On October 12, 2007, a Friday during a long holiday weekend, 14 children aged between 9 and 11 boarded a white school bus at the entrance of the Professora Olinda Brito de Souza Municipal School, in the Parque das Laranjeiras neighborhood, in Manaus. The bus belonged to the city government.
The driver was a contracted employee. The excursion’s destination was the Tarumã resort, about 20 km from the city center, in an area of streams well-known to families in the northern zone. Two teachers accompanied the group. One of them, Marinalva dos Santos Queiroz, 52, signed the attendance list at 7:12 am with 14 names.
The bus departed at 7:15 AM. At 8:40 AM, none of the teachers had called the school administration to report their arrival and that the children were already going down to the edge of the stream. This was the last recorded communication. The bus was found at 4 PM that same day, parked on the shoulder of highway AM-010, 11 km from the resort, with empty seats, open windows, and a cold engine.
Inside, there were backpacks, half-empty water bottles, a child’s sandal in the hallway, and the attendance list signed by Marinalva, folded on the bulletin board. The two teachers were found that night, walking along the side of the road towards Manaus, in a state of shock. Neither of them could give a coherent account of what had happened.
The children weren’t with them, they weren’t at the resort, they weren’t on the road, they weren’t anywhere. This isn’t the story of an accident on the river. Although the press reported it that way in the first few weeks, it’s not a story about a mass kidnapping. Although the police worked with that hypothesis for months, it’s the story of 14 families from the same neighborhood, the same school, the same bus, who on a holiday morning sent their children on a day trip and spent 18 years without a meaningful answer.
It’s the story of a city that came to a standstill, of an investigation that unfolded into three inquiries, two state parliamentary commissions of inquiry, and a federal trial that never went to trial. And it’s the story of what was found in 2025, in a densely wooded area 7 km from the resort, by a surveying team that was taking measurements for a transmission line—something that didn’t close the case, but rather changed forever what was believed to be known about it.
The Parque das Laranjeiras neighborhood is located in the northern part of Manaus, in an area of simple houses, uneven asphalt streets, and backyards where you can hear the noise of the neighbors all day long. In 2007, the Olinda Brito de Souza school served approximately 400 elementary school children. The field trip to Tarumã was an end-of-year tradition, a reward for the students with the best attendance.
The parents signed authorization forms, and some prepared snacks. Others gave their children R$2 to buy an ice pop by the river. None of them imagined that that Friday morning would be the last time they would see their children leave home with backpacks on their backs and flip-flops on their feet.
If this kind of story makes you rethink how much we really know about what happens far from urban centers, where the forest and silence guard what the city forgets, you can subscribe to the channel and leave your comment. From here, the story unfolds slowly. Each detail needs to be told in the order in which it happened, because order is part of what has never been explained.
“How many hours does a mother wait before setting out to search for something on her own? What happens when 14 families discover at the same time that their children have not returned? At what point does the state acknowledge that it has lost 14 children and yet cannot explain how? What can be found in a forest after 18 years? And what does this mean for those who never stopped waiting?” We are in Manaus, in October 2007.
The temperature had been 34 degrees Celsius since 6 a.m. The Rio Negro was receding. The streams in the northern zone were low, but with exposed sandbanks where children used to play. The Manaus Fire Department has three search teams for a metropolitan area covering more than 11,000 km², encompassing urban and peri-urban areas.
The nearest police station to the Tarumã beach resort is the 19th Integrated Police District, which was operating with reduced staff during that holiday. The list was handwritten in blue pen on a lined notebook page that Marinalva tore from the notepad on her desk. 14 names, 14 ages, 14 signatures of guardians collected in the previous days.
In Alto da Folha, the date is October 12, 2007. And the destination is Balneário do Tarumã. Departure at 7 am, return scheduled for 3 pm. The school coordinator, Rosângela Pimentel, checked the list and stamped it. She then saved a copy in the semester’s events folder and went to get coffee. When the police requested a copy 48 hours later, Rosângela couldn’t find the file.
She said it might be in the principal’s office, since she was on vacation. The list only reappeared 11 days later, inside a dead file, with coffee spilled over three of the 14 names, making them illegible. The complete identification of the children took another four days. This delay was unusual by the standards of the Manaus municipal education system at that time.
The school operated with three shifts, nearly 1000 students in total, and an office with two employees who handled all the paperwork. The files were physical, kept in plastic folders inside steel cabinets that rusted from the humidity. There was no digitization or backup. The semester’s events folder was kept along with meeting minutes, lunch receipts, and letters from the Department of Education that nobody read.
It was the type of organization that functions on a daily basis, but collapses when someone urgently needs information. Detective Airton Brandão Teixeira, responsible for the case, noted in his preliminary report that the delay in fully identifying the children compromised the first 48 hours of the investigation.
Without knowing precisely who the 14 children were, it was not possible to notify all the families at the same time, nor to cross-reference information about who had seen what before the bus departed. Three of the families were only located on October 14th, two days after the disappearance. One of them, the Monteiro Duarte family, lived in an informal settlement on the banks of a stream.
“In the Santa Etelvina neighborhood, with no formal address, no landline phone, and no neighbors registered at the school. The father, Raimundo Monteiro Duarte, worked as a bricklayer’s assistant on a construction site on the other side of town and only learned that his son had disappeared when he returned home on Sunday night and found a police note taped to the door.”
The original attendance list, the one Marinalva signed and left folded on the bus dashboard, was collected by forensics and analyzed. The handwriting was Marinalva’s. The ink was consistent with a normal ballpoint pen, but there was a detail that was only noticed weeks later, when the handwriting expert reviewed the document at the request of the Public Prosecutor’s Office.
“Next to two of the names, there were small pencil marks, two diagonal lines, almost invisible to the naked eye, as if someone had marked those names for a specific reason. When questioned, Marinalva said she did not remember making those marks. The forensic team was unable to determine whether the mark had been made before or after boarding.”
The two names listed belonged to twin brothers, Lucas and Luan Figueira Barbosa, 10 years old. Their mother, Neusa Figueira, was a lunch lady at the same school. That morning, Neusa had helped prepare the lunches and personally put her two sons on the bus. She didn’t know what those markings meant. Nobody knew.
What was known was that the list existed, that the names were real, that the children had boarded the bus, and that, at some point between leaving school and the end of that afternoon, they all disappeared without a plausible explanation. The list was the only documentary proof that the trip had taken place, and for years it was treated by investigators as little more than a bureaucratic record.
It was only much later that someone realized it could be the most important piece of evidence in the case, and that it had been lost for 11 days inside a rusty cupboard, covered in coffee. The white bus with the inscription “Manaus City Hall, education” on the side was found by a truck driver named Osvaldo Bentes Cavalcante, who was driving the route between Manaus and Itacoatiara, and stopped to urinate on the side of the road.
It was 4:12 PM. Osvaldo noticed that the vehicle’s doors were open and the windows were down, but there was no one around, which he found strange, since that stretch of the AM-010 highway had no rest stops or access to any rural property. It was dense forest on both sides, with that smell of damp earth that lingers in the air after the quick midday rain.
Osvaldo looked inside and saw backpacks scattered across the seats, a thermos on the floor, and a child’s sandal in the aisle. He didn’t touch anything. He called the Federal Highway Police from a public phone at the nearest gas station, 6 km away. The Federal Highway Police took 40 minutes to arrive at the scene.
When the two agents got out of the vehicle, the sun was already low and the light filtered through the treetops with that orange hue that precedes nightfall in the Amazon. The bus was exactly as Osvaldo had described it. The front and back doors were open, the engine was off and cold to the touch, and the keys were not in the ignition.
The dashboard showed no signs of damage. The odometer displayed a mileage inconsistent with the route between the school and the resort, plus the additional 11 km to that point on the highway. There were no skid marks on the asphalt, no signs of a collision, no blood, either inside or outside the vehicle. The police took photos with a compact digital camera, 13 photos in total, which would later be attached to the investigation.
One of the photographs showed the driver’s seat with the backrest reclined, as if someone had adjusted it for resting. Another photograph showed the central aisle with four backpacks lying on the floor, one of them open with a school notebook partially visible. The most reproduced photo in the following years, the one that appeared in newspapers and television programs, was that of a child’s pink rubber sandal, size 32, abandoned between the second and third seats on the right side.
It was never determined to which child it belonged. None of the families recognized it for sure. Three mothers said it could be their son’s or daughter’s. The sandal remained in the Civil Police’s evidence storage for years, inside a transparent plastic bag with a yellow label. What caught the experts’ attention when they finally examined the vehicle two days later was the arrangement of the objects.
The backpacks were not in the seats where the children would presumably be sitting. They were scattered unevenly, as if someone had moved them. Some were open, others closed, and two were stacked on the back seat, which is usually the least occupied. There were no lunchboxes. The ones the parents had packed were not found on the bus.
There were no water bottles with the children’s names on them, although several mothers stated that they had sent bottles that already had labels. What was there were generic, unlabeled bottles, half-bottles of water, the kind you buy from distributors. Nobody could explain where they came from. The bus was towed to the 19th DIP’s yard the following morning and cordoned off with duct tape.
It remained there for 8 months, exposed to sun and rain, until a supplementary expert analysis was requested. When the second analysis was carried out in June 2008, some of the internal surfaces already showed mold and deterioration, and the fingerprints collected in the first analysis were insufficient for cross-referencing data.
Most of the fingerprints were partial or overlapping. No fingerprints of the driver were identified, which was technically impossible if he had driven the vehicle that morning. Either the fingerprints had been erased, or Edivaldo Nonato de Souza never touched that steering wheel. Highway patrol officers found Marinalva dos Santos Queiroz and another teacher, Aparecida Lima Fonseca, walking along the shoulder of the AM-010 highway towards Manaus at around 8 pm.
Marinalva was barefoot and her dress was torn at the shoulder. Aparecida was carrying her own purse and crying incessantly, but making no sound. It was a silent, open-mouthed cry, which the highway patrolman who rescued them described in his report as a state of severe shock. Neither of them could explain where the children were.
Marinalva kept repeating that they were just playing and that the vegetation had closed in. Aparecida said nothing for the first six hours. They were taken to the emergency room on August 28 and given medication. The formal statement only occurred three days later, and when it did, the two women’s versions of events hardly coincided.
Marinalva was 52 years old and had been a teacher in the municipal school system for 23 years. I knew the Tarumã locker room because I had taken other classes there before. At least four times since 2002. She was the kind of teacher that parents knew by name, who greeted them at the bakery, and whom mothers would call over for coffee when she passed by on the street.
He had no criminal record, no problems with the school administration, and no record of incidents on previous field trips. She was a widow, lived alone in an apartment in the Francisca Mendes complex, and dedicated her free time to the neighborhood’s Evangelical Church. No one in Parque das Laranjeiras had any reason to distrust Marinalva.
And that made everything harder to understand. Aparecida Lima Fonseca was younger, 34 years old, and had only been at the school for two years. She was a substitute teacher, hired for a fixed term, and this was her first field trip with the students. “I knew the children, but not as well as Marinalva.”
She lived in the Novo Aleixo neighborhood with her husband and 4-year-old son. Her husband, Wellington Fonseca, told police that Aparecida had left home in high spirits that morning, that she had prepared a corn cake to take with her, and that she seemed calm when they found her on the road, 8 hours after her last communication with the school.
Aparecida didn’t have the cake, nor did she have the suitcase with her documents. In fact, the bag she was carrying didn’t belong to her. It was a black nylon bag that neither of them recognized as theirs, and they had no answers to any of the questions. At the emergency room, on August 28, both were evaluated by an on-duty team.
Marinalva had mild dehydration, abrasions on her feet consistent with prolonged walking without shoes, and a superficial cut on her right shoulder that she said was caused by a branch. She also had scratches on her legs and arms, dirt under her fingernails, a slight bruise on her left knee, and the same signs of dehydration.
The on-call doctor observed that neither of them had serious injuries and that the emotional state of both was incompatible with a simple walk down the road. Marinalva alternated between incoherent phrases and catatonic silence. Aparecida trembled uncontrollably, even after taking her medication.
What Marinalva said in the following days varied from session to session. In her first testimony, she said that the children were playing near the stream and that she had gone to the bus to get sunscreen. She said that when she returned she found no one, neither the children, nor Aparecida, nor the driver. She said that she called out, shouted, searched the banks of the stream, and then went into the woods, but found no one.
He said that at a certain point he decided to return to the road and walk towards the city. She couldn’t explain how Aparecida appeared beside him on the road, nor when it happened. In the second session, three days later, Marinalva included a new element. She said that before going to the bus she saw a man in a green shirt approaching along the trail and that the children went towards him.
When the detective asked why she hadn’t mentioned the man before, Marinalva said she wasn’t sure if she had seen him or if she had dreamed about him. In the third session, with her lawyer present, Marinalva only said that she didn’t remember anything between the moment they got off the bus and the moment she was on the road.
He said there was a void in his memory. He said he wanted to help, but couldn’t. Aparecida maintained her version of events in all three sessions. “I was sleeping on the bus.” She said she had a headache, that she took a pill after they got to the locker room, and that she fell asleep in the back seat while the children left with Marinalva.
She said she woke up with the bus moving, looked out the window and didn’t recognize where she was, and that at some point the bus stopped and she got off. “I couldn’t say who was driving, I couldn’t say if the children were on the bus when it started moving. I couldn’t say how long he slept.” The detective asked how she explained the scratches on her arms and the dirt marks on her legs.
“If he was sleeping inside the bus.” Aparecida didn’t answer. He looked at the lawyer who had requested the suspension of the session. At that point in the investigation, Detective Airton Brandão Teixeira had two witnesses who were present at the time of the disappearance of 14 children, and neither of them could offer a version that corroborated the physical evidence.
The contradiction between the accounts was not subtle; it was open, evident, and, for those who read the documents, deeply disturbing. The news reached Parque das Laranjeiras in a disorganized manner, as often happens in neighborhoods where everyone knows each other and where landlines are still more reliable than cell phones. Around 6 p.m., the mother of one of the students, Dalva Cristina Rebolsas, called the school, asking why the bus hadn’t returned yet. No one answered. Dalva called her neighbor, who was the mother of another child in the group. The neighbor also knew nothing. In less than an hour, seven mothers were at the school gate, which was closed and the lights were off.
No authority had made contact. No announcement had been made. The night doorman, who arrived at seven, knew nothing of any excursion. Dalva was a seamstress; she worked from home, but on a Singer sewing machine that was in the living room, next to the window that looked out onto the street.
That Friday, she spent the entire day sewing, listening to the radio, and waiting for her son to come home so they could watch TV together that night, as they always did when he didn’t have school the next day. Her son, Bruno Henrique Rebolsas de Lima, was 10 years old, had worn glasses since he was seven, and was the type of guy who preferred reading comic books to playing soccer.
Dalva wasn’t worried until 5 PM, because the return trip was scheduled for 3 PM and delays on school buses were common. At 5:30 PM, he called Marinalva’s cell phone, which was turned off. He called again at 6 PM. It was off. “You called the school. Nobody answered.” That’s when he left home. The scene outside the school that night gradually took shape.
Primeiro as sete mães, depois os dois pais que regressaram do trabalho e foram avisados pelas mulheres. Depois havia uma avó que tomava conta de um dos alunos enquanto a mãe trabalhava num supermercado no centro da cidade. A seguir vieram vizinhos que não tinham filhos na excursão, mas que ouviram o barulho, para saber o que se estava a passar.
Por volta das 21h, havia cerca de 40 pessoas na calçada do lado de fora da escola, e nenhuma delas tinha qualquer informação. Não havia policiais, não havia representantes da prefeitura, e não havia ninguém da Secretaria da Educação. O porteiro, um homem chamado Cícero, tentou ligar para a diretora e para a coordenadora.
A diretora não atendeu. Rosângela Pimentel atendeu e disse que ia verificar, mas nunca apareceu. Foi a Dalva quem decidiu ir à esquadra. Apanhou um autocarro público com outra mãe, Francisca das Chagas Oliveira, e foram para a 19ª Esquadra. Chegaram lá por volta das 22 horas.
O turno era composto por um único investigador, que ouviu os relatos das duas mulheres. Ele anotou as informações num formulário de boletim de ocorrência e disse que precisava esperar 72 horas para registrar formalmente o desaparecimento. Dalva perguntou se ele entendia que se tratavam de 14 crianças. O investigador disse que entendia, mas que esse era o procedimento.
Dalva perguntou se ele tinha filhos. O investigador não respondeu. Ele pediu que elas voltassem na segunda-feira. Francisca das Chagas voltou para casa e não dormiu. Sentou-se na cozinha com a luz acesa, olhando fixamente para o relógio na parede. Seu marido, que trabalhava como vigia noturno, ligou do trabalho e ela lhe contou o que estava acontecendo.
“Ele disse que ia pedir para sair mais cedo, mas não podia.” A Francisca ficou sozinha até de madrugada. A Dalva também não dormiu, mas em vez de ficar em casa, voltou a sair. Às 5 da manhã do dia 13, apanhou outro autocarro e foi sozinha, às escuras, para a estância do Tarumã. “Ele chegou ao local quando o sol estava a nascer e caminhou pelas margens do riacho durante duas horas, chamando o nome do filho.”
“Não encontrou nada, não viu ninguém.” “Ele voltou para casa às 9h da manhã, com cortes nos pés e o rosto queimado de sol, e encontrou dois veículos da Polícia Civil estacionados no portão da escola.” A busca oficial começou 18 horas após o desaparecimento das crianças. Naquela noite, algo mudou no Parque das Laranjeiras. Não foi algo visível.
Não era algo para o qual se pudesse apontar o dedo. Era uma alteração no tecido do bairro, uma fenda que percorria todas as casas, todos os quintais e todas as conversas ao portão. 14 famílias mandaram os filhos para a escola, e as crianças não regressaram. E ninguém, ninguém as avisara. O Corpo de Bombeiros de Manaus iniciou a operação de buscas na manhã do dia 13 de outubro, 18 horas após o último contato das professoras com a escola.
The team consisted of 12 men and two sniffer dogs. The area designated for the initial sweep was the surroundings of the Tarumã beach resort, within a 3 km radius. In five days of intensive searches, with the support of the army from the third day onwards, no material trace of the children was found: no clothing, no shoes, no food scraps, no identifiable footprints.
The sniffer dogs trained to locate recent human presence did not indicate any trace in any direction from the beach resort. The captain in charge of the operation, Humberto Farias Monteiro, told the press that this was not normal, that 14 children do not disappear without leaving a single trace.
In October 2007, the Tarumã bathing area was a semi-urbanized space in the western zone of Manaus, where the Tarumã stream flowed into a river beach that received hundreds of families on weekends and holidays. There were food stalls, an improvised dirt parking lot, rudimentary bathrooms, and an access trail that descended down a red earth embankment to the water’s edge.
On the busiest days, you could find açaí vendors, children jumping from branches, loud music blaring from parked cars, and the smell of fried fish mixed with cheap sunscreen. However, on the October 12th holiday, the resort wasn’t at its peak when the bus arrived. At 8:40 am, the place was practically empty.
The firefighters who arrived the next day found the tents closed, and no residents in the vicinity said they had seen a school bus or a group of children that morning. This was part of the problem. In 2007, there were no security cameras in the resort area, no guardhouse, no access control, and no record of vehicles entering or leaving.
The only point of reference was a bar called Cantinho do Tarumã, located 300 meters from the Igarapé slope, which, on that holiday, only opened after 11 am. The bar owner, Sebastião Vieira Lopes, told the police that he saw nothing unusual. He said that when he opened the door there were no buses parked nearby and that the morning was quiet.
The firefighters expanded the search radius to 5 km on the third day and to 8 km on the second. They entered the woods with machetes, cleared paths through areas of dense vegetation, and searched the banks of the stream, upstream and downstream. They inspected natural pools, seasonal swamps, and ravines where water accumulates during the dry season.
Divers descended to three points in the creek, where the depth exceeded 2 meters. They found nothing, neither the divers, nor the dogs, nor the army soldiers who joined the operation with jungle equipment. Captain Humberto Farias Monteiro gave an interview to TV Amazonas on the fifth day of the search.
He was a man of measured speech, with 30 years of service, and not given to dramatic pronouncements. But in that interview, by the stream, his uniform covered in mud, he said something that stuck in the memory of those who followed the case. “I’ve searched for many people in this forest. Adults, children, river dwellers, tourists, there’s always something.”
“A footprint, a piece of fabric, a trail in the mud. There’s nothing here. It’s as if these children never set foot on this ground.” On the seventh day, the operation was scaled back. On the tenth day, officially concluded, the Fire Department issued a 14-page report concluding that land and water searches in the Tarumã bathing area and its surroundings did not result in locating the children, nor in identifying any traces that could indicate their presence at the site after the arrival time recorded by the teachers. The report recommended opening a criminal investigation to determine the circumstances of the disappearance. The criminal investigation was already underway, but was not progressing. The police inquiry conducted by Detective Airton Brandão Teixeira, of the 19th Precinct, collected statements from the two teachers in three separate sessions throughout October 2007.
In the first session, Marinalva said that the children were playing by the stream when she went to the bus to get sunscreen and, upon returning, found no one there. On Monday, the version changed. She said that a man in a green shirt appeared on the trail and that the children followed him, thinking he was the guide.
In the third interview, she said she remembered nothing between the moment they got off the bus and the moment she was walking down the road. Aparecida, in all three sessions, maintained a single statement: “I didn’t see what happened. I was sleeping on the bus.” But the hospital report recorded dirt marks and scratches on Aparecida’s legs and arms, inconsistent with those of someone who had remained seated in the vehicle.
Detective Airton was 48 years old and spent most of his career in neighborhood police stations in the northern zone of Manaus. He wasn’t an investigator of complex cases, he didn’t have a specialized team, he worked with two investigators, a clerk, and a structure that he himself described years later, in an interview with a local newspaper, as insufficient to investigate the theft of a bicycle, much less the disappearance of 14 children.
But he was the one who received the case, and he was the one who had to work with what he had. In the first 30 days, Airton collected testimonies from 42 people. The two teachers, the coordinator Rosângela, school employees, the children’s parents, residents near the resort, employees of Trans Rotas Norte, and two military police officers who were patrolling the AM-010 highway that day.
None of the testimonies provided a concrete clue. The military police officers said they did not see the bus stopped on the highway during their morning patrol, but admitted that the patrol covered an 80 km stretch and that it was impossible to see everything. The residents around the resort, mostly riverside dwellers living in stilt houses 1 or 2 km from the stream, said that October 12th was just another day.
No one heard screams, no one saw any unusual movement. No one saw a white bus. What bothered the police chief was the complete lack of consistency between the accounts of the only two adult witnesses. In a case with 14 victims and only two people present, some discrepancy in the details would be expected.
Times, sequence of events, children’s positions. But what Airton found wasn’t a discrepancy, it was a structural contradiction. Marinalva said the children were on the edge of the stream. Aparecida said she was sleeping and didn’t see anything. Marinalva said she went to the bus and when she returned she didn’t find anyone.
Aparecida said she was always on the bus and didn’t see Marinalva arrive. Marinalva mentioned a man in a green shirt. Aparecida never mentioned anyone. Marinalva changed her story at each session. Aparecida repeated the same phrase with mechanical precision. The police chief requested that the Public Prosecutor’s Office conduct psychological examinations on both teachers to assess whether they were fit to testify or if they presented any dissociative condition that could explain the contradictions.
The examination was conducted in November 2007 by a forensic psychologist from the Public Security Secretariat, who concluded that both women showed signs of acute trauma but were fit to testify. The psychologist added in a note at the end of the report that Aparecida’s response pattern—exact repetition of the same phrase, absence of emotional variations, refusal to elaborate—was consistent with severe trauma and also with prior rehearsal.
“It wasn’t possible to distinguish one thing from another without prolonged monitoring, which was never authorized. The families followed the investigation from afar. And because nobody invited them closer. Dalva Cristina Rebolsas asked to be heard three times, in addition to her initial statement. I wanted to talk about Marinalva’s behavior in the days leading up to the excursion, about a conversation I had with her at the school gate, about a detail that I thought was important.”
The officer said he would write it down. Dalva was never called again. The bus driver, registered in the city hall documents as Edivaldo Nonato de Souza, 41 years old, hired by a third-party company called Transrotas Norte, could not be located at any time after the disappearance. The company reported that Edivaldo had been hired two months earlier and that his documents were in order, but when the police checked, the CPF (Brazilian taxpayer identification number) belonged to a person who had died in 2003 in Porto Velho. The address provided in the contract did not exist, and the photograph on the company badge did not match any record in the available databases.
Edivaldo Nonato de Souza, if that was indeed his name, was never found. No witness from outside the school was able to accurately describe him. Transrotas Norte employees said he was a quiet guy who arrived early and didn’t talk to anyone.
Transrotas Norte was a small company registered in 2005, headquartered in a warehouse in the Industrial District of Manaus. Its contract with the city government stipulated the provision of drivers and vehicles for school transportation in five areas of the city. The company had eight registered buses and 12 drivers on its staff.
The contracts were renewed annually through a simplified bidding process, and the documentary requirements, as was later discovered, were minimal: a driver’s license (category D), a criminal record certificate, and proof of residence. There was no biometric verification, no cross-referencing of CPF numbers with death records, and no in-person interview conducted by the city hall.
The process was bureaucratic on paper and lax in practice. Edivaldo Nonato de Souza, or whoever used that name, had started working at Transrotas Norte in August 2007. He had been managing the Parque das Laranjeiras route for two months when the excursion took place. The company employees, who were interviewed by the police, gave vague and sometimes contradictory descriptions.
Operations manager Nelson Praia Monteiro said that Edivaldo was a man of average height, with dark skin, short hair, who wore a cap and spoke little. Secretary Ivone Sarmento said that he was tall and thin and that he didn’t wear a cap, but a dark beret. The mechanic who looked after the buses said that he never saw Edivaldo’s face properly because he arrived when the workshop was still closed and left before closing time.
None of them had a personal photo. The only photo available was the company identification badge, a low-resolution image printed on ordinary photographic paper showing a face partially covered by a wide-brimmed cap. When the police tried to process this image into the recognition systems available at the time, they obtained no match. The address provided by Edivaldo on his employment contract was Japurá Street.
114, Lírio do Vale neighborhood. The police went to the location and found a vacant lot with an abandoned building. Neighbors said that no one had lived there for at least 3 years. The CPF number provided did not belong to a man named Edivaldo Nonato de Souza, born in 1966 in Porto Velho, Rondônia, who died in March 2003 in a traffic accident on BR364.
The death certificate was registered at the second registry office in Porto Velho. When Detective Airton requested a copy of the registration from the registry office, he received confirmation that the person had indeed died. Someone had used the identity of a dead man to hire himself as a school bus driver in a city 2000 km away.
Transrotas Norte was ordered to provide copies of all documents relating to Edivaldo’s hiring. The company handed over what it had: a handwritten registration form, a copy of his driver’s license, which would later prove to be fake, and the service contract with the manager’s signature. There was no record of an employment relationship in his work permit, nor a pre-employment medical examination, nor a training course for school transport, which has been mandatory by law since 2004.
The Public Prosecutor’s Office of Amazonas initiated a parallel process against Transrotas Norte for irregularities in the contracting process, but the company ceased operations in February 2008, before the conclusion of the process. The warehouse in the industrial district was returned. The buses were sold. Nelson Praia Monteiro moved to Boa Vista and could no longer be located.
Secretary Ivone Sarmento continued living in Manaus, but refused to give a new statement, claiming she had already said everything she knew. The ghost of Edivaldo Nonato de Souza haunted the case for years. Who was he? Why was he using false documents? Where did he go? And the police issued a search and seizure warrant in the name of the real CPF (Brazilian taxpayer ID) and the fictitious name, which was distributed to police stations in Manaus, Porto Velho, Boa Vista, and Belém.
No one matching that description was found. No one with that name appeared in any subsequent records, whether criminal, labor, or electoral. It was as if the person driving the bus that morning had existed for only two months, solely for that role, and then vanished. In 2009, the investigation was closed for the first time due to a lack of material evidence and the absence of suspects.
The Public Prosecutor’s Office of Amazonas appealed, and the case was reopened in 2010, under the supervision of a new prosecutor, Fernanda Lúcia Barroso. Fernanda requested a reconstruction of the bus route, an analysis of the Transrotas Norte contracts, and a hearing with all residents within a 5 km radius of the resort. None of this was fully accomplished.
Em 2012, uma CPI (Comissão Parlamentar de Inquérito) de nível estadual foi instaurada na Assembleia Legislativa do Amazonas para investigar o caso, mas concluiu seus trabalhos após 8 meses sem chegar a uma conclusão. Uma segunda comissão parlamentar de inquérito em 2016 produziu um relatório de 340 páginas recomendando que o caso fosse federalizado.
A recomendação nunca foi seguida. Durante esse tempo, as famílias organizaram vigílias, marchas e petições. Nenhuma delas teve repercussão fora do Amazonas. O arquivamento de 2009 chegou às famílias através de uma carta registrada enviada pelo cartório do 19º DEP. A carta era padronizada com o número do inquérito, a data da decisão e uma linha informando que o procedimento havia sido arquivado por falta de indícios suficientes para a denúncia.
Dalva Cristina Rebolsas recebeu a dela numa terça-feira à tarde enquanto costurava. Ela leu a carta duas vezes, colocou-a em cima da máquina de costura, sentou-se na cadeira e ficou ali sem se mover durante quase uma hora. A seguir, telefonou a Francisca das Chagas, que já tinha recebido a sua. Francisca estava a chorar e disse que não podia acreditar.
A Dalva disse que conseguiria. A Procuradora Fernanda Lúcia Barroso assumiu o caso em janeiro de 2010, depois do recurso do Ministério Público contra a exoneração ter sido aceite. Fernanda tinha 36 anos, era natural de Parentins e tinha-se especializado em crimes contra menores na Escola Superior do Ministério Público em Brasília.
Ela era a pessoa certa para o caso e, talvez por isso, tenha sido escolhida. Mas o que a Fernanda encontrou ao abrir os arquivos do caso não foi animador. Faltavam partes. As fotos da inspeção ao autocarro estavam incompletas. Das 13 originais, apenas nove tinham sido digitalizadas. Faltava um relatório conclusivo da lista de presenças com as marcas de lápis.
Os testemunhos dos professores continham erros de transcrição que a secretária não tinha corrigido. E a ficha de registo de Edivaldo Nonato de Souza na Transrotas Norte tinha desaparecido do arquivo físico. A Fernanda pediu a reconstituição dos autos. O pedido demorou 14 meses a ser parcialmente atendido.
Em 2011, Fernanda obteve autorização para refazer a rota do autocarro com um veículo semelhante, medindo os tempos de viagem desde pontos de visibilidade e secções da rodovia AM010 onde seria possível parar sem ser visto por outros condutores. A reconstrução revelou que, entre a estância balnear de Tarumã e o ponto onde o autocarro foi encontrado, existiam pelo menos quatro estradas de acesso não pavimentadas que conduziam a áreas de floresta densa, estradas de terra utilizadas por madeireiros ilegais e residentes de colonatos irregulares.
Nenhum destes pontos de acesso havia sido verificado nas buscas originais dos bombeiros, pois ficavam fora do raio de 3 km estabelecido inicialmente. A Fernanda solicitou uma nova operação de buscas naquelas zonas. O pedido foi negado pela Secretaria de Segurança Pública por falta de pessoal.
The 2012 CPI (Parliamentary Commission of Inquiry) was established at the request of a state deputy whose nephew attended the same school, although he did not participate in the field trip, right? The commission heard testimony from 28 people over 8 months, including Delegate Airton, Prosecutor Fernanda, representatives from the city hall, employees of the Department of Education, and three of the mothers.
The sessions were open to the public, but there were rarely more than 10 people in the audience. Local newspapers covered the hearings in their inside pages with short notes. TV Amazonas dedicated two reports to the subject, both in the afternoon. Outside of Amazonas, the case was practically unknown. The CPPIRO concluded its work in April 2013 without a formal conclusion, with a report that listed failures in the police service and irregularities in the Transrotas Norte contract, and the need for further investigation.
Recommendations were sent to the state executive branch and never received a response. The second CPI in 2016 was more ambitious. It produced a 340-page report, heard testimony from 51 people, including federal experts and an Interpol specialist on missing persons who was passing through Brasília and agreed to collaborate informally.
The report concluded that the case presented characteristics incompatible with the hypotheses of accident or drowning, that the original investigation had been negligent in critical points, and that federalization—the transfer of the case to the Federal Police and federal justice—was advisable given the demonstrated incapacity of the state authorities.
The recommendation was sent to the Attorney General of the Republic, but it was never heeded. The formal argument was that there were no new elements that justified federalization. The real argument, according to sources within the Public Prosecutor’s Office itself, who spoke to journalists on condition of anonymity, was that nobody wanted to inherit an unsolved case with the potential for political repercussions.
However, the families did what they could. Dalva and Francisca organized the first vigil in October 2008, on the one-year anniversary of the disappearance. They set up a table on the school grounds with 14 candles and 14 photographs. About 50 people showed up. In 2009, there were 60. In 2010, 40. In 2015, 22. In 2020, 9. Fewer people each year.
Each year the faces in the photographs grew more distant, harder to associate with a real child who once ran through those streets. The candles remained the same. Dalva bought them at the same neighborhood store, always white, always size 14. The marches through the city took place three times in 2009, 2012, and 2017. The largest was in 2009.
Approximately 300 people gathered from Parque das Laranjeiras and marched to the headquarters of the Public Security Secretariat in the city center, carrying banners with the children’s faces and demanding the reopening of the case. The secretary did not meet with the group. An aide came down, picked up a letter, and said that the government was paying attention.
The 2012 petition, with 4,000 signatures, was filed with the Legislative Assembly and prompted the second Parliamentary Inquiry Commission. The 2017 petition, with 2,000 signatures, was sent to the Ministry of Justice in Brasília and never received a formal response. At some point between the second Parliamentary Inquiry Commission and the discovery in 2025, the case ceased to be news and became a memory.
A memory that belonged to a neighborhood, to a group of families. There is a school that changed its name in 2014, becoming the Professor Raimundo Nonato Municipal School, and which painted its walls, replaced its gates, and tried, silently, to erase the association with the event. The children who attended the school in 2007 and who did not participate in the field trip grew up knowing that something had happened, but without fully understanding what.
Some changed schools, others stayed and learned not to ask questions. The new teachers were unfamiliar with the history. The old ones didn’t want to remember. The plaque for the Olinda Brito de Souza school was removed and stored in a warehouse belonging to the Department of Education, where it probably still is.
In March 2025, a surveying team from Eletronorte was conducting measurements for the installation of a new transmission line in the rural area north of Manaus, passing through a strip of secondary forest 7 km from the old Tarumã bathing resort. The technician in charge, Jonas Rebelo de Araújo, noticed that the terrain presented an irregularity in the relief, a rectangular depression covered with low vegetation, different from the natural pattern of the Amazonian soil.
Upon reporting the anomaly to their supervisor, the team requested a probe. What they found at a depth of 1.40 meters interrupted work indefinitely and led to the Federal Police being contacted within 24 hours. Jonas Rebelo de Araújo was 29 years old and had worked at Eletronorte since 2021. He was unaware of the case involving the children from Parque das Laranjeiras.
“He wasn’t from Manaus, he was from Santarém, he came to the capital to study at IFAM and stayed. That day in March, I was with three colleagues carrying out a topographic survey of a 20 km corridor where the transmission line would pass.” The work was technical and repetitive, involving measurements with a total station, marking points and analyzing the terrain profile.
The area where the anomaly was found was located in a secondary forest, vegetation that grew after previous deforestation, probably in the 1990s, and which now formed a dense corridor of thin trees and shrubs. The depression in the ground measured approximately 6 m long by 2.5 m wide. Its shape was regular, with defined contours, and the vegetation covering it was different from the surrounding area, low and less dense, as if the soil there had been disturbed at some point and the forest had not been able to fully recover.
Jonas noted the anomaly in his field report and took four photos with his cell phone. That night, at the hotel in Manaus, he showed the photos to his supervisor, who said it could be an old clay extraction pit or a disused septic tank. But Jonas didn’t think so. The shape was too regular. The depth of the depression, visible to the naked eye, suggested that something had been deliberately excavated.
The research was authorized two days later. A compact backhoe performed the initial excavation, removing the vegetation layer and the first 60 cm of soil. At 80 cm, the operator noticed that the earth was changing color. It went from the typical reddish-brown color of Amazonian soil to a darker, almost black shade, with a different consistency.
“At 120 cm, the mechanical shovel brought to the surface a fragment of synthetic fabric, deteriorated but recognizable. It was blue, with a striped pattern reminiscent of school backpack fabric. The operator stopped the machine. Jonas went down to the ditch, looked at the fragment, looked at the ground around him, and asked everyone to stop.”
The Federal Police in Manaus received the notification at 4 PM that same day. A team from the Regional Superintendency arrived at the scene the following morning, accompanied by experts from the National Institute of Criminalistics and a forensic anthropologist from the Amazonas Institute of Legal Medicine. The area was cordoned off within a 50-meter radius.
“The excavations were then carried out manually, using a brush and a trowel, following the forensic exhumation protocol. Over the next 10 days, the team recovered fragments of tissue from the soil, pieces of rubber consistent with children’s shoe soles, remnants of oxidized metal that appeared to be buckles or fasteners, and organic material in an advanced state of decomposition, which was sent to the forensic genetics laboratory in Brasília for DNA analysis.”
The maximum excavated depth was 1.80 meters. The depression contained material distributed in layers, as if it had been filled in stages. There was no coffin, no shroud, and no containment structure. The material was directly on the ground, mixed with soil and roots. The preliminary report from the Federal Police, issued in April 2025, confirmed that the material found was consistent with human remains of multiple individuals of short stature and that the fragments of fabric and footwear were compatible with children’s clothing from the period between 2005 and 2010. However, the report did not confirm the victims’ identities. DNA analysis would take months and has not yet established the cause of death. But for those following the case, 18 years had already passed. The report said enough. The news reached the families even before it was published in the press.
The Federal Police sent agents to Parque das Laranjeiras to personally inform the parents registered in the investigation that material possibly related to the case had been found. Dalva Cristina Rebolsas received the visit on a Saturday morning. Two federal agents, a man and a woman, sat in her living room and explained, in careful terms, that traces had been found in an area near the Tarumã resort and that it would be necessary to collect genetic material from family members for comparison. Dalva listened without interrupting, and when the agents finished, she asked only one thing: “Did you find glasses?” Dalva’s son, Bruno Henrique, wore glasses. The agents said they could not give details about the objects found. Dalva nodded and signed the consent form for DNA collection. Francisca das Chagas Oliveira received a visit soon after.
She reacted differently. She didn’t ask anything, she didn’t sign anything that day. She asked the agents to return on Monday when her husband was home. The agents returned. Francisca signed, but didn’t say a word during the entire procedure. The national press covered the discovery prominently for about two weeks.
The Jornal Nacional dedicated 3 minutes to the case. Fantástico produced an 8-minute report, with aerial images of the excavation site and interviews with family members. For the first time in 18 years, the case of the children from Parque das Laranjeiras was discussed outside of Amazonas. Social media amplified the story. Theories, speculations, and accusations emerged.
None of this changed the course of the investigation, which continued at its usual pace. From the point of view of forensic science, it is slow, methodical, and silent. Eighteen years separate the morning of October 2007 from the discovery in March 2025. In that interval, six of the 14 families moved from Parque das Laranjeiras.
Two mothers died, one from cancer and the other from a heart attack. A father was hospitalized with a persistent psychiatric problem. The siblings of the missing children grew up carrying a burden that no report describes and no compensation can alleviate. Dalva Cristina Rebolsas, the first mother to call the school that October night, continues to live in the same house, on the same street, with her son’s room exactly as it was.
When news of the discovery in Tarumã arrived, Dalva didn’t turn on the television; she sat in a chair on the porch and watched the street until it got dark. Bruno Henrique’s room has remained untouched since October 12, 2007. The single bed with the Spider-Man bedspread, the ceiling fan that makes noise when turned on, the Monica’s Gang comic books.
Monica’s belongings are piled on the plywood shelf her grandfather made before he died. The spare pair of glasses Dalva bought at the downtown optician in September 2007, which she never used, is kept in its case in the bedside table drawer. The bedroom door is open during the day. Dalva walks through it several times, without even going from the living room to the kitchen, or from the kitchen to the backyard.
She doesn’t go in, she doesn’t close it, she doesn’t touch it. Francisca das Chagas Oliveira moved from Parque das Laranjeiras in 2013, after her husband was transferred to a factory in the industrial district. And they rented a house in Novo Aleixo. Francisca told a friend that she couldn’t stand passing by the school every day anymore, that whenever she saw uniformed children on the sidewalk, her whole body froze, that she needed to go to a place where nobody knew her.
In Novo Aleixo, Francisca began attending a Catholic church and participating in a group of bereaved mothers that met on Wednesday nights in the parish hall. The group had no official name. The women sat in a circle, drank coffee, and talked about their children. Some had lost children to violence, others to illness.
“Francisca was the only one who didn’t know what had happened to hers. Once she told the group that the worst thing wasn’t the pain, it was the uncertainty. The pain had a place in the body. The uncertainty didn’t stay anywhere and, at the same time, it was everywhere. Neusa Figueira, the lunch lady who was the mother of twins Lucas and Luan, never left Parque das Laranjeiras again.”
She continued working at the school until 2014, when her name was changed and her contract was not renewed. No one gave a formal explanation. Neusa realized that the new administration didn’t want anyone around to remind people of what had happened. After that, Neusa worked as a cook in two neighborhood snack bars, without a formal contract, until her health no longer allowed it.
In 2019, N. was diagnosed with severe depression and began receiving the Continuous Benefit Payment (BPC) from the INSS (National Institute of Social Security). She lived alone. Her husband had left home in 2011. Neighbors said that Neusa almost never left the house and that sometimes they heard her talking to herself in a low voice, as if she were conversing with someone who wasn’t there.
Raimundo Monteiro Duarte, the bricklayer’s assistant, who only learned of his son’s disappearance two days later, was admitted to the Eduardo Ribeiro psychiatric hospital in Manaus in 2015 with a case of psychosis that doctors associated with prolonged and unprocessed grief. Raimundo had spent years searching for his son on his own, entering wooded areas on weekends, exploring trails and streams in the rural area of Manaus, asking riverside dwellers if anyone had seen anything in October 2007. He never found anything, never received any leads, but he didn’t give up. His wife said that he would leave home on Saturday morning and return on Sunday night, dirty, exhausted, and sometimes with a fever. One day he didn’t return on Sunday. He was found on Tuesday, sitting in a clearing near the tarumã tree, saying things that didn’t make sense.
He didn’t recognize the woman when she arrived. He was hospitalized that week and never received a full discharge. The siblings of the missing children, those who remained, formed a silent web among themselves that no newspaper reported on. They grew up together in the same neighborhood, on the same streets, and shared something that no outsider could fully comprehend.
They knew what it was like to come home and see their mother staring into space. They knew what it was like to hear their father talk about their brother in the present tense, not as if he were coming back at any moment. They knew what October 12th meant to the rest of the country, Children’s Day. For them, it was the day the family split in two.
Some of these siblings grew into functional adults with jobs, families, and routines. Others did not. Two of them dropped out of school before finishing high school. One had problems with alcohol. One went to live with an aunt in Belém to escape it all. None of them speak publicly about the case. None of them need to. The discovery in March 2025 reopened wounds that had never healed because they had never been treated.
“The families who still lived in Manaus were called to meetings with the Federal Police, the Federal Public Prosecutor’s Office, and the Federal Public Defender’s Office. They were informed that the DNA analysis could take six months to a year. They were also informed that, even with a positive identification, determining the cause of death depended on the state of preservation of the remains, which, after 18 years in the Amazonian soil, might be insufficient for definitive conclusions.”
In other words, they were informed that the answer they had been waiting for for 18 years might come incomplete, or not at all. Dalva heard all this in a meeting room at the Federal Police headquarters, sitting in a plastic chair with her wallet on her lap and dry eyes. Next to her was Francisca. Behind them was Neusa, who had been brought by a niece.
At the back of the room were three adult siblings of the missing children who had asked to watch but didn’t want to sit at the table. The federal prosecutor, who led the meeting, spoke for 40 minutes using technical terms and legal caution. At the end, he asked if anyone had any questions. Dalva raised her hand and asked the same question she had asked the agents in her living room.
“Did you find any eyeglasses?” The prosecutor looked at the papers and said he couldn’t confirm the specific items recovered at the scene. Dalva lowered her hand and didn’t ask anything else. Outside the police station, a reporter from a news website was waiting with a cameraman. When the families came out, he approached Dalva and asked how she was feeling.
Dalva looked at him, looked at the camera, and said, “I feel the same way I’ve felt for 18 years.” Then she turned her back and left. Marinalva dos Santos Queiróz died in 2021 of natural causes in a public hospital in Manaus. She was 66 years old. She was never formally charged in the case. She never changed her last statements, of which she remembered nothing.
In recent years, he lived a reclusive life, never leaving his house and receiving no visitors other than his sister, who lived in the São José neighborhood. Neighbors in the Francisca Mendes housing complex said that Marinalva aged rapidly after 2007, stopped going to church, and kept her window curtains always closed.
When she died, six people attended the funeral. None of the children’s families attended. No one expected them to. Aparecida Lima Fonseca is still alive. She lives in Manaus, at an address that is not public. She has never given interviews and has never been indicted. Her aversion, “I was sleeping on the bus,” remains in the records as the last formal record of her words about the case.
Her husband, Wellington, left his job in 2008 due to family health reasons and never returned. What happened inside that house between 2007 and the present day, nobody knows. Nobody asked. Nobody has the right to demand that you speak. But Aparecida’s silence continues to be, for the families and for those who followed the case, one of the most difficult elements to accept.
The case of the children from Parque das Laranjeiras remains open at the Federal Police. DNA analysis of the material found in March 2025 is underway at the Institute’s forensic genetics laboratory. The National Institute of Criminalistics, in Brasília, has a new investigation. There is no official deadline for its conclusion.
There are no formal suspects and no explanation that connects all the dots. The list with the pencil marks, the bus that deviated 11 km from its route, the teachers on the road, the phantom driver, the dogs that found no tracks, the material buried 7 km from the ranch. Each piece of the case points in a different direction.
None of them complete the picture. Years passed and Parque das Laranjeiras remained what it always was. A neighborhood of simple people, of streets where children play until dark, of houses with half-open doors and the sound of the television escaping through the window. The school changed its name, but the promenade is the same.
The bus stop on the corner is the same. The mango trees that provided shade at dismissal time are still there, older, thicker, their roots lifting the asphalt. Anyone passing by today who doesn’t know the history sees nothing different. It’s just a neighborhood, just a school, just a street like so many others in Manaus, in any city in the world.
In Brazil, life goes on because there is no alternative. But for those who lived through what happened on October 12, 2007, that part of the city was never the same again. There is a fracture that doesn’t appear on the map, that isn’t mentioned in any report, that no CPRI documented. It is the fracture that remains within those who woke up one day.
“And the son wasn’t in bed, the slipper wasn’t near the door, the backpack wasn’t on the living room floor. It’s the space that opens up between the last image – the child getting on the bus, waving from the window, disappearing around the bend – and the silence that followed. 18 years of silence. 18 years of a phone that doesn’t ring.”
And for 18 years, a door has opened every night with the same hope and closed with the same emptiness. The families who waited didn’t ask for much; they asked for an answer. They asked for someone to look at them and tell them what had happened. They didn’t ask for grand justice, they didn’t ask for tributes. They asked for the minimum that any father or mother asks for when their child doesn’t come home.
To know, simply to know. And for 18 years, the answer didn’t come. Perhaps it will arrive now, incomplete, translated into reports, genetic codes, and terms that no one in that neighborhood uses in their daily lives. Or perhaps it won’t arrive at all. Perhaps one day the case will close with a technical report that explains one part and leaves the other in the dark, as happens with so many other cases in this country, where people disappear more easily than one imagines and where the memory of those who remain is often the only proof that the person existed. Today, when we know where the people we love are, when we send a message and get a reply, when we hear the key in the lock and know who it is, it’s easy to forget what it means not to know, what it means to look at the door and have no certainty. What does it mean to carry a photo in your pocket for years? To show it to strangers, to ask if they saw it, if they heard it, if they know anything about it.
What does it mean to lie in bed at night and not be able to close your eyes properly because your mind won’t shut off, because the questions won’t stop, because the silence of the room is the same silence as the phone, the police station, the school, with all the apparatus that should have protected you but didn’t? The story doesn’t end with an explanation; it ends as most true stories end, with questions that remain unanswered, with people who have learned to live with what they never managed to solve, and with the silent certainty that somewhere in this country, at this very moment, someone is still waiting. No.
