1 2 3 Histological Assessment of Organs in Sexually Mature and Post-spawning Steelhead Trout and Insights into Iteroparity 4 Zachary L. Penney1 and Christine M. Moffitt1,2 * 5 6 7 1 8 penn4282@vandals.uidaho.edu; 208-885-7139 (phone) 9 2 Department of Fish and Wildlife Sciences, University of Idaho, Moscow, Idaho, USA; US Geologic Survey, Idaho Cooperative Fish and Wildlife Research Unit, Department of Fish 10 and Wildlife Sciences, University of Idaho, Moscow, Idaho, USA, cmoffitt@uidaho.edu; 208- 11 885-7047 (phone); 208-885-9080 (fax) 12 * Corresponding author, Christine M. Moffitt 13 For Reviews in Fish Biology and Fisheries 14 Special issue: Teaming up Atlantic and Pacific Salmonid Biologists to Enhance Recovery of 15 Endangered Salmon in North America 16 17 18 19 20 21 This draft manuscript is distributed solely for purposes of scientific peer review. Its content is deliberative and predecisional, so it must not be disclosed or released by reviewers. Because the manuscript has not yet been approved for publication by the U.S. Geological Survey (USGS), it does not represent any official USGS finding or policy." 22 Abstract 23 Steelhead (Oncorhynchus mykiss) are anadromous and iteroparous, but repeat-spawning rates are 24 generally low. Like other anadromous salmonids, steelhead fast during freshwater spawning 25 migrations, but little is known about the changes that occur in vital organs and tissues. 26 Intuitively, it would seem that any fish capable of repeat spawning would not undergo the same 27 irreversible cellular degeneration as occurs in semelparous salmon and begin replacing energy 28 soon after spawning. Using Snake River steelhead as a model we examined the gastrointestinal 29 tract of post-spawn steelhead (kelts) to document if freshwater feeding was occurring, and used 30 histological analysis to assess the cellular architecture in the pyloric stomach, ovary, liver, and 31 spleen in sexually mature and kelt steelhead. We found 38% of the kelts contained food or fecal 32 material in the gastrointestinal tract and that feeding was related to external condition. We found 33 a significant renewal of villi in the pyloric stomachs of kelts was associated with feeding. No 34 vitellogenic oocytes were observed in sections of kelt ovaries, but perinucleolar and early/late 35 stage cortical alveolus stages were present suggesting iteroparity was possible. We documented a 36 negative correlation between the quantity of perinucleolar oocytes in ovarian tissues and fork 37 length suggesting that larger steelhead may invest more into a single spawning event. Liver and 38 spleen tissues of kelts were found to show minimal cellular deterioration. Together, these 39 findings indicate that the physiological processes causing rapid senescence and death in 40 semelparous salmon are not the same in steelhead, and recovery begins in freshwater. 41 42 Keywords: Iteroparity, Histology, Steelhead, Fasting 1 1 2 Introduction Within the salmonidae family two reproductive life history strategies exist: semelparity and 3 iteroparity. Semelparity is exclusive to the Pacific salmon (Oncorhynchus sp) that spawn once 4 and die, although exceptions to this have occasionally been documented (Tsiger 1994; Unwin et 5 al. 1999). All other salmonids are iteroparous, but the degree of inter and intra-specific post- 6 spawning survival is highly variable (Crespi and Teo 2002; Quinn and Meyers 2004). Steelhead 7 O. mykiss, the anadromous form of rainbow trout, and Atlantic salmon Salmo salar are 8 considered iteroparous, but show complex polytypic life history strategies that involve 9 residualism (Viola and Schuck 1995, Christie et al. 2011), anadromy (Schaffer and Elson 1975; 10 McDowall 1987; Pascual et al. 2001), and combinations of life history strategies (Docker and 11 Heath 2003; Thrower et al. 2004; Pearse et al. 2009; Null et al. 2013). Iteroparity can increase 12 genetic diversity of stocks and individual lifetime fitness (Seamon and Quinn 2010), as well as 13 provide a safeguard against complete brood year failures (Narum et al. 2008). Consequently, 14 anadromy can complicate many of the benefits afforded by iteroparity. Foremost, the energetic 15 and physiological costs of migrations between seawater and freshwater can be high and repeat 16 spawning rates among steelhead and Atlantic salmon are generally low (<10%) (Fleming and 17 Reynolds 2004). 18 Within steelhead, a range of life history strategies can exist, including stream-maturing 19 (summer/fall run) and ocean-maturing (winter/spring run) reproductive strategies (Behnke 1992). 20 Stream-maturing steelhead return to freshwater early (June-November) and undergo the final 21 stages of gonadal maturation while in freshwater, whereas ocean-maturing steelhead return to 22 freshwater (December-April) much closer to the time of spawning. As a spring spawner, low 23 rates of iteroparity in stream-maturing steelhead have been commonly attributed to energy 2 24 limitations due to longer migration distances and the extended time in freshwater compared to 25 ocean-maturing steelhead (Burgner et al. 1992). However, high post-spawning mortalities in 26 ocean-maturing steelhead (Busby et al. 1996) has not been well explained. Considering the 27 differences in migration distance and time spent in freshwater between the two ecotypes, it 28 appears that factors other than simple energetic depletion may contribute to post-spawning 29 survival. 30 A defining characteristic expressed by both semelparous and iteroparous anadromous 31 salmonids, including steelhead, is the prolonged fasting that accompanies freshwater re-entry to 32 spawn. Most anadromous salmonids enter a period of voluntary anorexia during spawning 33 migrations and rely on lipids and protein stored in somatic and visceral tissues. It has been 34 theorized that by discontinuing active consumption in freshwater during spawning, Pacific 35 salmonids and Atlantic salmon can curtail the energy required for digestion, which can account 36 for as much as 40% of basal metabolism in feeding fish (Wang et al. 2006). Through fasting, 37 anadromous salmonids may able to conserve energy by significantly lowering the energetic 38 demands of their basal metabolism, thus allocating the bulk of their stored energy to support 39 upstream migration, completion of gonadal maturation, the physical exertion of spawning, and in 40 the case of iteroparous species, emigration back to the ocean. 41 Post-spawn steelhead and Atlantic salmon, known as kelts (Allan and Ritter 1976), are 42 known to re-initiate feeding activity in freshwater following spawning (Quinn and Myers 2004), 43 but it is presumed that majority of somatic energy is replaced in the ocean. However, the low 44 proportion of repeat-spawners in steelhead and Atlantic salmon populations suggests that most 45 kelts do not survive despite attempts at feeding. In some cases, such as stream-maturing 46 Columbia River steelhead the rates of repeat-spawning are so low that the stocks have often been 3 47 regarded, as functionally semelparous (Burgner et al. 1992). Little is known about the effects that 48 prolonged fasting and catabolism have on organ systems in steelhead or Atlantic salmon, 49 especially those involved in digestion, energy storage, and vitellogensis. Studies with other non- 50 mammalian vertebrates has shown that there is a gradual reduction of intestinal epithelium 51 (atrophy) and nutrient transport capacities of the gastrointestinal tract during fasting, but that 52 digestive processes are rapidly restored at the onset of feeding (Wang et al. 2006; Secor et al. 53 2002). It therefore seems apparent that for steelhead or Atlantic salmon kelts to recover 54 energetically, their digestive functions must also be restored. Additionally, little is known about 55 how other organ systems involved in energy storage, blood filtration, waste removal, and red 56 blood cell production respond in recovering kelts. 57 In the following study, Snake River steelhead were used as a model to qualitatively and 58 quantitatively evaluate the histological architecture of liver, spleen, gastrointestinal, and ovary 59 tissues to provide insight on the physiological effects of prolonged fasting in an iteroparous 60 salmonid. Histological assessments provide information about tissue function at the cellular level 61 and can reveal physiological information that may not be apparent via external condition and 62 non-lethal measures of nutrition (i.e. blood chemistry, fat-meter). The primary objectives of this 63 study were to (1) determine of Snake River kelts were attempting to replace energy via during 64 freshwater emigration, and (2) describe the gross changes in cellular architecture in pre- 65 spawning and post-spawning fish and evaluate the capacity for repeat-spawning based on the 66 functionality of tissues. 67 68 Methods Sample locations and procedures 4 69 Sections of the stomach, ovarian, liver, and spleen tissues of Snake River steelhead were 70 sampled in 2009 and 2010 at two phase of the reproductive cycle, first at or near sexual maturity 71 (mature) and secondly in migrating kelts (Table 1). Samples collected at maturity were obtained 72 from hatchery stocks returning to Dworshak National Fish Hatchery (DNFH) (46°30’N, - 73 116°19’W) or intercepted at Nez Perce Tribal Hatchery on the Clearwater River, ID (46°30’N, - 74 116°39’W ). Samples from kelts were collected from mixed stocks of Snake River steelhead at 75 the Lower Granite Dam juvenile bypass facility (46°39’N, -117°26’W), WA (Fig 1). We 76 sampled to select representative proportions of female and male steelhead at both phases for 77 comparison, but few male kelts were ever captured (Table 1). 78 Before sampling, steelhead were euthanized with a lethal dose (200mg/L) of MS-222 79 (Finquel, Argent Laboratories, Redmond, WA) buffered with NaHCO3 or in some cases CO2 was 80 used at DNFH. Fish were measured for fork length (nearest 0.5 cm). All mature steelhead were 81 considered to be in good external condition, whereas the condition of kelts was rated as good, 82 fair, or poor. Kelts in good condition exhibited no or only minor injuries, minimal fungal 83 infection, minor fin deterioration, silver or bright coloration, firm tissue texture, and were active. 84 Kelts in fair condition exhibited moderate injuries (non-life threatening), fungal infections over 85 1-10% of the body, moderate fin deterioration, pale coloration, and moderate activity. Kelts in 86 poor condition exhibited severe injuries (life threatening), fungal infections greater than 10% of 87 the body, severe fin deterioration, pale coloration, spongy tissue texture, and were listless. The 88 gastrointestinal tract of kelts was examined at the time of necropsy for the presence of 89 identifiable food or evidence of digestion via the presence of fecal material. Sections of the 90 stomach, liver, spleen, pyloric stomach, and ovarian tissues (kelts only) were removed using a 91 scalpel and transferred to jars containing 10% buffered neutral formalin (pH 6.8) at a 10:1 ratio 5 92 of formalin for tissue fixation. After fixation the stomach was sagittally bisected to provide a 93 cross section of the pyloric stomach. Fixed tissues were shipped to Colorado Histo-Prep (Fort 94 Collins, Colorado) for processing, where tissues were dehydrated, embedded in paraffin, 95 sectioned at ~ 4 to 6 μm, mounted on glass microscope slides, and stained with hematoxylin and 96 eosin (H&E) (Luna 1968). 97 Histological analysis 98 Histological analyses and scoring were accomplished with a compound light microscope 99 (Leitz Laborlux) fitted with a Leica EC3 camera and photographic software (LAS EZ 1.8.0; 100 Leica Microsystems, Cambridge, Ltd). Evaluations were made at magnifications of 4X, 10X, and 101 40X, which provided fields of view at 9.6 mm2, 1.5 mm2, and 0.1 mm2, respectively. We 102 evaluated tissue architecture with a categorical scoring system (Table 2). All scoring was 103 blinded. 104 Pyloric stomach 105 We evaluated and scored cross sections of the pyloric stomach using six metrics. We 106 quantified the detachment of the submucosa from the surrounding muscularis and stratum 107 compactum as severe, moderate, and minor to no detachment. The density of villi, which are 108 comprised of columnar epithelial cells and goblet cells lining the stomach lumen was scored as 109 low, moderate, and high. The extent of villiar invagination was estimated based on depth or 110 distance of the tip of villi (lamina epithelialis) to the stratum compactum. Villi invagination was 111 then scored as <1/4 the distance to the stratum compactum, 1/4 to 1/2 the distance to the stratum 112 compactum, or > 1/2 the distance to the stratum compactum. The cellular integrity of columnar 113 epithelial cells comprising the villi was evaluated and ranked by observing the severity of 6 114 deterioration of the cell membrane, cytoplasm and nucleus. The results were scored as severe, 115 moderate, and minor to no (absent) deterioration. The length to width ratio of columnar epithelial 116 cells was assessed to examine changes in overall cell size. As the name suggests, it was assumed 117 that columnar epithelial cells would exhibit length to width ratios similar to a rectangle (length > 118 width) rather than square (length = width). The length to width ratio of columnar epithelial cells 119 was ranked either as < 1/2 length to width ratio or > 1/2 length to width ratio. These measures 120 were validated via micrometer-calibrated measurements of photographs. Finally, we evaluated 121 the lamina epithelialis for the presence or absence of goblet or mucous cells to provide an 122 indication of mucous secretion. All scores for the pyloric stomach were assessed over the entire 123 tissue, regardless of magnification. 124 Ovary 125 We enumerated the quantity of perinucleolar oocytes (pre-lipid deposition), the quantity of 126 early/late cortical alveolus stage oocytes (lipid deposition present), and the quantity of 127 vitellogenic oocytes ( lipid and yolk deposition) in three to four fields for each ovary section. 128 Only oocytes sectioned through the nucleus or exhibiting clear lipid or yolk deposition were 129 enumerated in each field and it was assumed that all oocytes were randomly distributed within 130 the ovary. To validate this assumption, we compared the frequency of oocyte counts by category 131 and field. We found 92% of the ovarian tissues (N = 51) showed no significant variations across 132 fields and concluded pre and post-lipid oocyte distribution was random and could be averaged 133 across all fields by dividing the total sum of pre and post-lipid oocytes by the total number of 134 fields examined. 135 Liver 7 136 Five metrics were assessed (Table 2): The density of melanomacrophage aggregates 137 dispersed in the parenchyma was scored as none visible, low (< 5 aggregates), and high (> 5 138 aggregates). The separation of hepatocytes within their respective cords was evaluated for the 139 amount of detachment and scored as light, moderate, and minor or none. The overall separation 140 between heptatocyte cords (sinusoid spaces) was estimated and classified as less than 10μm or 141 greater than 10μm. The proportion of vacuolization within parenchyma was observed and ranked 142 as absent, between 1-10% in the field of view, and greater than 10% in the field of view. 143 Hepatocyte cellular structure was scored to consider integrity of the cell membrane, cytoplasm, 144 and nucleus and ranked as severe, moderate, or minor to no (absent) cellular deterioration. The 145 density of melanomacrophage aggregates, the apparent hepatocyte detachment, and sinusoid 146 spacing were scored over the entire liver tissue area at 10X. Scores for the degree of 147 vacuolization and hepatocyte integrity from each microscopic field assessed at 40X were 148 averaged and rounded up. 149 Spleen 150 Spleen tissues were evaluated to consider the distribution of white pulp nodules within the 151 red pulp as discrete, or diffuse wherein no clear separation of the white and red pulp was visible. 152 The percent of fields occupied by red pulp was estimated and categorized as 0-39%, 40-60%, or 153 61-100%. We evaluated the presence or absence of maturing erythrocytes and leukocytes by 154 noting the presence of differentiating cells or cell types. All scores for the spleen were derived 155 from assessments of the entire tissue, regardless of the magnification used. 156 Statistical analysis 8 157 Statistical comparisons of the frequency of scoring metrics by category within each tissue 158 were made to explore the differences within and between both mature and kelt phases. Within 159 mature steelhead we evaluated differences between sexes. Within the sample of female kelts we 160 evaluated metrics by condition, and also to evaluated differences between kelts with and without 161 presence of food at necropsy. Because sample sizes for male kelts were small (poor = 3, fair = 5, 162 and good = 5) we did not make statistical comparisons between males and female kelts. We 163 compared metrics between mature and kelt steelhead for good condition females only. 164 Statistical comparisons of the frequency of scores by category were made using exact Chi- 165 square analyses with Monte Carlo estimates derived from 20,000 permutations. These estimates 166 were used to account for occasional low frequencies (< 5 per cell) in which case the use of an 167 asymptotic Chi-Square test was not valid. Comparisons of oocyte counts in kelts by condition 168 were tested using Wilcoxon rank-sum tests. The relationship of fish length to several scoring 169 metrics transformed to progressive numerical scores was examined using Spearman correlations. 170 All statistical tests were conducted using SAS 9.3 (SAS Institute, Carey, North Carolina), and 171 statistical significance against a null hypotheses was reported as a P value. 172 173 174 Results Evidence of feeding Of the 65 migrating kelts examined, 25 had food or fecal material in the gastrointestinal tract. 175 Food items included salmon smolts, other small unidentified fish, fish eggs, and various 176 terrestrial and aquatic invertebrates. The proportion of feeding kelts significantly decreased with 177 decreasing condition (58% of good condition versus 7% of poor condition; P < 0.01, Table 3). In 9 178 male kelts, we detected no significant relationship between the proportion of feeding fish and 179 fish condition. 180 Pyloric stomach 181 We found no significant variation between male and female mature steelhead in any of the 182 metrics scored in the pyloric stomach, and all fish were considered in good condition. However, 183 we found significant differences attributed to fish condition in female kelt scores of submucosa 184 integrity, villi density, villi invagination, and columnar epithelial cell integrity (P < 0.02; Fig 2). 185 We found 23 of the 38 female kelts in good or fair condition had minor to no detachment of the 186 submucosa from the muscularis of the pyloric stomach, whereas the majority of poor condition 187 kelts (12/14) exhibited severe and moderate detachment of the submucosa. The villi density was 188 highest in good condition female kelts (Figs 3 and 4). The invagination of villi to the stratum 189 compactum was highly variable between good, fair, and poor condition kelts. Invaginations in 190 good condition kelts were all > 1/4 the distance to the stratum compactum and 60% of those 191 evaluated were > 1/2 the distance. Fair condition kelts exhibited a wide range of villi 192 invagination, and nearly all poor condition kelts showed villi invagination < 1/2 the distance to 193 the stratum compactum (Fig 2). Regardless of condition, the majority of kelts showed minor to 194 no signs of columnar epithelial cell deterioration. 195 Comparisons between stomach tissues of mature and kelt steelhead showed several 196 differences in structural patterns likely related to feeding recovery. Over 60% of mature 197 steelhead had low densities of villi but over 70% of kelts had significantly higher villi density (P 198 < 0.0001; Fig 5). We found significant differences between the extent of invagination of villi 199 between mature and kelt steelhead (exact P < 0.0001). The depth of villi invagination in mature 10 200 steelhead was very shallow, and nearly 80% of villi measured were <1/4 the distance to the 201 stratum compactum. In kelts, villi invagination was more pronounced and all were > 1/4 the 202 distance to the stratum compactum. No significant variations in submucosa integrity, columnar 203 epithelial cell integrity, length to width ratio of columnar epithelial cells, or presence of goblet 204 cells were found between mature or kelt steelhead. 205 We found a significant negative correlation of submucosa integrity with fish size in mature 206 steelhead (r = -0.4654; P <0.007; N = 33). However, compared to kelts, the range in length of 207 mature fish sampled (72 to 82 cm) was different from that of kelts (52 to 82 cm) sampled. We 208 found no such relationship with fork length in kelts. No mature steelhead exhibited severe 209 submucosa deterioration and very few (< 5) showed moderate deterioration, and the lack of 210 correlation could be influenced by sample sizes in each scoring category. 211 Ovary 212 No vitellogenic oocytes were observed in any of the ovarian tissues from kelts. However, 213 perinucleolar and early/late stage cortical alveolus stage oocytes were observed indicating that 214 Snake River steelhead are capable of repeat-spawning. Oocyte atresia was infrequent. We found 215 no variations in perinucleolar and early/late stage cortical alveolus oocytes by fish condition (Fig 216 6). However, we detected a negative correlation between fish fork length and perinucleolar 217 oocytes (r = -0.606 ; P < 0.002; N = 24). There was no significant correlation with length and 218 counts of early/late stage cortical alveolus oocytes. 219 Liver 11 220 We detected a significantly higher proportion of vacuolization in liver parenchyma of mature 221 steelhead males compared to that observed in mature females (P = 0.043; Fig 7). Mature female 222 steelhead had significantly lower proportions of cellular deterioration in hepatocytes than 223 proportions observed in males (P <0.001). We found no significant differences in the density of 224 melanomacrophage aggregates, the amount of hepatocyte detachment, or sinusoid spacing 225 between male and female mature steelhead. 226 In kelts, we observed that hepatocyte integrity varied among good, fair, and poor condition 227 females. Over 60% of good condition (N=24) and poor condition (N=14) kelts had minor to no 228 hepatocyte deterioration, whereas approximately 80% of fair condition kelts (N=13) showed 229 moderate deterioration (Fig 9). In samples with deterioration, we noted disintegration or loss of 230 the cellular membrane and nucleation within hepatocytes. Hypertrophy of hepatocytes was 231 occasionally observed and some exhibited vacuolization within the cytoplasm (Fig 9). We found 232 no significant variation in the proportion of melanomacrophage aggregates, attachment of 233 hepatocytes, spacing of sinusoids, or vacuolization among good, fair, and poor female kelts. 234 Our comparisons of liver tissues between good condition mature and kelt steelhead 235 determined that the proportion of mature females with minor to no hepatocyte cellular 236 deterioration was significantly higher than that of kelts (P = 0.006; Figs 8 and 9). Good condition 237 female kelts exhibited moderate hepatocyte detachment within cords (P = 0.002), in contrast 238 with samples from mature fish that had little to none. Variations in the degree of vacuolization in 239 livers between mature and kelt steelhead were noted (P = 0.05). We observed approximately 240 60% of livers from mature steelhead with vacuolization scores between 1 to 10%, and 20% with 241 scores > 10%, while nearly 50% of kelts showed a complete absence of liver vacuolization. 242 Small areas of melanomacrophage aggregates were present in all mature and kelt livers and 12 243 clustered near bile ducts and veins. Sinusoid spacing was consistent between phases and rarely 244 exceeded 10 μm between hepatocyte cords. 245 Spleen 246 We found significant differences in the proportion of red pulp in male versus female mature 247 steelhead (P = 0.001; Fig 10). Proportionally, 70% of male steelhead exhibited red pulp ratios 248 greater than 61%. In migrating female kelts, differences in the proportions of cellular 249 differentiation in the spleen were associated with fish condition. We observed some evidence of 250 red blood cell differentiation as well as leukocyte production or infiltration in 75% of kelt 251 spleens. Only fair and poor condition kelt spleens were observed without cell production or 252 monocyte activity (Fig 11). No other metrics were associated with fish condition. Regardless of 253 reproductive phase, we observed that spleens with higher proportions of white pulp had greater 254 amounts of cell differentiation in the white pulp. In spleens with elevated proportions of red 255 pulp, mature erythrocytes were observed as the dominant cell type. 256 We found a statistically significant relationship between fish length and the proportion of red 257 pulp in kelts (r = 0.4347; P = 0.034; N = 24). Although this trend suggested that larger kelts had 258 higher proportions of red pulp in spleens, these relationships were driven by two large kelts (73 259 and 75 cm) with red pulp ratios > 61%. 260 Discussion 261 The energetic and physiological deterioration of semelparous Pacific salmon during 262 spawning has been documented extensively (Green 1916; Robertson and Wexler 1960; 263 Robertson et al. 1961; Hane et al. 1959; Brett 1995; McPhee and Quinn 1998; Barry et al. 2010; 264 Morbey et al. 2005), but has not been well described in anadromous iteroparous salmonids 13 265 (Belding 1934). Intuitively, it would seem that any fish capable of repeat spawning would not 266 undergo the same irreversible cellular degeneration as occurs in semelparous salmon. We found 267 little evidence of extensive cellular deterioration in the pyloric stomach, ovary, liver or spleen in 268 mature or kelt steelhead suggesting that the physiological processes causing rapid senescence 269 and death in semelparous salmon are likely not the same in steelhead. Furthermore, the 270 observation of active feeding in migrating kelts indicates that these fish are attempting to replace 271 energy even while still in freshwater. 272 Gastrointestinal response 273 Steelhead are not believed to feed actively during freshwater migrations prior to spawning, 274 but occasionally freshwater feeding has been documented in mature Pacific (Garner et al. 2009) 275 and Atlantic salmon (Johansen 2001). Reductions or loss of villi and microvilli in the 276 gastrointestinal tract is a common characteristic found in fasting or starving poikilothermic 277 organisms (i.e. fish, snakes) (Ehrlich et al. 1976; Secor et. al. 2002; Krogdahl and Bakke- 278 McKellep 2005; Wang et al. 2006). Considering that the majority of mature steelhead at DNFH 279 were held on site for 2-3 months and not provided any opportunities to feed prior to sampling, 280 we are confident these fish were in a fasting state. The pyloric stomach of mature steelhead had 281 low densities of villi and almost little villiar invagination. These observation were similar to the 282 histological assessments of post-spawn sockeye O. nerka and Chinook O. tshawytscha salmon 283 (Robertson and Wexler 1960). The reduction of villi in fasting fish has also been documented in 284 non-salmonid species. Zeng et al. (2012) showed that the gastrointestinal tract and liver in food 285 deprived juvenile catfish (Siluris meridionalis) were down-regulated as a physiological response 286 to endure unfavorable environmental conditions. However, it is important to note that cellular 14 287 deterioration of the columnar epithelial cells rarely occurred in mature steelhead suggesting that 288 an irreversible degeneration was not occurring. 289 From an evolutionary perspective, the volitional anorexia in anadromous semelparous and 290 some iteroparous salmonids is likely the result of returning to a less productive environment 291 (freshwater) at a generally much larger size (McDowell 1987; Gross 1987; Fleming 1998). Wang 292 et al. (2006) hypothesized that reductions in organ size and enzymatic activity, specifically in the 293 gastrointestinal tract, could provide considerable energy savings to fasting or starving organisms 294 via lowering basal metabolic demands. This hypothesis is especially relevant to fasting 295 anadromous salmonids preparing to spawn, which primarily rely on somatic energy stores for 296 migration and spawning. Bioenergetically, it is conceivable that anadromous salmon returning to 297 spawn in a less productive environment at a larger size would save energy by fasting for three 298 primary reasons: (1) they would not actively pursue food with little energetic gain, (2) lower 299 basal metabolic costs by not operating the gastrointestinal tract, and (3) the use of stored somatic 300 energy forgoes the costs of specific dynamic action (loss of energy during digestion). Therefore 301 we agree with Wang et al (2006) hypothesis and further hypothesize that the gastrointestinal tract 302 of iteroparous salminds enters a cellular stasis or hibernation during reproduction, which is 303 supported by the observed renewal of villi and invagination found in kelts. 304 It is unclear if steelhead kelts must first ingest food to begin gastrointestinal recovery. Secor 305 et al. (2002) found that following prolonged periods of fasting in pythons that intestinal mass and 306 microvilli density rapidly increased following the ingestion of specific amino acids and peptides. 307 The columnar epithelial cells and goblet cells comprising the villi are important in the secretion 308 of enzymes, mucous, as well as the uptake of nutrients in the gastrointestinal tract (Mader 2001). 309 Presumably, the observed increases in villi density and invagination provide more surface area 15 310 for the digestion and absorption of food. Significant variations in villi density and invagination 311 did occur between good, fair, and poor kelts, which is possibly due to the finding that good 312 condition kelts were more likely to feed than fair or poor condition kelts. However, comparisons 313 between feeding and non-feeding good condition female kelts found no significant variations in 314 any of the pyloric scoring metrics. In reconditioning experiments with Atlantic salmon kelts, 315 Johnston et al. (1987) found that many kelts began feeding soon after spawning, but that many 316 had difficulties in swallowing feed and only after swallowing food several times did the rejection 317 of feed become less prevalent. This suggests that gastrointestinal regeneration may not occur at 318 the same rate or may be more difficult for some kelts than others. One possibility is that non- 319 feeding kelts had simply evacuated food from the gastrointestinal tract prior to sampling. 320 Carnivorous species like steelhead have short gastrointestinal tracts and the residence period of 321 food in the gut is relatively short (Arrington et al. 2002), thus it is possible feeding was 322 undetected in some kelts. Additionally, it has been shown that acute stress can alter the cellular 323 composition and reduce the permeability of the gastrointestinal tract (Olsen et al. 2005; Olsen et 324 al. 2008). Therefore, it is possible that stresses not detected via external examination could also 325 potentially explain the variations in cellular microstructure between feeding and non-feeding 326 kelts. 327 In addition to feeding, the gastrointestinal tract is also important in osmoregulation in 328 migrating kelts. As in smolts, the migration of kelts from a hypo- to hyper-tonic environment 329 requires a change in the osmoregulation of ions from the gills to the gastrointestinal tract. The 330 process of the re-acclimation to seawater has been shown to be greatly accelerated in kelts. 331 Talbot et al. (1992) found that Atlantic salmon kelts were able to re-adapt to seawater within 48 332 hours, thus indicating that the gut and all organs responsible for the salt and water balance had 16 333 recovered some or all function 4 to 6 weeks after spawning. This quick acclimation is likely 334 beneficial to kelts on two levels. Not only does it reduce the stress and time required to re- 335 acclimate in the estuary allowing for quicker access to high energy food in the ocean, but it likely 336 aids in the recovery from many freshwater-specific external infections such as (Saprolegnia sp.) 337 and parasitic copepods (Salmincola sp). 338 Ovarian composition 339 We observed perinucleolar (pre-lipid) and early/late stages of cortical alveolus oocytes (post- 340 lipid) in Snake River kelts, suggesting that these populations are capable of re-maturation. Like 341 their freshwater resident conspecifics, steelhead exhibit group synchronous ovarian development 342 characterized by the development of one primary group of oocytes at a time, which are 343 accompanied by numerous “resting” previtellogenic oocyte stages (Blazer 2002). This type of 344 ovarian development pattern has also been defined as determinant fecundity, where the 345 recruitment of oocytes in relation to secondary growth is arrested prior to spawning (Rideout and 346 Tomkiewicz 2011). Interestingly, Robertson and Wexler (1960) found small nucleated oocytes 347 present in the ovaries in spawning sockeye salmon, but not in the ovaries of Chinook salmon. 348 Exceptions to repeat spawning have been found for some salmonids that were previously 349 considered to follow strict semelparous life histories. Tsiger et al. (1994) found that male masu 350 salmon O. masou maturing and spawning in freshwater could later adopt anadromous life 351 histories and return to spawn again. Unwin et al. (1999) found that precocious male Chinook 352 salmon could be spawn and remature to spawn again under experimental conditions. 353 All oocytes found in Snake River steelhead ovaries were in previtellogenic stages indicating 354 that vitellogenesis was not occurring. This finding is not surprising considering that these fish 355 had recently ovulated the primary group of oocytes during spawning (<1 – 2 months earlier). 17 356 Oogenesis is an energetically expensive process for all fish species (Rideout and Tomkiewicz 357 2011), thus immediate gonadal recrudescence would not be expected to occur in a post-spawning 358 fish with depleted somatic energy. Johnson et al. (1987) found oogenesis and vitellogenesis 359 proceeded in Altantic salmon kelts only after restoration of body energy reserves. In our studies, 360 the exact date and location of spawning were unknown for kelts, but it is unlikely that any of 361 these fish had adequate time to restore their total body energy to re-initiate oogenesis and 362 vitellogenesis, even with the re-initiation of feeding. 363 Recent studies by Null et al. (2013) on acoustically tagged steelhead kelts in the Sacramento 364 documented that 10% of kelts rematured or residualized in freshwater. Although residualized 365 kelts grew less than those that returned to the ocean, their survival was higher. Steelhead that re- 366 mature can remain in the ocean a year or more before return migration (Burgner et al.1992). 367 Variations in spawning periodicity, known as skip-spawning, are common and can occur in 368 mature fish due to low somatic energy, poor physical health, or poor environmental conditions 369 (Rideout et al. 2005; Rideout and Tomkiewicz 2011). In anadromous species, migrating long 370 distances, skip spawning is common. We found no female kelts in our study that had retained 371 ripe eggs, and some follicular atresia was observed. The absence of vitellogenisis in Snake River 372 kelt ovaries in our studies suggests that the pre-vitellogenic oocytes were in a resting stage, 373 presumably until re-feeding was initiated and somatic energy reserves restored. 374 With the extended migration distance to return to the ocean it is likely that Snake River 375 steelhead would exhibit skip-spawning to re-build energy stores. Jonsson et al. (1991) found 376 smaller Atlantic salmon kelts (< 60cm) were more likely to exhibit consecutive spawning, 377 whereas larger kelts were more likely to exhibit skip-spawning (> 90 cm), and smaller kelts had 378 higher survival. The residualized and anadromous steelhead kelts tagged by Null et al. (2013) 18 379 were observed as consecutive spawners, and were <60 cm. It has been shown that larger 380 iteroparous salmonids generally exhibit lower post-spawning survival over smaller salmonids 381 likely due to difficulty in replacing energy stores (Fleming 1998). 382 We found the number of pre-lipid oocytes was negatively correlated with fork length. Our 383 finding is of particular interest as stocks in the Snake River basin that contain two steelhead run- 384 types, known as A and B-run (Brannon et al. 2004). The A-run steelhead return to freshwater 385 earlier (July to August), have shorter marine residences (1-2 years), are smaller in size (<70 cm), 386 and generally spawn earlier (March-April) than B-run fish. The B-run steelhead return later 387 (August to October), have longer marine residence (2-3 years), are generally larger (>70 cm), 388 and spawn later (April-June). During our sampling of kelts in 2009 and 2010, we observed lower 389 proportions of large kelts in the juvenile bypass system at the Lower Granite Dam (unpublished 390 data). 391 We hypothesize that the lower number of perinucleolar oocytes in larger B-run kelts may 392 relate to greater energetic investments in size to increase fecundity, egg size, and juvenile 393 survival, thus exhibit a life history similar to semelparity (Fleming 1998; Crespi and Teo 2002; 394 Fleming and Reynolds 2004). Observations of acoustically tagged Snake River steelhead by 395 Jones (2013) do not support a higher mortality of large B-run kelts following spawning. Of the 396 30 B-run kelts tagged at Fish Creek on the Lochsa River, ID, 29 were detected at the Lower 397 Granite Dam forebay indicating they successfully migrated from the spawning system. 398 Unfortunately, estimates of the quantity of B-runs steelhead passing the Lower Granite Dam via 399 routes other than the bypass facility are unavailable. Regardless, the successful migration of 400 large B-run kelts does not suggest that size is as important as restoring energy. 19 401 Liver and spleen responses 402 Our assessments of mature and kelt steelhead livers showed little evidence of severe liver 403 deterioration or loss of function at the cellular level. Of the two phases examined, tissues in kelts 404 had the most hepatocyte detachment and deterioration, although samples from < 10% of kelts 405 showed severe hepatocyte deterioration. Robertson and Wexler (1960) reported pronounced 406 hepatocyte degeneration, that included diminishments in cytoplasm and occasionally a complete 407 loss of nucleation in spawning semelparous salmon. 408 Compared to white muscle tissue the liver is not a primary energy storage tissue in 409 anadromous salmonids (Brett 1995), but it does serve as a readily accessible energy source 410 during fasting or starvation. In semelparous pink salmon O. gorbuscha, Phleger (1971) found 411 that the liver lost the ability to synthesize triglycerides following freshwater re-entry and 412 spawning. Mommsen et al. (1980) examined the enzymatic activity of Fraser River sockeye 413 salmon O. nerka livers during spawning migrations and found a gradual decrease in metabolic 414 enzymes and protein content as fish progressed toward spawning. In salmonids, carbohydrates 415 comprise only a very small component of white muscle tissue composition and are a not a 416 significant source of energy (Brett 1995), however the liver does store glycogen as a small 417 source of energy for immediate use when needed. French et al. (1983) found that that the liver of 418 post-spawn sockeye salmon increased in its capacity to synthesize glucose following complete 419 exhaustion of somatic lipids, perhaps providing a final energy source for nest guarding. Thus, it 420 appears that a complete loss of liver function is rare even among spawning semelparous species 421 and may explain the overall lack of hepatocyte deterioration found in Snake River steelhead 422 livers. 20 423 The effects of starvation and prolonged fasting on the cellular architecture of the liver have 424 been documented for a variety of fish species (Hochachka 1961; Connor et al. 1964; Vijayan and 425 Moon 1992; Hur et al. 2006; Rios et al. 2007; Kullgren et al. 2010). Among the noticeable 426 histological changes in the livers of starving fish are losses of glycogen or lipid vacuolization 427 (Wolf and Wolfe 2005; Hur et al. 2006). In our study, the only significant variation in liver 428 vacuolization was observed between male and female pre-spawn steelhead, where males had a 429 higher proportion of vacuolization. French et al. (1983) found that liver glycogen reserves in 430 Fraser River sockeye were highest in females just before spawning, but did not examine males. 431 As livers are important to vitellogenin synthesis, it is possible that glycogen reserves in female 432 were used to complete oogenesis before spawning. This difference between sexes may also 433 partially explain the variations in hepatocyte integrity between male and females. Wolf and 434 Wolfe (2005) note that it is not uncommon to observe hepatocye hypertrophy in the livers of 435 sexually maturing females due increases in estrogen vitellogenin. Therefore, it is possible that 436 the variations in vacuolization and hepatocyte integrity between mature male and female 437 steelhead reflected these differences in energy allocation for oogenesis. 438 Regardless of condition, sex, or phase all steelhead livers had > 5 melanomacrophage 439 aggregates. Agius and Roberts (2003) noted the melanomacrophage aggregates are generally 440 prolific in the kidney, spleens, and livers in starving fish. We did not observe large numbers of 441 aggregates, but perhaps a more refined method of scoring aggregates would have resolved 442 potential variation among factors of condition, sex, and mature and kelt phases. 443 We found little evidence of cellular deterioration or loss of spleen function in either mature 444 or kelt steelhead. Only fair and poor condition kelts showed a complete loss of cellular activity in 445 the spleen suggesting other potential pathological problems. Robertson and Wexler (1960) 21 446 detected a reduction in the quantity or complete disappearance of lymphoid cells and increase in 447 fibrous tissue in the spleen of spawning semelparous Chinook and sockeye salmon, but found 448 only 1 of 16 spleens examined with cellular necrosis. 449 We found a higher proportion of red pulp in males over females in comparisons between 450 mature steelhead. Rebok et al. (2011) found that white pulp increased in the spleens of female 451 Ohrid trout Salmo letnica during spawning, but did not examine males. Rebok et al. (2011) 452 postulated that gonadal maturation likely reduced the proportion of circulating lymphocytes 453 thereby potentially increasing the white pulp in the spleen of females. Variations in red pulp can 454 be caused by the use of erythrocytes for exercise and activity. Strenuous exercise influences the 455 proportion of erythrocytes stored in the spleens of O. mykiss, and other teleosts (Kita and Itazawa 456 1989; Franklin et al.1993). It is possible that the spleens of mature male steelhead in our study 457 contained higher proportions of erythrocytes in red pulp in preparation for competition on the 458 spawning grounds. 459 We found no variations in the arrangement of white pulp nodes by sex, condition, or phase in 460 our study, but red and white pulp nodes in teleosts are reported to be more diffuse than in 461 mammals (Mumford et al. 2012). Although not scored, melanomacrophage aggregates were 462 present in all steelhead spleens. Like the liver and kidney, prolonged fasting and starvation 463 generally increases the number of melanomacrophage aggregates (Agius and Roberts 1981, 464 Mizuno et al. 2002), thus future evaluations of spleen in response to prolonged fasting during 465 spawning may be warranted. 466 Summary 22 467 This study was the first to examine in detail the histological architecture of organs of 468 steelhead at maturity and as migrating kelts. Although many changes can be detected between 469 these two stages, the majority of steelhead showed little cellular atrophy that would be indicative 470 of loss of cellular function. Moreover, our measures of tissue histology have been paired with 471 assessment of plasma blood profiles (unpublished data) that indicate kelts are in homeostasis. 472 The evidence of developing oocytes, changes in the epithelial structures of the pyloric stomach 473 between maturity and kelt migration, and observations of feeding, all provide strong support for 474 potential iteroparity. 475 476 23 477 Acknowledgments 478 Funding for this study was provided by the Columbia Inter-Tribal Fish Commission, Doug 479 Hatch, project manager. We thank the staff at Dworshak National Fish hatchery for their 480 cooperation with sampling. The US Army Corps of Engineers, and Nez Perce Tribe provided 481 assistance with logistics of sampling at Lower Granite Dam juvenile fish bypass facility. We 482 thank Dr. Rajan Bawa and staff at Colorado Histo-prep for processing tissues samples for 483 microscopic analysis. Ann Norton, Boling Sun, James Nagler, Chris Williams, Andrew Pierce, 484 Victor Cajas of the University of Idaho provided invaluable assistance in interpretation of 485 samples, laboratory and field assistance. We thank Amy Long and Vicki Blazer for their reviews 486 of earlier drafts. Additional sampling support was provided by Aaron Penney, Jessica Buelow, 487 Kala Hamilton, Andrew Paper, William Schrader, Martin Perales, Veatasha Dorsey, and Bryan 488 Jones. This study was performed under the auspices of University of Idaho protocol # 2009-10. 489 Any use of trade, firm, or product names is for descriptive purposes only and does not imply 490 endorsement by the U.S. Government. 491 24 492 References 493 Agius C, Roberts RJ (2003) Melano-macrophage centres and their role in fish pathology. 494 Journal of Fish Diseases 26: 499-509 495 496 497 498 499 Allan I.R.H, Ritter J.A. (1976) Salmonid terminology. Journal Du Conseil International 37(3): 293-299 Arrington DA, Winemiller KO, Loftus WF, Akin S ( 2002) How often do fishes “run on empty”. Ecology 83(8):2145-2151. 500 Barry TP, Marwag A, Nunez S (2010) Inhibition of cortisol metabolism by 17α,20β-P: 501 Mechanism mediating semelparity in salmon? General and Comparative 502 Endocrinology 165:53-59 503 504 505 506 507 508 Behnke RJ (1992) Native trout of western North America. American Fisheries Society, Bethesda, Maryland Belding DL, (1934) The cause of the high mortality in the Atlantic salmon after spawning. Transactions of the American Fisheries Society 64(1): 219-224 Blazer VS (2002) Histopathological assessment of gonadal tissue in wild fishes. Fish Physiology and Biochemistry 26: 85-101 509 Brannon EL, Powell MS, Quinn TP,Talbot A (2004) Population structure of Columbia 510 River basin Chinook salmon and steelhead trout. Reviews in Fisheries Science 511 12:99-232 512 Brett JR (1995) Energetics. In: Groot C, Margolis L, Clarke WC (ed) Physiological 513 ecology of Pacific salmon. UBC Press, Vancouver British Columbia, pp 3-68 514 Burgner RL, Light JT, Margolis L, Okazaki T, Tautz A, Ito S (1992) Distribution and 515 origins of steelhead trout (Oncorhynchus mykiss) in offshore waters of the North 25 516 Pacific. International North Pacific Fisheries Commission. Bulletin Number 51: 517 1-92 518 Busby PJ, Wainwright TC, Bryant GJ, Lierheimer LJ, Waples RS, Waknitz FW, 519 Lagomarsino IV (1996) Status review of west coast steelhead from Washington, 520 Idaho, Oregon, and California. NOAA Technical Memorandum NMFS-NWFSC- 521 27. Northwest Fisheries Science Center, Seattle. Available: 522 http://www.nwr.noaa.gov. (September 2004) 523 Celius T, Walther BT, (1998) Oogenesis in Atlantic salmon (Salmo salar L.) occurs by 524 zonagenesis preceding vitellogenesis in vivo and in vitro. Journal of 525 Endocrinology 158:259-266 526 Christie MR, Marine ML, Blouin MS, (2011) Who are the missing parents? 527 Grandparentage analysis identifies multiple sources of gene flow into a wild 528 population. Molecular Ecology: 1-14 529 Crespi BJ, Teo R, (2002) Comparative phylogenetic analysis of the evolution of 530 semelparity and life history in salmonids fishes. Evolution 56(5): 1008-1020. 531 Docker MF, Heath DD, (2003) Genetic comparison between sympatric anadromous 532 steelhead and freshwater resident rainbow trout in British Columbia, Canada. 533 Conservation Genetics 4: 227-231 534 Dyck JC van, Pieterse GM, van Vuren JHJ (2007) Histological changes in the liver of 535 Oreochromis mossambicus (Cichlidae) after exposure to cadmium and zinc. 536 Ecotoxicology and Environmental Safety 66:432-440 26 537 Ehrich K., Blaxter JHS, Pemberton R, (1976) Morphological and histological changes 538 during the growth and starvation of herring and plaice larvae. Marine Biology 35: 539 105-118 540 Fleming IA, (1998) Pattern and variability in the breeding system of Atlantic salmon 541 (Salmo salar), with comparisons to other salmonids. Canadian Journal of 542 Fisheries and Aquatic Sciences 55 (Supplement 1): 59-76 543 Fleming IA, Reynolds JD, (2004) Salmon breeding systems. In: Hendry AP, Stearns SC 544 (ed) Evolution illuminated salmon and their relatives. Oxford University Press, 545 USA, pp 264-294 546 Franklin CE, Davison W, Mckenzie JC, (1993) The role of the spleen during exercise in 547 the Antarctic teleost, Pagothenia Borchgrevinki. Journal of Experimental Biology 548 174: 381-386 549 French, CJ, Hochachka PW, Mommsen TP, (1983) Metabolic organization of liver 550 during spawning migration sockeye salmon. American Journal of Physiology 551 245(6):R827-830 552 Garner, SR, Heath JW, Neff BD, (2009) Egg consumption in mature Pacific salmon 553 (Oncorhynchus sp.). Canadian Journal of Fisheries and Aquatic Sciences 66: 554 1546-1553 555 Gauthier D, Desjardins L, Robitaille JA,Vigneault Y (1989) River spawning of 556 artificially reconditioned Atlantic salmon (Salmo salar). Canadian Journal of 557 Fisheries and Aquatic Sciences 46: 824-826 27 558 Green C W (1916) On Some Quantitative Physiological Changes in the Pacific Salmon 559 during the Run to the Spawning Grounds. Transactions of the American Fisheries 560 Society, 45:1, 5-12 561 562 563 Gross MR, (1987) Evolution of diadromy in fishes. American Fisheries Society Symposium 1: 14-25 Hane S, Robertson OH (1959) Changes in plasma 17-hydoxycorticosteroids 564 accompanying sexual maturation and spawning of the Pacific salmon (Oncorhynchus 565 tschawtscha) and rainbow trout (Salmo gairdnerii). Proceedings of the National 566 Academy of Sciences of the United States of America 45(6):886-893 567 Hatch D, Anders P, Evans A, Blodgett J, Bosch B, Fast D, Newsome T, (2002) Kelt 568 Reconditioning: A Research Project to Enhance Iteroparity in Columbia Basin 569 Steelhead (Oncorhynchus mykiss)". 2001 Annual Report. Project No. 200001700, 89 570 electronic pages (BPA Report DOE/BP-00004185-2) Bonneville 571 572 573 Hochachka PW (1961) Liver glycogen reserves of interacting resident and introduced trout populations. Journal of Fisheries Research Board of Canada 18(1): 125-135 Hur JW, Jo JH, Park IS (2006) Effects of long-term starvation on hepatocyte 574 ultrastructure of olive flounder Paralichthys olivaceus. Ichthyological Reseach 53: 575 306-310 576 577 578 579 Johansen MJ (2001) Evidence of freshwater feeding by adult salmon in the Tana River, northern Norway. Journal of Fish Biology 59: 1405-1407 Johnston CE, Hambrook MJ, Gray RW, Davidson KG, (1992) Manipulation of reproductive function in Atlantic salmon (Salmo salar) kelts with controlled 28 580 photoperiod and temperature. Canadian Journal of Fisheries and Aquatic Sciences 49: 581 2055-2061 582 Johnston CE, Gray RW, McLennan A, Paterson A, (1987) Effects of photoperiod, 583 temperature, and diet on the reconditioning response, blood chemistry, and gonad 584 maturation of Atlantic salmon (Salmo salar) held in freshwater. Canadian Journal of 585 Fisheries and Aquatic Sciences 44:702-711 586 Jones B, (2013) Migratory and physiological characteristics of steelhead kelts from the 587 Clearwater River, Idaho, and Lower Granite Dam, Washington. Master’s thesis, 588 University of Idaho 589 Jonsson N, Hansen LP, Jonsson B, (1991) Variation in age, size, and repeat spawning of 590 adult Atlantic salmon in relation to river discharge. Journal of Animal Ecology 60(3): 591 937-947 592 Kita J, Itazawa Y, (1989) Release of erythrocytes from the spleen during exercise and 593 splenic constriction by adrenaline infusion in the rainbow trout. Japanese Journal of 594 Ichthyology 36: 48-52 595 Krogdahl Å, Bakke-McKellep AM (2005) Fasting and refeeding cause rapid changes in 596 intestinal mass and digestive enzyme capacities of Atlantic salmon (Salmo salar L.). 597 Comparative Biochemistry and Physiology, Part A 141: 450-460 598 Kullgren A, Samuelsson LM, Joakim Larsson DG, Gjornsson BT, Bergman EJ (2010) A 599 metabolic approach to elucidate effects of food deprivation in juvenile rainbow trout 600 (Oncorhynchus mykiss). American Journal Physiology-Regulatory, Integrative, and 601 Comparative Physiology 299:R1440-R1448 29 602 603 Luna LG, (1968) Manual of histological staining methods of the Armed Forces Institute of Pathology, 3rd edition. McGraw-Hill Book Company, New York 604 Mader SS, (2001) Biology, 7th edition. McGraw-Hill Inc. New York, NY. pp 786-789 605 McDowall RM, (1987) The occurrence and distribution of diadromy among fishes. 606 607 608 609 American Fisheries Society Symposium 1: 1-13 McDowall RM, (1997) The evolution of diadromy in fishes (revisited) and its place in phylogenetic analysis. Reviews in Fish Biology and Fisheries 7: 443-462 McPhee MV, Quinn TP, (1998) Factors affecting the duration of nest defense and 610 reproductive lifespan of female sockeye salmon, Oncorhynchus nerka. 611 Environmental Biology of Fishes 52: 369-375 612 Mizuno S, Misaka N, Miyakoshi Y, Takeuchi K, Kasahara N, (2002) Effects of starvation 613 on melano-macrophages in the kidney of masu salmon (Oncorhynchus masou). 614 Aquaculture 209: 247-255 615 Mommsen TP, French CJ, Hochachka PW (1980) Sites and patterns of protein and amino 616 acid utilization during the spawning migration of salmon. Canadian Journal of 617 Zoology 58: 1785-1799 618 619 620 Morbey YE, Brassil CE, Hendry AP (2005) Rapid senescence in Pacific salmon. The American Naturalist 166(5):556-568 Mumford S, Heidel J,Smith C, Morrison J, MacConnell B, Blazer V (2007) Fish 621 Histology and Histopathology. United States Fish and Wildlife Service Manual. 622 http://training.fws.gov/EC/Resources/Fish_Histology/Fish_Histology_Manual_v4.pdf 30 623 Narum S R, Hatch D, Talbot A, Moran P, Powell M, (2008) Iteroparity in complex 624 mating systems of steelhead Oncorhynchus mykiss (Walbaum). Journal of Fish 625 Biology 72:45-60 626 Null R E, Niemela KS, Hamelberg SF, (2013) Post-spawn migrations of hatchery-origin 627 Oncorhynchus mykiss kelts in the Central Valley of California. Environmental 628 Biology of Fishes 96:341-353 629 Olsen RE, Sundell K, Mayhew TM, Myklebust R, Ringǿ E (2005) Acute stress alters 630 intestinal function of rainbow trout Oncorhynchus mykiss (Walbaum). 631 Aquaculture 250: 480-495 632 Olsen, R.E., K. Sundell, E. Ringǿ, R. Myklebust, G. Hemre, T. Hansen, Karlsen Ø (2008) 633 The acute stress response in fed and food deprived Atlantic cod, Gadus morhua L. 634 Aquaculture 280: 232-241 635 Pascual M, Bentzen P, Rossi CR, Mackey G, Kinnison MT, Walker R (2001) First 636 documented case of anadromy in a population of introduced rainbow trout in 637 Patagonia, Argentina. Transactions of the American Fisheries Society. 130: 53-67 638 Phleger CF, (1971) Liver triglyceride failure in postspawning salmon, Lipids 6: 347-349 639 Pearse DE, Hayes SA, Bond MH, Hanson CV, Anderson EC, MacFarlane RB, Garza JC 640 (2009) Over the Falls? Rapid evolution of ecotypic differentiation in 641 steelhead/rainbow trout (Oncorhynchus mykiss). Journal of Heredity 100(5): 515- 642 525 643 644 QuinnTP, Myers KW (2004) Anadromy and the marine migrations of Pacific salmon and trout: Rounsefell revisited. Reviews in Fish Biology and Fisheries 14:421-442 31 645 Rebok K, Joranova M, Tavciovska-Vasileva I (2011) Spleen histology in the female 646 Ohrid trout, Salmo letnica (Kar.) (Teleostei, Salmonidae) during the reproductive 647 cycle. Archives of Biological Science 63 4:1023-1030 648 649 650 651 Rideout RM, Rose GA, Burton MPM, (2005). Skipped spawning in female iteroparous fishes. Fish and Fisheries 6:50-72 Rideout RM, Tomkiewicz J (2011) Skipped spawning in fishes: More common than you think. Marine and Coastal Fisheries 3(1): 176-189 652 Rios FS, Donatti L, Fernandes MN, Kalinin AL, Rantin FT (2007) Liver histopathology 653 and accumulation of melano-macrophage centres in Hoplias malabaricus after long- 654 term food deprivation and re-feeding. Journal of Fish Biology 71:1393-1406 655 Robertson OH, Wexler BC (1960) Histological changes in the organs and tissues of 656 migrating and spawning Pacific salmon (Genus Oncorhynchus). Endocrinology 657 66(2): 222-239 658 Robertson OH, Wexler BC, Miller BF (1961) Degenerative changes in the cardiovascular 659 system of spawning Pacific salmon (Oncorhynchus tshawtscha). Circulation Research 660 9: 826-834 661 Schaffer WM, Elson PF, (1975) The adaptive significance of variations in life history 662 among local populations of Atlantic salmon in North America. Ecology 56(3): 577- 663 590 664 Secor SM, Lane JS, Whang EE, Ashley SW, Diamond J, (2002) Luminal nutrient signals 665 for intestinal adaptation in pythons. American Journal of Gastrointestinal Physiology 666 283: G1298-G1309 32 667 Talbot C, Stagg RM, Eddy FB (1992) Renal, respiratory and ionic regulation in Atlantic 668 salmon (Salmo salar L.) kelts following transfer from freshwater to seawater. Journal 669 of Comparative Physiology 162(4): 358-364. 670 Thrower,FP, Hard JJ, Joyce JE (2004) Genetic architecture of growth and early life- 671 history transitions in anadromous and derived freshwater populations of steelhead. 672 Journal of Fish Biology 65 (Supplement A): 286-307 673 Tsiger VV, Skirin VI, Krupyanko NI, Kashlin KA, Semenchenko AY (1994) Life history 674 form of male masu salmon (Oncorhynchus masou) in South Primore, Russia. 675 Canadian Journal of Fisheries and Aquatic Sciences 51: 197-208. 676 Unwin MJ, Kinnison MT,Quinn, TP (1999) Exceptions to semelparity: Postmaturation 677 survival, morphology, and energetic of male Chinook salmon (Oncorhynchus 678 tshawytscha). Canadian Journal of Fisheries and Aquatic Sciences 56: 1172-1181/ 679 Vijayan MM, Moon TW (1992) Acute handling stress alters hepatic glycogen 680 metabolism in food-deprived rainbow trout (Oncorhynchus mykiss). Canadian Journal 681 of Fisheries and Aquatic Sciences 49: 2260-2266 682 Viola A E, Schuck ML, (1995) A method to reduce the abundance of residual hatchery 683 steelhead in rivers. North American Journal of Fisheries Management 15:488–493 684 Wang T, Hung CY, Randall DJ (2006) The comparative physiology of food deprivation: 685 686 687 688 689 From feast to famine. Annual Review of Physiology 68:223-251 Wingfield B, (1976) Holding summer steelhead adults over to spawn a second year. Northwest Fish Culture Conference 27: 63-64 Wolf JC, Wolfe MJ (2005) A brief overview of nonneoplastic hepatic toxicity in fish. Toxilogical Pathology 33: 75-85 33 690 Zeng LQ, Li FJ, Li XM, Cao AD, Fu SJ,Zhang YG (2012) The effects of starvation on 691 digestive tract function and structure in juvenile southern catfish (Siluris meridionalis 692 Chen). Comparative Biochemistry and Physiology, Part A 162: 200-211 34 693 Table 1. Number of fish sampled and tissues examined, separated by spawning year, and 694 reproductive stage. All but four sexually mature males and one female in 2010 were sampled at 695 Dworshak National Fish Hatchery. All kelts were sampled at Lower Granite Dam. 696 697 Phase Sex 698 699 2010 Total Liver, spleen and pyloric stomach Sexually mature 700 701 2009 Kelt 702 F 16 17 33 M 0 14 14 F 24* 29* 0 M 3 10 13 703 Ovary 704 Sexually mature F 0 11 11 705 Kelt F 24 29 53 706 * Indicates one sample was missing from analysis of pyloric stomach and liver. 707 708 35 709 Table 2. Summary of histological evaluations by tissue type and magnification in pyloric 710 stomach, liver, spleen, and ovarian tissues. Numerical scores and criteria are presented for each 711 metric. Tissue Metric (magnification) Scoring criteria Pyloric stomach Submucosa integrity (4X) 1 = Severe detachment from muscularis ; 2 = Moderate detachment from muscularis; 3 = Light or no detachment from muscularis Villi density (4X) 1 = Low ; 2 = Moderate; 3 = High Villi invagination (10X) 1 = <1/4 distance to stratum compactum; 2 = 1/4 ≤ 1/2 distance to stratum compactum; 3 = 1/2 distance to stratum compactum Columnar epithelial cell integrity 1 = Severe cell deterioration, (40X) poor nucleation; 2 = Moderate cell deterioration, nucleation present ; 3 = Light to no cell deterioration, nucleus present L:W ratio of columnar epithelial cells 1 = ≤1/2 length to width ratio; 2 (40X) = >1/2 length to width ratio; Presence of goblet cells (40X) 712 1 = Absent; 2 = Present 36 Table 2, continued. Tissue Liver Metric (magnification) Melanomacrophage aggregates (10X) Scoring criteria 1 = Absent; 2 = < 5; 3 = > 5 Hepatocyte dettachment from rosette 1 = Severe detachment (10X) 2 = Moderate; 3 = Light to no detachment Hepatocyte rosette spacing (10X) 1 = <10μm; 2 = >10μm Vacuolization in field of view (40X) 1 = absent; 2 = 1 to 10% 3 = > 10% Hepatocyte cell integrity (40X) 1 = Severe cell deterioration, poor nucleation ; 2 = Moderate cell deterioration, nucleation present; 3 = Light to no cell deterioration, nucleation present Spleen Distribution of white pulp 1 = Discrete; 2 = Diffuse Percentage of red pulp coverage (10X) 1 = 0-39%; 2 = 40-60% ; 3 = 61100% Cell production/differentiation (40X) 713 1 = Absent ; 2 = Present 37 Tissue Metric (magnification) Scoring criteria Ovary Quantity of perinucleolar oocytes (4X) Directly quantified and averaged in 3 to 4 separate zones Quantity of early/late cortical alveolus oocytes (4X) Directly quantified and averaged in 3 to 4 separate zones Quantity of vitellogenic oocytes (4X) Directly quantified and averaged in 3 to 4 separate zones 714 38 715 Table 3. Summary of proportion of kelts with food or evidence of feeding at time of necropsy at 716 Lower Granite Dam, by sex and fish condition. Monte Carlo permutation exact P values are 717 provided for comparisons by sex and condition. Sex Female Male 718 719 720 721 722 Condition Good N 24 Food (%) 14 (58) P Fair 14 6 (43) 0.0061 Poor 14 1 (7) Good 5 2 (40) Fair 5 2 (40) Poor 3 0 Total 65 25 (38) 0.5809 39 723 724 Figure Captions Fig 1 Map of Columbia/Snake River sub-basin with sampling sites indicated. 725 726 Fig 2 Significant scoring metrics for the pyloric stomach among good, fair, and poor condition 727 kelts for scores of A) submucosa integrity, B) villi density, C) villi invagination, and D) 728 columnar epithelial cell integrity. 729 730 Fig 3 Photomicrograph (4X) of pyloric stomach architecture in (a) mature steelhead (good 731 submucosa integrity, low villi density, <1/4 invagination), (b) poor condition kelt (poor 732 submucosa integrity, low villi density, <1/4 invagination), (c) fair condition kelt (fair submucosa 733 integrity, moderate villi density, and <1/2 invagination), and (d) good condition kelt (good 734 submucosa integrity, high villi density, >1/2 invagination). 735 736 Fig 4 Photomicrograph (40X) of columnar epithelial cell integrity in the pyloric stomach villi of 737 (a) poor condition kelt (severe to moderate cellular deterioration), (b) fair condition kelt 738 (moderate to minor cellular deterioration), and (c) good condition kelt (minor to no cellular 739 deterioration). 740 741 Fig 5 Significant scoring metrics for pyloric stomach histology between good condition female 742 sexually mature and kelt steelhead: A) villi density and B) villi invagination. 743 744 Fig 6 Photomicrograph (4X) of kelt ovaries (a) kelt ovary with high quantity of perinucleolar 745 (pre-lipid) and some early and late stage cortical alveolus (lipid deposition present) oocytes and 746 (b) ovary from a large kelt (> 70cm) with only early and late stage cortical alveolus oocytes. 40 747 748 Fig 7: Significant scoring metrics for the liver between male and female pre-spawn steelhead: A) 749 vacuolization and B) hepatocyte integrity. 750 751 Fig 8: Significant scoring metrics for liver histology between good condition female sexually 752 mature and kelt steelhead: A) hepatocyte detachment, B) vacuolization, and C) hepatocyte 753 integrity. 754 755 Fig 9. Photomicrograph (10X) of liver cellular architecture in (a) mature steelhead (high 756 vacuolization, minor to no hepatocyte detachment, minor to no hepatocyte deterioration), (b) 757 poor condition kelt (no vacuolization, severe hepatocyte detachment, severecellular 758 deterioration), (c) fair condition kelt (no vacuolization, minor to no hepatocyte deterioration), 759 and (d) good condition kelt (vacuolization present, minor to no hepatocyte attachment, minor to 760 no hepatocyte deterioration). 761 762 Fig 10 Comparison of red pulp coverage between the spleen of male and female pre-spawn 763 steelhead. 764 765 Fig 11 Photomicrograph (10X) of spleen cellular architecture in (a) mature steelhead (40-60% 766 red pulp, diffuse white pulp, and cellular activity present) (b) poor condition kelt (40-60% red 767 pulp, diffuse white pulp, loss of cellular activity), (c) fair condition kelt (61-100% red pulp, 768 white pulp discrete), cellular activity present), and (d) good condition kelt (0-39% red pulp, 769 white pulp diffuse, cellular activity present). 770 41 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 Fig 1. 42 795 796 100 797 A Good Fair Poor 799 Percent of fish 80 798 800 B 60 40 20 0 801 Severe Moderate Minor/absent Low Submucosa detachment 802 100 High Villi density C D 80 Percent of fish 803 804 Moderate 60 40 805 20 806 0 <1/4 808 809 811 >1/2 Invagination to stratum compactum 807 810 1/4 to 1/2 Fig 2. Severe Moderate Minor/absent Columnar epithelial cell deterioration 43 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 Fig 3. 44 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 Fig 4. 45 845 846 Mature Kelt 847 100 A 848 80 849 850 60 851 40 852 20 Percent of fish 853 854 855 856 0 Low Moderate High Villi density 100 B 857 80 858 60 859 860 40 861 20 862 863 0 <1/4 864 865 866 867 1/4 to 1/2 >1/2 Villi invagination to stratum compactum Fig 5. 46 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 Fig 6. 47 100 Vacuolization A Male Female 80 60 40 20 Percent of fish 0 Absent 1-10% >10% Hepatocyte deterioration 100 B 80 60 40 20 0 Severe 887 888 889 890 Fig 7. Moderate Minor/absent 48 Hepatocyte detachment 100 A Mature Kelt 80 60 40 20 0 Severe Percent of fish 100 Moderate Minor/absent Vacuolization B 80 60 40 20 0 Absent 1-10% >10% Hepatocyte deterioration 100 C 80 60 40 20 0 Severe 891 892 893 Fig 8. Moderate Minor/absent 49 894 895 896 897 898 899 900 901 902 Fig 9. 50 903 904 905 100 Male Female 906 80 Percent of fish 907 908 909 0 911 0-39% 913 916 917 918 919 920 921 40-60% 61-100% Proportion of red pulp 912 915 40 20 910 914 60 Fig 10. 51 922 923 924 925 926 927 928 929 930 931 932 933 934 935 Fig 11.