zonobiome IV. Zonobiome VIII. Zonobiome IV. Subalpine (oceanic) forest. Mediterranean Humid. Zonobiome IV. Zonobiome IV. Quercus agrifolia forests (Northern Californian province). Mediterranean Humid. Zonobiome IV. Boreal spruce forest on highlands. Zonobiome IV.Table 4 Summarized descriptions (bioclimate. zonobiome or zonoecotone and brief diagnosis) of the vegetation types arranged alphabetically according to the abbreviations used in the text. Mediterranean Dry. Boreal Continental. Mediterranean Semiarid. Northern Californian province.

) THI > 230: THI ≥ 280: THI < 280: ARI > 0.) Tn < 5: OTI2w < 2: OTI2w > 2: THI > 300.5. OTI > 10: LITHO (1 st. Ts2 < 22. %Pcm1 < 7.) %Psu > 2: SEPI < 4: LARREA-1 (1 st.) OEI < 7: OEI > 7: OTI3 > 0. For abbreviations of climate data see Table 1.) COI < 10: COI > 10: OEI > 25. OTI > 1: %Psu < 2: ATRI (2 st. %Pwi < 34: Tp < 50: BETUL (1 st. ARI > 0.) %Pcm1 > 20.) OEI < 25.) THI < 270: ARI < 0. %Pwi > 34: AMAB (2 st.5: BETUL (2 st.80: OEI > 12.) OTI2w < 8: PSEU (73 st. Tsu > 10. OTI < 10: OTI3 > 0.) COI > 13 OTI2w > 8: SITKA-2 (7 st.) ARI > 0. %Pcm1 > 7.7: AGRI-3 (31 st.) Without the two previous premises: COI > 13: Continental variant: COI < 13: Oceanic variant: COI > 18. Tsu < 10. ARI < 5.) SEPI > 4: LARREA-2 (3 st.) Tp < 190: AGRI-2 (10 st.) COI < 11: P < 280. ARI < 0. with Quercus agrifolia when COI < 14) and DOUG-2 (2 st. %Psu < 5: SEQU-2 (1 st. OTI < 1: OEI > 2.) COI < 13: SITKA-2 (49 st.5: PICEA-2 (4 st.) Tp < 50. and for PNV see Table 4 1 1b 2 2b 3 3b 4 4b 5 5b 6 6b 7 7b 8 8b 9 9b 10 10b 11 11b 12 12b 13 13b 14 14b 15 15b 16 16b 17 17b 18 18b 19 19b 20 20b 21 21b 22 22b 23 23b 24 24b 25 25b 26 26b 27 27b 28 28b 29 29b 30 30b 31 31b 32 32b 33 33b 34 34b 35 35b 36 OTI3 > 1. SEPI < 5: DOUG-1 (15 st.) Tp < 76: Tp > 76: %Pcm1 < 20.) Tn > 5: ENGEL (7 st.) THI < 300.) OTI < 2: OTI > 2: THI < 380: DOUG-1-SAV/CHAPA-4 (17 st.8: LITHO (4 st. TEV > 380: PICEA-1 (31 st. OTI > 10: LITHO (5 st.) THI > 380: COI > 11: CHAPA-3 (13 st.) COI < 12: SITKA-2 (7 st. ARI < 0.45: COI > 12: PSEU (10 st.8: OTI3 < 1.) 2 15 3 5 4 6 8 7 9 10 13 11 12 14 16 18 17 19 22 20 21 23 27 24 25 26 28 29 31 30 32 33 34 35 36 17 . Pn-a < 250: CHAPA-1 (3 st. Pn-a n 250: CHAPA-2 (20 st.) OEI < 2.) P > 280.4: LITHO (3 st.) OTI3 < 0.5.8: Tn > 22: Tn < 22: Tp > 50.20 and elevation > 500 msl: AGRI-1 (3 st..) COI < 18.7 OEI > 22. %Psu > 5: ARI > 0. OTI < 10: SEQU-1 (6 st.80: GARRY (5 st.. TEV < 380: COI < 20.) OEI < 22.). ARI > 5. ARI ≤ 0. with DOUGL-2 on ultramafic soils) ARI < 0.4: SEQU-1 (3 st. with Quercus douglasii when COI > 8) THI < 230: GARRY (45 st.) COI > 20.Table 5 Key to PNV types based on climatic differences. Ts2 > 22.45: ARTHU (28 st.) Tp > 50: SITKA-1 (23 st. on ultramafic soils) Tp > 190: AGRI-1 (15 st. SEPI > 5: WILIS (1 st.) THI > 270: CHRYS-1 (1 st.

36b OEI < 12.8: CHRYS-2/CHRYS-3 (3 st.) 18 . ARI > 0.

along Lithocarpus densiflorus dominates the forest canopy: SEQU-1. 8 Subalpine forests. Q. U. Boreal and temperate coniferous forests: 2b Sclerophyllous and/or deciduous trees (mainly Quercus species) dominate tree canopy. understories with Cornus canadensis. menziesii as dominant in the forest canopy. Mountains of southern California province: CHRY1. T. Continental Q. densiflorus always lacks. never dominate tree canopy. plicata and P. menziesii as co-dominant tree. Mediterranean forests and woodlands: 3 With Picea glauca and/or P. plicata and P. mariana as dominant trees. along Tsuga heterophylla. never dominates the forest canopy: 7 P. mariana lack: 4 Betula nana. mertensiana but with T. P. Mertensia paniculata: PICEA-1. Canadian and Alaskan Coastal Ranges): SITKA-2b. Coastal Sitka spruce forests thriving on the foggiest areas of the Pacific coast: 6b Picea sitchensis usually lacks. Thuja plicata. Arbutus menziesii and P. but. For abbreviations see Table 4 1 Forests and woodlands: 1b Scrubs: 2 Needleleaf forests: species of the genera Picea. 2 22 3 11 4 5 6 7 8 9 10 12 13 14 16 15 17 18 21 19 19 . Thuja plicata. menziesii and Polystichum munitum usually lack: 8b Montane forests. if present. Coastal and Cascade Ranges: AMAB. Boreal oceanic Sitka spruce forests: SITKA-1. menziesii. Picea engelmannii or Abies amabilis dominate tree canopy. menziesii usually occur: GARRY. from Point Conception northwards. amabilis. Pacific fir-Sitka spruce forests inhabiting the rainiest areas (Cascades. A. agrifolia forests thriving on central and northern California. Temperate Sitka spruce forests: SITKA-2. Mixed evergreen forests thriving on the driest submediterranean areas of the temperate zone: ARTHU. Q. Arctostaphylos rubra and several dwarf willows dominate in the understory: PICEA-2. menziesii var. 16b Without the above features: 17 Dense mixed evergreen forests. as well as in the mountains of southern California: 15 With Umbellularia californica as co-dominant. 10 With Arbutus menziesii as co-dominant tree. Tsuga heterophylla. L. heterophylla. munitum. californica and P. Abies lasiocarpa usually occurs. 11 Sequoia sempervirens. menziesii mixed forests: PSEU. Usually with Tsuga mertensiana and Chamecyparis nootkatensis as co-dominant trees. Oceanic mixed evergreen forests living on the Northern Californian province: LITHO. Q. Thuja or Tsuga are dominant trees. agrifolia may occur but never dominating the forest canopy: 14 With Malosma laurina and other thermophilous taxa thriving on southern California. 12b Q. With Tsuga mertensiana as codominant living in very snowy areas: SITKA-1a. if occurs. menziesii dominates. 7b Without T. Pseudotsuga. groves and savannas dominated by the deciduous oak Quercus garryana¸Q. 15b U. 9b Abies amabilis co-dominating the forest canopy. 4b Without the above characteristics. Without T. Sclerophyllous species of Quercus usually lack and. Geocaulon lividum. densiflorus. menziesii are dominant tres. menziesii var. Continental oak forests: WILISZ. 5 Sequoia sempervirens. wiliszenii may occurs but never dominating the forest canopy: 13 Quercus agrifolia dominates the forest canopy: 13b Q. T. 10b Without A. often with Lithocarpus densiflorus. kelloggii (mainly in southern areas). Without A. 5b S. laurina. menziesii var. chrysolepis forests: 19 With Pseudotsuga macrocarpa. plicata. menziesii and P. Inland forests of the Canadian Coastal Ranges: ENGEL. dominates the forest canopy: SEQU-2. Quercus douglasii is the dominant tree along with Pinus sabiniana: 18 L. along T. agrifolia forests thriving on the Northern Californian province: AGRI-3. 18b Without the above features. from Point Conception southwards: AGRI-1. sitchensis. sempervirens lacks: 6 Picea sitchensis dominates or co-dominates. californica may occur but never co-dominating. Pseudotsuga menziesii var. munitum is common in the understory: 9 Picea engelmannii dominates or co-dominates. sempervirens as dominant tree: 12 Quercus wiliszenii dominates the forest canopy. Pseudotsuga menziesii var. With Quercus chrysolepis as dominant or codominant: 17b Blue oak woodlands and savannas. 14b Without M. heterophylla: SITKA1b. 16 Forest stands. Coastal Sitka spruce forests living on the driest coastal temperate areas and in the wettest and foggiest Mediterranean areas: SITKA-2a. Polystichum munitum is common in the understory. menziesii. menziesii var. Boreal continental coniferous forests: 3b Picea glauca and P.Table 6 Key to PNV types based on floristic and ecological differences. agrifolia forests thriving on central and southern California: AGRI-2. 11b Without S. With Abies amabilis co-dominating the forest canopy. Thuja plicata and Pseudotsuga menziesii var. menziesii var. P.

Without the above features. Usually on ultramafic soils: DOUG-2 Without Quercus durata: DOUG-1. Yucca brevifolia and other Mohavean taxa: MID-1. With Atriplex confertiflora. Arctostaphylos. Without the above characteristics. Coastal chaparrals of southern California: Without Malosma laurina. Sclerophyllous and drought-deciduous chaparrals. Thermophylous areas of the Southern Californian province from San Diego northwards to Point Conception: CHAPA-2. Colorado creosote bush. tridentata: Mohave creosote bush. With Quercus wiliszenii var. With Quercus durata. On semiarid areas occurs as open savanna (DOUG-1-SAV). Species of the genera Quercus. Q. 20 23 24 25 26 27 28 20 . Rhus or Adenostoma are dominants or co-dominants: With Malosma laurina. Mediterranean shrubs: With Larrea tridentata as dominant shrub: Without L. john-thuckeri and Juniperus californica. Without Opuntia basilaris. On ultramafic soils: CHRY-3. Yucca brevifolia and other Mohavean taxa: MID-2. Ornitostaphylos oppositifolia and other Martirense endemics: CHAPA-1. Mountains of the Southern Californian province: CHAPA-3. A. Without Pinus sabiniana and Quercus durata: CHRY-2.19b 20 20b 21 21b 22 22b 23 23b 24 24b 25 25b 26 26b 27 27b 28 28b Without Pseudotsuga macrocarpa: With Pinus sabiniana and Quercus durata. frutescens. Continental chaparrals of central California: CHAPA-4. With Xylococcus bicolor. With Opuntia basilaris. Sarcobatus vermiculatus and other subhalophilous plants: ATRI. polycarpa. Subcontinental and continental chaparrals: San Diegan chaparrals. With Rhus ovata and Quercus dumosa as main dominant scrubs. Boreal shrubs: BETULA.

8 2a. %Pcm1 < 20 9a. MEDITERRANEAN MACROBIOCLIMATE 5 Mediterranean Arid Bioclimate 3 Mediterranean Semiarid Bioclimate 4 Mediterranean Dry Bioclimate Mediterranean Humid Bioclimate BOREAL MACROBIOCLIMATE (Continental bioclimate) 6 9. OEI < 7 4b. Tp < 76 8a. Tn > 5 9b. OTI2w > 2 10b. Tn > 22 5b. Tn < 5 10a. TEMPERATE MACROBIOCLIMATE 7 TEMPERATE MACROBIOCLIMATE (Cold bioclimate) 8 BOREAL MACROBIOCLIMATE (Oceanic bioclimate) 9.Table 7 Key to bioclimates. OTI3 < 1. OEI < 2 2b.8 1b. Tn > 0 7a. TEMPERATE MACROBIOCLIMATE TEMPERATE MACROBIOCLIMATE (Cold bioclimate) 10. Tn = 0 6b. OEI > 7 5a. OTI2w < 2 2. %Pcm1 > 20 8b. OEI > 2 3a. TEMPERATE MACROBIOCLIMATE (warm bioclimates) Dry bioclimate Humid bioclimate 21 . OTI < 2 3b. Tp > 76 7b. OTI > 2 4a. OTI3 > 1. Tn < 22 6a. For abbreviations of climate variables see Table 1 1a.

For abbreviations of climate variables see Table 1 CLA Stations Classification 1 All stations 2 3 4 5 6 Boreal vs. Tp 3 BIO OTI3. OEI. k-means) for the six variables showing most significance (p < 0. Temperate Inside Boreal (Bioclimates) Inside Temperate (Bioclimates) Inside Mediterranean (Bioclimates) NSA 457 245 396 61 184 212 CLT 3 MB 2 MB 2 MB Variables Tn. bioclimates). Temperate Mediterranean vs. Tn. SCC: Stations correctly classified. %Pcm1 SCC %SCC SIC 407 236 378 61 140 184 89 96 95 100 76 87 Changes between Temperate Dry and Temperate Humid stations in border zones Changes between Mediterranean Dry and Mediterranean Humid stations in border zones 50 Submediterranean stations classified as Mediterranean Some Boreal Oceanic stations classified as Temperate Oceanic and viceversa 16 Submediterranean stations classified as Mediterranean Humid. macrobioclimates. see Table 1. OTI2w. Tp. %Pcm1 OTI. BIO. 2 Mediterranean Humid stations classified as Submediterranean 2 BIO OTI3. %PCm1 OTI3. %Pcm1 4 BIO OEI For abbreviations of variables.0001). OTI.Table 8 Summary of the results obtained in the cluster analysis (CLA. CLT: clusters tested (MB. NSA: number of stations analyzed. OTI2w. SIC: Stations incorrectly classified 500 501 22 .

More detailed maps and geographical coordinates are provided in electronic supplementary material (Maps S1 to S4 and Table S1.502 503 504 505 506 Figure 1 Locations of the weather stations sampled. 23 . respectively).

Note the prevalence of windward rains in the two areas showing lower precipitation. WIND: 60 stations found on the windward slopes of the Coastal Ranges and Cascades. which were in turn grouped into four climate groups. 24 . Stations included in this category show the highest rainfall records (average precipitation: 2088 mm). directly accessed by wet fronts.507 508 509 510 511 512 513 514 515 Figure 2 Distribution of average precipitation for three categories of stations. Rainfall records are significantly lower (average precipitation: 946 mm). LEE: 235 stations influenced by the rainshadow. but in areas with less rain (Mediterranean and Temperate Submediterranean) this type or orographic rainfall prevails. COAST: 162 stations located on coastal plains. approximately at sea level. Rainfall in these stations is lower than in the previous group (average precipitation: 1495 mm). as discussed in the results section.

seasonal precipitation percentages (c) and four indices (d) recorded at 457 stations along a north-south transect in the study area. For abbreviations see Table 1. 25 . For polynomial fit equations and R2 statistics see Tables 2 and 3.: Latitude (degrees).516 517 518 519 520 Figure 3 Regression lines for seven temperature variables (a. b). Lat.

Percentages for 457 stations distributed according to their position with respect to the wet fronts. 2). TEM: 184 stations with a Temperate MB. BOR: 61 boreal stations with a Boreal MB. For abbreviations see Table 1.521 522 523 524 525 526 Figure 4 Seasonal precipitation patterns. The term INLAND included WIND and LEE stations (see Fig. 26 . MED: 212 stations with a Mediterranean MB.

27 . Graphs represent a single value (average precipitation) for all stations in the range of one degree of latitude. 2).527 528 529 Figure 5 Latitudinal distribution of precipitation for the three categories of stations (see Fig.

3. 6b. 2. 8b: mountain stations). 8. 10. Northern Coastal Ranges and Cascades (6a: winward. Aleutian province. (C) Boreal stations: 9. 4. 5. Northern Coastal Ranges. British Columbia and Southern Alaska Coastal Ranges (8a: lowland stations. leeward of the Alaskan Coastal Ranges and stations located inland. Sitkan province (9a: windward of the Alaskan Coastal Ranges. (A) Mediterranean stations: 1. Hudsonian province. 11. (B) Temperate stations: 6. 1b: leeward). 7b: leeward).530 531 532 533 534 535 536 537 Figure 6 Scatter plots for 457 stations represented by their bioclimates. 9b. Puget Trough and Georgia Depression. leeward). Vancouver Island (7a: winward. 3b: leeward). at the head of fjords). 28 . 7. Interior valleys of Oregon and Washington. Central Coastal Ranges (2a: winward. Southern Coastal Ranges (1a: winward.

538 539 Figure 7 Scatter plots for 457 stations represented by PNV (For abbreviations see Table 4). 29 .

Red lines mark the approximate location of the rainshadow areas: 1. slightly modified). Puget Trough and Willamette Valley. 30 . Great Valley of California. Alaskan-Yukon plateaus and lee of the Alaskan ranges. and for PNV see Table 4. and PNV types for 457 sampled stations. Alexander Archipelago-Nahwitti-Hecate-Georgia depressions. 2. Maps S1 to S3 Coastal. 4. Washington. 5. and California (S3). Mohave Desert. after [11]. wind and lee weather stations in Alaska (S1). bioclimatic classification. British Columbia. Map S4 Bioclimates for 457 sampled stations.540 541 542 543 544 545 546 547 548 Supplementary material Figure S1 Physiographic map of the North American Pacific basin and phytogeographical units of the Pacific Northwest (lower left. For abbreviations of climate data see Table 1. Oregon (S2). Table S1 Climatic parameters and indices. 3.